Technical ReportPDF Available

Patient safety interventions and their implementation: Current status and directions for the future

Authors:
  • Swiss Patient Safety Foundation

Abstract and Figures

In order to be able to launch successful improvement programmes at national level, it is important to select the urgent topics in terms of content and to design adequate methods for improvement programmes with which changes in the healthcare system can be initiated as broadly as possible and implemented sustainably. In the present project, future programmes are to be prepared thematically and methodologically. Thus, this project aimed (1) to identify potentially effective fields of action and patient safety measures for national quality improvement projects from a content-thematic perspective, (2) to provide an overview of implementation practices and to assess their usefulness for the Swiss context, (3) to evaluate the progress! pilot programmes, and (4) to outline possible methods and contents of future patient safety intervention programmes.
Content may be subject to copyright.
Zurich, April 2021
Patient safety interventions and their implementation:
Current status and directions for the future
For the attention of the Swiss Federal Office of Public Health
Project-Nr. 736.2-6/8
Dr. Annemarie Fridrich; Dr. Yvonne Pfeiffer; Andrea Niederhauser; Prof. Dr. David Schwappach
Contact:
Dr. Annemarie Fridrich
Swiss Patient Safety Foundation
+41 43 244 14 89
fridrich@patientensicherheit.ch
Recommended citation:
Fridrich A, Pfeiffer Y, Niederhauser A., & Schwappach D. Patient safety interventions and their
implementation: Current status and directions for the future. Swiss Patient Safety Foundation. Zurich: 2021.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch II
Content
List of Figures ............................................................................................................................................... III
List of Tables ............................................................................................................................................... IV
Abbreviations ................................................................................................................................................ V
Executive summary ....................................................................................................................................... 1
Introduction .................................................................................................................................................... 4
1 InterventionsProblem areasPopulations ................................................................................. 5
1.1 National report on the quality and safety of healthcare in Switzerland ............................................ 5
1.2 CIRRNETdatabase ........................................................................................................................ 6
1.3 Search for activities/action plans in other countries ......................................................................... 7
1.4 Literature research ........................................................................................................................... 7
1.5 Innovative and underrepresented topics ........................................................................................ 11
1.6 Conclusions for future programmes regarding topics and interventions ....................................... 12
2 Implementing practices: relevant concepts, steps and tools ................................................. 15
2.1 Different approaches to achieve adoption ..................................................................................... 16
2.2 Effectiveness vs. implementation research: hybrid design of programmes ................................... 17
2.3 Phases in implementing evidence-based interventions ................................................................. 20
2.4 Recommended for supporting implementation: ImpRes tool ......................................................... 21
2.5 Choosing an implementation framework: recommendation and selection criteria ........................ 25
2.6 Developing implementation strategies ........................................................................................... 29
2.7 Logic model of the intervention and its implementation ................................................................. 31
2.8 The role of evidence in implementing evidence-based practices .................................................. 32
3 Review and assessment of the progress! programmes .......................................................... 35
3.1 Basis and structure of the progress! programmes ......................................................................... 35
3.2 Reflection ....................................................................................................................................... 36
3.3 Conclusions and recommendations for future pilot programmes .................................................. 41
4 Directions for future improvement programmes ...................................................................... 43
4.1 The current state of improvement efforts in healthcare ................................................................. 43
4.2 Designing large-scale programmes: scenarios for three kinds of interventions ............................ 46
4.3 Suggested topics for future national programmes in Switzerland ................................................. 50
Literature ...................................................................................................................................................... 54
Appendices .................................................................................................................................................. 59
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch III
List of Figures
Figure 1. Absolute frequencies of forwarded incident reports to the CIRRNET database in 2019 per incident
type according to WHO categorisation (https://www.who.int/patientsafety/taxonomy/icps_full_report.pdf) ...6
Figure 2. Absolute frequencies of included review studies per incident type according to WHO categorisation
(https://www.who.int/patientsafety/taxonomy/icps_full_report.pdf) .................................................................8
Figure 3. Absolute frequencies of included review studies per implementation strategy (categories for
implementation strategies based on Waltz TJ, Powell BJ, Matthieu MM, et al. Use of concept mapping to
characterize relationships among implementation strategies and assess their feasibility and importance:
results from the Expert Recommendations for Implementing Change (ERIC) study. Implement Sci.
2015;10(1):109. doi:10.1186/s13012-015-0295-0) .........................................................................................9
Figure 4. Traditional translational pipeline from preintervention, efficacy, effectiveness, and dissemination
and implementation studies, from Brown CH, Curran G, Palinkas LA, et al. An Overview of Research and
Evaluation Designs for Dissemination and Implementation. Annu Rev Public Health 2017;38:122.
doi:10.1146/annurev-publhealth-031816-044215, published under the terms of the Creative Commons
Attribution-ShareAlike 4.0 International License (CC-BY-SA 4.0; https://creativecommons.org/licenses/by-
sa/4.0) ........................................................................................................................................................... 17
Figure 5. Three aims of the use of theoretical approaches in implementation science and the five categories
of theories, models and frameworks, from Nilsen P. Making sense of implementation theories, models and
frameworks. Implement Sci 2015;10:53. doi:10.1186/s13012-015-0242-0, published under the terms of the
Creative Commons Attribution License (CC BY 4.0; https://creativecommons.org/licenses/by/4.0/). ......... 25
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch IV
List of Tables
Table 1. Effectiveness vs. implementation trials in improvement programmes ........................................... 18
Table 2. Implementation strategy examples (based on Perry CK, Damschroder LJ, Hemler JR, Woodson TT,
Ono SS, Cohen DJ. Specifying and comparing implementation strategies across seven large implementation
interventions: a practical application of theory. Implement Sci. 2019;14(1):32. doi:10.1186/s13012-019-0876-
4) ................................................................................................................................................................... 30
Table 3. The implementation process of three different kinds of interventions ............................................ 48
Table 4. Range of relevant problem areas for Switzerland categorised by the relevance of the problem (major
vs. moderate) and the availability and feasibility of solutions (no or little vs. practicable solutions available).
...................................................................................................................................................................... 51
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch V
Abbreviations
AHRQ Agency for Healthcare Research and Quality
ANQ Swiss National Association for Quality Development in Hospitals and Clinics
EPIS Exploration, Preparation, Implementation, Sustainment
ERIC Expert Recommendations for Implementing Change
IHI Institute for Healthcare Improvement
ImpRes Implementation Science Research Development
MoMo Morbidity and Mortality Conferences
NPHS National Programme for HIV and Other Sexually Transmitted Infections
NR National Report on the Quality and Safety of Healthcare in Switzerland
PBM Patient Blood Management
SIRC Society for Implementation Research Collaboration
WHO World Health Organization
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 1
Executive summary
In order to be able to launch successful improvement programmes at national level, it is important to select
the urgent topics in terms of content and to design adequate methods for improvement programmes with
which changes in the healthcare system can be initiated as broadly as possible and implemented sustaina-
bly. In the present project, future programmes are to be prepared thematically and methodologically. Thus,
this project aimed (1) to identify potentially effective fields of action and patient safety measures for national
quality improvement projects from a content-thematic perspective, (2) to provide an overview of implemen-
tation practices and to assess their usefulness for the Swiss context, (3) to evaluate the progress! pilot pro-
grammes, and (4) to outline possible methods and contents of future patient safety intervention programmes.
The first aim of the project was to identify relevant topics and potentially effective patient safety actions for
national intervention projects from a content-thematic perspective. In order to simultaneously incorporate
content from science and practice, international and national findings, as well as established and innovative
concepts, we used the following sources of information: The National Report on the Quality and Safety of
Healthcare in Switzerland, the CIRRNET database, activities/action plans in other countries, literature re-
search, and the expert knowledge of the Swiss Patient Safety Foundation. It was not possible to compile a
list of ready-made interventions to be implemented on a 1:1 basis, as the effectiveness of interventions is
very much dependent on the context and the evaluation of promising interventions often has shortcomings.
For example, methodological design is often poor, often it is not clear which parts of multi-component inter-
ventions are effective, information on cost-benefit evaluation is often missing, and long-term results are rarely
reported. Moreover, there is often little empirical evidence for the Swiss context. Across all sources, one
problem area emerged as particularly dominant: medication safety with the two sub-topics prescription and
administration mentioned particularly frequently. Most studies focus on acute hospital care and aim primarily
at changing individual clinician behaviour. This contrasts with national and international recommendations
that also identify for example ambulatory care and mental health care as priority areas for action and em-
phasise the effectiveness of more systems-oriented measures. Further information on the most frequently
identified interventions, problem areas, and populations are presented in Chapter one.
The second aim of the project was to explore relevant concepts, steps and tools for implementing future
safety improvement programmes in Swiss healthcare. There are many steps to be taken starting from an
initial idea, for example the identification of a useful best practice, which could be developed into an inter-
vention, to its system-wide adoption: the practice needs to be thoroughly tested for its effectiveness in terms
of improving clinical outcomes, and potential negative side effects need to be considered. In a next step,
system-wide implementation needs to be prepared in identifying the most promising way of implementing
the intervention in various contexts. Implementation quality needs to be rigorously assessed, for being able
to compare implementation approaches. In a final step, evaluation on various dimensions can be carried out,
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 2
evaluating implementation and clinical effectiveness to draw conclusions for future improvement efforts. The
goal of this whole process is a sustainable and long-term implementation, and, ultimately, the improvement
of the safety and quality of the care delivered to patients. We identified a highly useful tool, which is designed
to support efforts to sustainably implement evidence-based interventions: ImpRes (Implementation Science
Research Development Tool). It supports many important decisions and concentrates guidance and tools to
take informed decisions along the implementation process. We recommend using the EPIS (Exploration,
Preparation, Implementation, Sustainment) framework in future programmes/implementation projects to
structure the implementation process, factor important concepts, and to gain an understanding of the pro-
cesses and influences crucial to implementation efforts. Chapter two discusses in detail the requirements for
national improvement programmes, relevant tools, models and approaches that are recommended for future
use.
Since 2012, the Swiss Patient Safety Foundation has been mandated by the Federal Office of Public Health
to develop and conduct pilot programmes with the aim of improving patient safety in healthcare: the
“progress!” programmes. For this project, the results and findings from the three completed pilot programmes
progress! have been reviewed; experiences from the two current programmes have been integrated if
appropriate. Furthermore, the pilot hospitals of the first three progress! programmes were contacted again
by email/telephone. This information was complemented by findings from the literature, the National Report
on the Quality and Safety of Healthcare in Switzerland as well as personal experiences of the internal project
managers of the progress! programmes. Overall, it seems that the expectations and demands placed on the
progress! programmes have moved further and further away from the initial mandate and intention of the
programmes over time. While it was defined at the beginning that the progress! programmes should rely on
evidence-based interventions and therefore focus on implementation outcomes rather than effectiveness
outcomes, it became increasingly clear during the course of the programmes that there is not enough
evidence for patient safety interventions in the Swiss setting and therefore both implementation and
effectiveness need to be considered in such pilot programmes. In addition, a significant conceptual gap
emerged: namely, that it was not defined how and by whom the findings from the pilot programmes were to
be implemented at national level. We are convinced that the conceptual weaknesses of the progress!
programmes can be overcome in future programmes through a revised and methodologically sound
programme concept and the appropriate time and financial resources. Chapter three provides
recommendations that should be incorporated into the design of future improvement programmes.
In the last two decades, there has been a great effort to improve healthcare quality globally. However, there
is only little evidence that preventable patient harm has been considerably reduced. In the current situation
it is difficult to disentangle what the main problem is: a) That we do not have sufficient and robust measure-
ment of the effects of programmes and interventions, or b) that the interventions and programmes imple-
mented are not sufficiently effective. It therefore is necessary to develop measures indicating whether a
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 3
change has happened that complement national adverse event rates. Assessing and measuring progress in
patient safety on various levels, e.g., national, or institutional, is an important undertaking that urgently needs
to be tackled in order to provide the necessary resources for successful improvement programmes. Suc-
cessful implementation needs to rely on a systematic process, addressing barriers and facilitators to reach
adoption of the intervention at the front-end of care.
Chapter four describes the implementation process of three different kinds of interventions to illustrate how
potential programmes for patient safety improvements could be designed. Patient safety interventions tar-
geting actual work practices at the front end (e.g., the safe administration of drugs) can be differentiated from
interventions that aim at increasing competencies of the workforce relevant for safety (e.g., the ability to learn
from errors as an individual and as an organisation). As a third kind of intervention, we illustrate a strong
systems-oriented intervention to improve patient safety.
Furthermore, chapter four shows a range of relevant problem areas for Switzerland, categorised by the rel-
evance of the problem (major vs. moderate) and the availability and feasibility of solutions (no or little vs.
practicable solutions available). Based on the relevance of the problem and the availability, feasibility, and
scope of solutions we give recommendations for action. We recommend approaching several problem areas
from all three problem-solution-combinations at the same time to advance patient safety on a broad scale.
The list covers interventions that aim at making clinical processes safer. For sustainably changing
healthcare, investing in the development of a better safety culture and learning from errors throughout the
whole system is crucial. i.e., programmes targeting meta-competencies should also be conducted.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 4
Introduction
In order to be able to launch successful improvement programmes at national level, it is important on the
one hand to select the urgent topics in terms of content, for example on the basis of a known problem
situation such as catheter-associated infections. On the other hand, it is important to design adequate meth-
ods for improvement programmes with which changes in the healthcare system can be initiated as broadly
as possible and implemented sustainably. With the expanded requirements arising from the KVG revision,
the Foundation has taken on the task of identifying core intervention fields and revising the methodological
foundation of the progress! programmes. The results are presented in this report.
The report is structured according to the following objectives:
1. To identify potentially effective fields of action and patient safety measures for national quality im-
provement projects from a content-thematic perspective.
2. To provide an overview of implementation practices and to assess their usefulness for the Swiss
context.
3. To evaluate the progress! pilot programmes.
4. To outline possible methods and contents of future patient safety intervention programmes.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 5
1 InterventionsProblem areasPopulations
The first aim of the project was to identify relevant topics and potentially effective patient safety actions for
national intervention projects from a content-thematic perspective. There is a continuum of evidence to sup-
port interventions ranging from very objective sources such as systematic reviews over somewhat objective
sources as research studies to more subjective sources such as practitioner reports to very subjective
sources such as expert opinions/personal experience [1]. While very objective sources can be considered
reliable, they often lack innovative approaches, which are more likely to be considered in subjective sources.
To reflect this diversity, we used the following information bases:
- National report on the quality and safety of healthcare in Switzerland
- CIRRNETdatabase
- Activities/action plans in other countries
- Literature research
- Expert knowledge of the Swiss Patient Safety Foundation
1.1 fvolunNational report on the quality and safety of healthcare in Switzerland
In order to obtain an first overview of the situation in Switzerland, the areas of action mentioned in the Na-
tional Report on the Quality and Safety of Healthcare in Switzerland (NR) by Vincent and Staines [2] were
examined. The authors argue that future national programmes should not focus on “the introduction of spe-
cific interventions such as the surgical safety checklist” [p. 47] but on programmes that are “more ambitious
in scope, scale of implementation and sustainability” [p. 47]. Specific issues could still be embedded in large
national programmes, but the programmes should primarily focus on clinical areas or wider themes to be
applicable across the healthcare system. Based on 28 short reports, Vincent and Staines [2] identify four
priority areas for action on a national level and draft recommendations to improve patient safety and quality
of healthcare in these areas:
- Medication safety
- Improving the care of frail older people
- Improving the care of people with mental health problems
- Teamwork for safety and quality
Two of these four priority areas (medication and teamwork) address treatment or treatment/care collabora-
tion, while the other two (frail older people and people with mental health problems) address a specific pop-
ulation. However, it must be noted that these areas are not clearly separable. For example, the topic of
medication safety with a focus on poly-medication plays a particularly important role in the care of frail older
people. The NR thus focuses on identifying the general problem areas relevant to Switzerland. However, it
does not address whether good solutions to these problems already exist or, if so, whether they are suitable
for implementation in the Swiss health system.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 6
1.2 CIRRNETdatabase
The Critical Incident Reporting and Reacting Network (CIRRNET) is a national platform that allows affiliated
healthcare organisations to anonymously and voluntary submit error reports from their local Critical Incident
Reporting Systems (CIRS). We decided to analyse the incident reports forwarded to the CIRRNET database
in 2019 (N=596) to find out which incident types were most frequently reported in 2019. The incident types
were analysed with respect to the subject area of interest and categorised according to the primary incident
types proposed by the World Health Organization (WHO) [3]. Figure 1 shows that a majority of forwarded
incident reports refer to the category of Medication/IV Fluids, followed by the categories Documentation and
Clinical Process/Procedure. Less affected, but still worth mentioning, are the categories Clinical Administra-
tion, Medical Device/Equipment, and Behaviour whereas incidents reports related to Nutrition, Patient Acci-
dents, and Oxygen/Gas/Vapour were only rarely forwarded to the CIRRNET database in 2019.
However, it should be noted that a report could be assigned to more than one category, and that the CIR-
RNET reports are not representative for all areas of healthcare in Switzerland, as the majority of participating
CIRRNET institutions are acute care hospitals; outpatient care and other sectors are therefore underrepre-
sented. Another point that limits the validity is the fact that high reporting rates on a category do not auto-
matically mean that it is the most important issue. It is possible that certain categories are reported on par-
ticularly frequently because they are more likely to be noticed or easier to report than others, but which may
have greater consequences or occur more frequently. It was not possible to categorise these incident reports
into specialties because the majority of reports were categorised as other, followed by the categories non-
specialty and internal medicine intensive care unit. However, this could be an indication that most of the
problems relate to several specialties and therefore could not be specifically assigned.
Figure 1. Absolute frequencies of forwarded incident reports to the CIRRNET database in 2019 per incident type ac-
cording to WHO categorisation (https://www.who.int/patientsafety/taxonomy/icps_full_report.pdf)
0
50
100
150
200
250
300
Note: N=596 incident reports; multiple categorisation possible
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 7
1.3 Search for activities/action plans in other countries
There are several institutions in Europe, but also worldwide, who are very involved in patient safety issues.
Some of them regularly publish publicly available guidelines that express their priorities for patient safety
related interventions, others spotlight priority patient safety issues for the coming years on their websites. In
order to get an overview of the priorities of these actors, we searched the websites of the major actors (no
claim to completeness) and compiled a list of their current core activities and/or focus areas for the coming
years (see Appendix 1).
A quick look at the list reveals two general themes that are common to all: medication safety and healthcare
associated infections. Other topics that are frequently mentioned are improved care for people with mental
health problems, fall prevention and patient/family involvement.
1.4 Literature research
While the NR the CIRRNET database, and the activities/action plans of other countries provide information
on problem areas that seem to be relevant to improve patient safety, we conducted a literature search to find
out which problem areas currently have solutions with sufficient evidence. Since the number of published
studies on patient safety interventions is very large (a simple search with the term patient safety intervention
shows more than 100,000 results), it was necessary to limit the scope. In order to identify fields of action and
interventions with strong evidence on the one hand and recent, innovative approaches on the other, we
conducted two literature reviews with different focuses:
First, we conducted a literature research limited to systematic reviews and meta-analyses published in the
last 10 years. This ensures that interventions and areas of action with high relevance are identified for which
data are available to assess effectiveness. Second, we conducted a literature research limited to studies
published in the first half of 2020. This in turn ensures that we do not disregard innovative approaches.
Literature research with focus on systematic reviews and meta-analyses
We did a PubMed research with the following search query:
(patient safety) AND (systematic review[Title]) ((intervention[Title/Abstract]) OR (quality improve-
ment[Title/Abstract])) AND ("2010"[Date - Publication] : "3000"[Date - Publication])
This search returned 430 results on June 4, 2020, of which 323 had to be excluded because they either
contained exclusively qualitative data or no patient-safety related results, or no abstract was available in
English. The 107 remaining review articles were categorised in terms of population and problem area (safety
target). Additionally, the incident types according to WHO [3] and the implementation strategies (see section
2.6) according to Waltz et al. [4] were used to categorise the review articles (multiple categorisation possible).
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 8
This proved to be difficult, as some reviews refer to the investigation of a specific field of action (e.g., medi-
cation errors), others to the effectiveness of an implementation strategy (e.g., simulation training) and others
to the effectiveness of a specific patient safety intervention (e.g., surgical checklist). Although the review
articles do not always focus on a specific incident type or implementation strategy, the use of the two estab-
lished category systems mentioned above was appropriate to meet the need for a consistent terminology
and to facilitate comparability of results. For example, the CIRRNET reports (see section 1.2) are also cate-
gorised according to WHO. In addition, the number of included studies, main effectiveness statements re-
garding the intervention studied, and information on limitations were extracted from the abstracts. All infor-
mation was compiled in an Excel file (see Appendix 2).
The included reviews covered a median of 18 studies (range, 2297). Most reviews were limited to either a
clinical setting (e.g., acute care hospitals) or a specific population (e.g., older adults). The majority of reviews
focused on studies from the acute care hospital setting (48%). Few reviews focused on an outpatient setting
(7%), intensive care units (7%), or multiple settings simultaneously (5%). Among reviews focusing on specific
populations (20% in total), only paediatrics/children/infants (9%) and older/frail adults (4%) were represented
frequently. In terms of incident type, most studies targeted interventions that were classified in the category
Medication/IV Fluids (26%), followed by the categories Clinical Administration (16%), and Resources/Organ-
isational Management (14%); see Figure 2.
Figure 2. Absolute frequencies of included review studies per incident type according to WHO categorisation
(https://www.who.int/patientsafety/taxonomy/icps_full_report.pdf)
0
5
10
15
20
25
30
Note: N=107 review studies; multiple categorisation possible.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 9
The most frequently targeted problem areas per category were the following:
- Medication/IV Fluids: Prescribing; Administration
- Clinical Administration: Transfer of care; Handover; Discharge
- Resources/Organisational Management: Teamwork; Organisation of teams
- Clinical Process/Procedure: Clinical decision making; Diagnostics
- Behaviour: Safety Culture
- Healthcare Associated Infections: Bloodstream; Urinary drain/tube (Type of infection); Bacteria
(Type of organism)
- Patient Accidents: Falls
- Documentation: Medical records; Checklists
- Nutrition: Feeding (infants)
An overview of the implementation strategies (see Figure 3) analysed in the included reviews shows that
strategies aimed on changing individual behaviour are used more often than strategies that aim to change
than structural or system conditions. In many cases, multi-component interventions targeting several behav-
ioural and/or several structural aspects, also called bundles, are used. Only few reviews focus on studies
examining the effects of patient involvement.
Figure 3. Absolute frequencies of included review studies per implementation strategy (categories for implementation
strategies based on Waltz TJ, Powell BJ, Matthieu MM, et al. Use of concept mapping to characterize relationships
among implementation strategies and assess their feasibility and importance: results from the Expert Recommendations
for Implementing Change (ERIC) study. Implement Sci. 2015;10(1):109. doi:10.1186/s13012-015-0295-0)
0
5
10
15
20
25
30
35
40
45
50
Train and
educate
stakeholders
Multi-component
interventions Support
Clinicians Provide
interactive
assistance
Change
infrastructure Engage patients Use evaluative
and iterative
strategies
Develop
stakeholder
interrelationships
Datenreihen1
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 10
It is remarkable that many reviews conclude that some evidence could be found, but followed by criticism of
the quality the methodology, the study design, the selection of outcomes and reporting methods, the evalu-
ation strategy, neglect of the context, the lack of information on sustainability or long-term results and missing
information on cost-benefit evaluation. They commonly report wide variability in terminology, outcomes,
measurement methods, and local adaptation, and criticise the inadequate description of instruments and
interventions, which makes replication difficult. For multi-component interventions, often only the overall ef-
fectiveness is reported; the information on the extent to which individual components were effective, is miss-
ing or cannot be assessed because of the evaluation strategy.
Literature research with focus on currently published studies
Since the research on reviews and meta-analyses may not include particularly new and innovative ap-
proaches due to the delay in publication, an additional literature search was conducted for interventions
published in the first half of 2020. We did a PubMed research with the following search query:
((patient safety[Title/Abstract]) AND ("2020/01/01"[Date - Publication] : "3000"[Date - Publication]) AND
((intervention[Title/Abstract]) OR (quality improvement[Title/Abstract]))) NOT (review[Title])
This search provided 452 results on July 7, 2020, of which 362 had to be excluded because of the same
exclusion criteria mentioned for the categorisation of the reviews. Further, studies were excluded if no results
were available yet.
The remaining 90 articles were also categorised in terms of population, problem area, incident types [3], and
implementation strategies [4]. In addition, main efficacy statements were extracted from the abstracts. All
information was compiled in an Excel file (see Appendix 3).
The majority of studies focused on interventions in the acute care hospital setting (57%). Some studies
included intensive care units (13%) while only few studies focused on outpatient settings (6%) or specifically
psychiatric settings (3%). Among studies focusing on specific populations (21% in total), only paediatrics/chil-
dren/infants (12%) were represented frequently. Compared to the reviews, the picture is similar, with a
slightly higher representation of studies on paediatric and intensive care units. In terms of incident type, most
studies targeted interventions classified in the category Healthcare Associated Infection (24%), followed by
the categories Clinical Process/Procedure (18%), and Medication/IV Fluids (13%).
With regard to the targeted problem area of the studied interventions, it is remarkable that several studies
were published on optimising the use of alarms/alerts and in the case of healthcare-associated infections,
many studies concern bloodstream infections (incl. sepsis). In addition, there is also a study that addresses
a very recent topic: patient flow in the context of Covid.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 11
Similar to the reviews, the majority of the analysed studies from the first half of 2020 applied implementation
strategies that aimed at changing staff behaviour through training and educating stakeholders (68%), sup-
porting clinicians (9%) or providing interactive assistance (10%).
Overall, the majority of studies reported some improvements in clinical outcomes as well as self-reported
measures of staff/team perceptions, while a few studies reported no significant reductions at all. However, it
must also be considered that studies that do not achieve any effects have a significantly lower probability of
publication, thus negative results may be underrepresented.
1.5 Innovative and underrepresented topics
It is likely that not all relevant topics are covered by the above sources for the following two reasons: a) topics
that are difficult to distinguish or are not primarily assigned to the term patient safety, and b) innovative topics
and difficult topics for which there are few/no solutions to date are underrepresented in the above analyses.
In order to identify these topics as well, the above analyses were compared with the expert knowledge of
the Swiss Patient Safety Foundation.
One issue identified is patient blood management (PBM). This WHO-supported initiative aims to increase
patient safety while reducing costs at the same time. PBM has already proven effective both internationally
[5] and in Switzerland [6]. Nevertheless, it is only being implemented in a few Swiss hospitals. The Alliance
Rouge, an open interest group, promotes the implementation of PBM in Switzerland and provides compre-
hensive information/support for interested hospitals (for further information visit https://alliance-rouge.ch/). A
similar situation is related to nutritional support: Here, a high quality randomized trial conducted in Switzer-
land reported highly-favourable results with decreases in mortality by screening for and treating malnutrition
in inpatients [7]. This intervention would certainly qualify for large scale improvement activities but has not
been labelled a patient safety interventionthough it certainly has such dimensions.
The problem areas and interventions targeted in the reviews and studies reported in section 1.4 do not
sufficiently take into account future conditions, such as increasing digitalisation of the health system. For
example, telemedicine has recently gained importance, especially due to the Corona pandemic. However,
the topic is still rather underrepresented in the reviews presented above; only one review investigated diag-
nostic errors and clinical decision support systems in the context of telemedicine.
The problem of interruptions is examined in a review with regard to medication administration. We would like
to emphasise the relevance of this problem at this point, as interruptions also play an important role in other
areas of patient safety (e.g., medication prescription, handover, etc.). Due to the different settings in which
interruptions play a role, there is not a one fits all- intervention that is particularly promising. Nevertheless,
the problem of interruptions should be taken into account in all fields of patient safety and addressed with
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 12
appropriate interventions. These should include structural approaches in particular as it is empirically proven
that interruptions comprise a serious threat to patient safety throughout the care continuum [8].
In addition, we would like to highlight the topic of clinical decision support. It is addressed in some reviews,
often as part of a bundle of interventions, e.g., with regard to triage, diagnosis, drug dosing and monitoring.
However, we think that the potential of clinical decision support systems, especially with regard to diagnostic
errors, has not yet been sufficiently systematically investigated.
The findings presented here are also in line with the overview on Existing and Emerging Strategies for Re-
ducing Medical Error and Increasing Patient Safety provided by the Agency for Healthcare Research and
Quality (AHRQ) [9].
1.6 Conclusions for future programmes regarding topics and interventions
Across all sources, one problem area emerges as particularly dominant: medication safety. It is identified as
a high priority topic in the NR (see 1.1), there were many reports forwarded to the CIRRNET on medication
safety in 2019 (see 1.2), it is addressed as a priority topic by all international key players (see 1.3), and there
are many reviews/studies on interventions in this area (see 1.4). The sub-topics prescription and administra-
tion are mentioned particularly frequently. Interventions on medication reconciliation supported by clinical
pharmacists have proven to be particularly promising in other countries [1013]. It should be noted though,
that the role, education, expertise, and availability of clinical pharmacists differ across countries and these
results may not be easily transferrable between countries and settings. Electronic tools supporting the pre-
scription tasks are also studied frequently, but reviews report mixed results and often poor methodological
quality of included studies [1416].
In the reviews of interventions to prevent medication administration errors, bundles that included various
actions for training and educating stakeholders were shown to be successful, although it was not possible to
prove which of the individual actions was most successful [17]. One review that examined the effectiveness
of interventions trying to reduce interruptions during medication administration showed weak evidence of the
effectiveness of interventions to significantly reduce interruption rates and medication administration errors
[18] although interruptions have been implicated as a potential contributory factor to medication administra-
tion errors.
Other problem areas that are addressed by many reviews are transfer of care, handover, and discharge at
various interfaces (e.g., intensive care unit to general ward; emergency service providers to emergency
department nurses) which are classified under the main category clinical administration. Most of the inter-
ventions are based on the implementation strategy train and educate stakeholders. However, the methodol-
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 13
ogy and the outcomes examined are very heterogeneous between the studies and the results on effective-
ness are inconsistent. Nevertheless, there seems to be a great interest in improving handovers at various
interfaces.
Although the topics of patient blood management (PBM) and nutritional support do not appear in the reviews
due to the necessarily limited search terms, we recommend paying attention to intervention bundles on PBM
and nutritional support: For their effectiveness, sound evidence has already been reported, internationally
and for Switzerland.
With regard to different populations, it appears that the groups older adults and children are often treated as
specialties. Paediatric topics are treated separately, especially in the empiric studies/reviews. This could be
since children/infants require dosages for medication that deviate from the standard dose, that the overall
case numbers are lower compared to adults, that children/infants are partly unable to provide information
themselves, and that parents are relevant as an additional stakeholder group. The separate treatment of
patient safety issues in relation to older adults is evident in both literature research and the recommendations
of international key players as well as in the NR [2]. Due to the increased likelihood of comorbidities and
poly-medication, and in the case of very high age/high fragility, possibly even decreased for communica-
tion/mental capacity, this seems plausible.
In some cases, special disciplines such as gynaecology or orthopaedics are also considered separately in
studies and recommendations for action. However, in the majority of studies, different disciplines are involved
in interventions to improve patient safety, making it difficult to assess importance or effectiveness by spe-
cialty, as shown, for example, in the CIRRNET analysis.
The analyses presented above have some limitations: The literature research does not represent a system-
atic literature review, as the methods required for this (e.g., two independent raters for review categorisation)
would have exceeded the resources and scope of the project. Further, it must be noted that the selected
sources of information refer to past and present findings, but do not consider future conditions, such as
increasing digitalisation of the health system. In addition, the findings are subject to a certain mainstream
bias, as the focus is on already frequently studied interventions/fields of action and recognised expert organ-
isations; innovative and rare approaches as well as specific contextual conditions may thus remain un-
derrepresented. Although the work reported above does not claim to be complete, the following key findings
can be derived:
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 14
Key finding 1:
It is not possible to compile a list of ready-made interventions to be implemented on a 1:1 basis, as the ef-
fectiveness of interventions is very much dependent on the context and the evaluation of promising inter-
ventions often has shortcomings. For example, methodological design is often poor, information on cost-
benefit evaluation is often missing, and long-term results are rarely reported. Moreover, there is often little
empirical evidence for the Swiss context.
Key finding 2: The greatest discrepancy in patient safety issues is in care for patients with mental health
problems. This topic area is often described as a priorityboth internationally [19] in the Swiss context [20].
At the same time, however, there are only a few reviews and current studies, which show ambiguous results
regarding the effectiveness of quality improvement interventions, as for example interventions for involuntary
psychiatric patients.
Key finding 3: With regard to sectors, studies and recommendations for action relating to acute hospital
care are predominant, followed by the area of emergency/intensive care. Outpatient settings including out-
patient care in hospitals, and GP practices, as well as nursing homes and care at home are mentioned as
priority areas by expert commissions, but the number of studies is still rather low.
Key finding 4: With regard to empirical studies, it is noticeable that bundles/multi-components are often
investigated, but the effectiveness of the individual components is rarely evaluated, since the various com-
ponents are implemented simultaneously. It is therefore rarely known what the mechanisms of action are,
and which components are most important for positive effects.
Key finding 5: Most improvement programmes focus on changing individual clinician behaviour rather than
structural or system conditions although changing individual behaviour (e.g., training, use of warnings) seem
to be less effective than system-level changes (e.g., automating a safety check, forcing functions) [21]. This
seems a rather weak strategy in general. For example, ensuring access to a shared, electronic medication
and patient record may be much more effective in ensuring medication safety than training staff in obtaining
a medication history individually. Also, in view of the high turnover in the nursing sector and the constant
rotation of physicians, person-based corrective actions must be viewed critically with regard to the long-term
effectiveness of interventions.
Conclusion:
The findings provide helpful indications regarding relevant problem areas and promising measures/bundles,
but also point out knowledge gaps. Where promising solutions already exist, they will need to be adapted
and tested to the Swiss context in a pilot test to determine whether and, if so, in what form they can be best
implemented and whether they are effective and cost-efficient. Section 4.3 will provide recommendations on
specific areas of action that should be targeted in near future.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 15
2 Implementing practices: relevant concepts, steps and tools
In this chapter, we first explain different approaches to achieve adoption as well as the phases of an imple-
mentation project. We then present the results of our work selecting useful tools and concepts for imple-
menting future safety improvement programmes in Swiss healthcare. There exists a large host of frame-
works, concepts, and tools centred around implementing interventions that were screened in a large litera-
ture review and evaluated for their usefulness for future programmes, involving evaluations of internal and
international experts. A particular focus is laid on a selected tool and framework, the choice of implementation
strategies and the need for evaluating the impact of a programme in assessing implementation outcomes.
Well-designed interventions to improve the quality and safety of clinical processes can only unfold their in-
tended effects if they are fully implemented into the work contexts they apply to: Clinicians not taking up new
guidelines, healthcare professionals taking shortcuts, potentially missing important knowledge of how to ap-
ply the intervention’s main procedures – there are many obstacles that may hinder the successful implemen-
tation of originally well-planned and -intentioned improvement interventions. Recent literature has reported
numerous examples of implementations failing to reach their intended improvements of clinical outcomes or
successful implementations failing to be replicated in other institutions or countries [2224].
Implementation research is an important domain of expertise to make sure that large-scale, e.g., national,
improvement efforts are effective in improving care. There are many steps to be taken starting from an initial
idea, for example the identification of a useful ‘best practice’, which could be developed into an intervention,
to its system-wide adoption: The practice needs to be thoroughly tested for its effectiveness in terms of
improving clinical outcomes, and potential negative side effects need to be considered. In a next step, sys-
tem-wide implementation needs to be prepared in identifying the most promising way of implementing the
intervention in various contexts. In Switzerland, this may affect the different language regions, different sizes
of hospitals, different clinical specialties, or different regional laws, for example. It is paramount to assure
that implementation is successful, so that the intervention is widely and sustainably adopted by the
healthcare workers. Thus, implementation quality needs to be rigorously assessed, for being able to compare
implementation approaches. Sustainably scaling up a best practice to national implementation means to
continue assessing intervention effects as well as implementation quality at least in some institutions, in
order to be able to evaluate long-term effectiveness of the intervention and in order to make sure that the
adoption of a new practice is indeed kept up [25,26]. Thus, in a final step, evaluation on various dimensions
can be carried out, evaluating implementation and clinical effectiveness to draw conclusions for future im-
provement efforts. The goal of this whole process is a sustainable and long-term implementation, and, ulti-
mately, the improvement of the safety and quality of the care delivered to patients.
The basic process that was outlined here will be discussed in detail in the following and its requirements for
national improvement programmes illustrated. Furthermore, the relevant tools, models and approaches that
were selected to be recommended for future use are presented.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 16
2.1 Different approaches to achieve adoption
All approaches of bringing new practices to the frontline of healthcare work have the goal to reach adop-
tion, also called uptake, by the practitioners. Two approaches are often differentiated: dissemination and
implementation. There is one important definitory aspect that may cause confusion in understanding the
different approaches: Both approaches aim at implementation of a practice as their goal, while only one
approach is called implementation. Thus, dissemination is also a way to generate the uptake of new prac-
tices, i.e., to reach implementation of a practice in healthcare settings. With this in mind, the differences
between the approaches are described in the following:
Large-scale implementation efforts can be
- dissemination-oriented: Aiming at the voluntary uptake of an intervention in raising awareness to the
problem, engaging important stakeholders, and showcasing positive examples, e.g., along the idea of
the breakthrough series collaboratives [27,28]. The intent of dissemination is spreading knowledge
and materials like toolkits [29] to increase the voluntary uptake of evidence-based practices [30]. Prior
pilot programmes in Switzerland adopting this kind of approach have been criticized for lack of evi-
dence of their effectiveness by the NR [2]. Dissemination strategies mainly rely on communication
efforts [31] (media, conferences, etc. [32]) portraying the benefits of the intervention from example
institutions. Traditionally, in medicine, dissemination is applied to alter existing practice and bring new,
evidence-based clinical practices [33] to the frontline, for example when a certain professional asso-
ciation issue a new guideline recommending a new, improved way of conducting a certain aspect of
their clinical work.
- or implementation-oriented: Aiming at systematically integrating a practice in a certain healthcare
setting, applying orchestrated, even directive, action such as incentives or policy-based reinforcing
mechanisms. For example, a strong form of implementation would be to integrate the adoption of a
certain practice in the contracts between providers and assurances. Implementation uses specific
strategies to integrate evidence-based practices systematically into routine practice (see section 2.6
on strategies). For these large-scale improvement programmes, implementation research plays a ma-
jor role, because tracking evidence of effectiveness and implementation quality is crucial to support
and guide the journey to full adoption.
For example, a surgical unit that has adopted the safe surgery checklist may have done so because a sur-
geon attended a talk on the effectiveness of the checklist and then used tools provided to implement the
practice (what would be a dissemination programme) or because there is a national programme supporting
and incentivizing the application of the checklist, its use becoming even part of contracts (what would be an
implementation programme).If new practices are performed in one or few places and they spread for exam-
ple by word-of-mouth to other places, this is called diffusion. Diffusion is not actively managed but describes
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 17
a process of a new practice being adopted without any centralised coordinating. In this report, we do not
elaborate on the diffusion of safety interventions.
2.2 Effectiveness vs. implementation research: hybrid design of programmes
Despite the need for action on improving the safety of healthcare being quite pressing, any action that will
be taken has to rely on sound evidence that justifies the investment of major resources in a national pro-
gramme. Additionally, the NR [2] demanded future programmes to provide robust evidence of their effective-
ness. Therefore, gathering, evaluating and producing evidence will be a major pillar of future programmes in
every phase along the lifecycle of a programme.
As a first clarification, it is important to differentiate between effectiveness and implementation research [34].
Effectiveness research investigates whether and how a specific intervention influences the targeted out-
comes in applying scientific methods. Issues studied in effectiveness studies are centred around the question
of whether a certain programme or intervention does work. Prior to that, there may be a phase in which a
new intervention is developed from scratch, if no prior evidence or developments exist (see Figure 4: prein-
tervention phase/efficacy studies).
Figure 4. Traditional translational pipeline from preintervention, efficacy, effectiveness, and dissemination and imple-
mentation studies, from Brown CH, Curran G, Palinkas LA, et al. An Overview of Research and Evaluation Designs for
Dissemination and Implementation. Annu Rev Public Health 2017;38:122. doi:10.1146/annurev-publhealth-031816-
044215, published under the terms of the Creative Commons Attribution-ShareAlike 4.0 International License (CC-BY-
SA 4.0; https://creativecommons.org/licenses/by-sa/4.0)
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 18
In contrast, the question asked in implementation research is how to make a programme work in the poten-
tially multiple contexts it applies to after its effectiveness has generally been confirmed in a limited set of
contexts. Thus, implementation research starts when the effectiveness of an intervention has been estab-
lished. In this phase, different implementation strategies may be tested within and across institutions in order
to best prepare for larger scale-up. The yellow box in figure 4 explains the different phases of implementation
research. They are based on the implementation framework (EPIS [35]) that is presented in section 2.5.
Designing a programme for quality or safety improvement usually involves both, effectiveness and imple-
mentation research resulting in project designs that are called hybrid designs [26]. For each programme to
be implemented system-wide, the existing level of evidence about its effectiveness and about appropriate
implementation strategies defines the extent to which effectiveness and implementation research is required.
Recommendation: Define for each programme or intervention the need for effec-
tiveness and implementation research based on the already existing level and ro-
bustness of evidence. If a hybrid design is chosen, define in which steps of the pro-
cess effectiveness is measured.
Table 1 compares effectiveness and implementation research along four design characteristics of implemen-
tation projects (see also [26]). For investigating effectiveness, the intervention is at the centre of the project,
while for testing implementation, the implementation strategy is the object under study.
Table 1. Effectiveness vs. implementation trials in improvement programmes
Design Characteristics
Effectiveness Trial
Implementation Trial
Test
Intervention (practice, tool, programme,
etc.)
Implementation strategy (facilitation, train-
ing, reminders, etc.)
Typical Unit of Randomization
Institution, unit, individual, or patient
Provider, unit, or system
Typical Unit of Analysis
Institution, unit, individual, or patient
Provider, unit, or system
Summative Outcomes
Health outcomes (e.g., behaviour, sur-
vival), and clinical process outcomes
(e.g., number of medication prescription
errors)
Adoption and appropriate use of interven-
tion (see Implementation outcomes in
section 2.4)
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 19
As mentioned above, the need for keeping track of the effectiveness of a specific intervention usually leads
to combinations of effectiveness and implementation research in implementation programmes. These com-
bined approaches are called hybrid designs. Depending on the extent to which effectiveness or implemen-
tation is focused upon in a programme, different types of designs emerge. Curran et al. [25] systematised
them into the following three categories (they are also discussed in the ImpRes guide [36] that we present
in section 2.4):
Hybrid design type 1 focuses primarily on the effectiveness of the intervention and only as a secondary
aim on the best strategy to implement the intervention. This approach could be used for interventions that
have not enough evidence base yet and need a thorough testing and potentially refinement before scaling
them up for a national programme. Using hybrid design type 1 brings an advantage in that the actual test of
the effectiveness is coupled with some basic assessments of the implementation of the intervention in the
real world. This makes the subsequent preparation phase easier in which the implementation strategies will
be chosen and tested.
For hybrid design type 2, there are two concurrent goals that are pursued simultaneously: Testing the
effectiveness and the best implementation strategy for a certain intervention within one step. This approach
allows to move faster along the implementation phases in combining the testing and the preparation phases
(see section 2.2 and [25]). It is specifically recommendable if the following three conditions are met: The
intervention and the chosen implementation strategies both should have a strong face validity, be supported
by indirect evidence (for example about the intervention being well implemented in other settings) and con-
stitute minimal risk to the setting, for example not replacing a highly effective other clinical procedure. Pro-
jects within the hybrid design type 2 are usually more complex with more trade-offs to consider than type 1
designs.
Hybrid designs of type 3 focus strongly on the implementation strategies of an intervention, but still assess
its effectiveness, too. A pure implementation project would do the implementation part of the programme
after effectiveness was established and solely focus on implementation. However, continuing to assess the
effectiveness measures, such as clinical outcomes, has several advantages. When changes to the interven-
tion during implementation are made, often called local adaptations, their consequences on the intervention’s
effectiveness can be traced. Furthermore, when a certain intervention is encouraging the implementation of
a clinical intervention for which effectiveness has not been established fully, it is appropriate to apply a hybrid
design type 3. To conclude, hybrid designs of type 3 are recommended, when additionally, to the precondi-
tions described for type 2 designs, the following two additional conditions apply: The intervention is strongly
recommended for adoption by important stakeholders in the healthcare system, and there is evidence that
the implementation strategy being tested is feasible in the clinical and organisational context targeted by the
intervention.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 20
Each implementation project thus needs a set of appropriate measures to be able to assess effectiveness
and implementation success. This means, the aims of each need to be defined and the related measures
need to be operationalized. Our literature and resources research has identified a very valuable tool that we
decided to use in future implementation work. It supports also hybrid designs and can be applied in each
process step along the translation of an intervention into system-wide practice [37]: ImpRes [38]. We will
present it in section 2.4.
2.3 Phases in implementing evidence-based interventions
Once a patient safety intervention has been developed and its effectiveness established, its implementation
in the healthcare context needs to be organised and planned for. Four basic phases of implementation are
differentiated in figure 4:
1) exploration phase, i.e., identifying a need and an effective evidence-based practice addressing this
need, explore feasibility of implementation
2) preparation phase, i.e., find out barriers and facilitators of implementation of a certain intervention,
assess need for adaptation and develop a detailed implementation plan; has also been called testing
phase in the Institute of Healthcare Improvement (IHI)-model [39]
3) implementation phase, i.e., initiate full-scale implementation with ongoing monitoring of the imple-
mentation process, and possibility to adjust the implementation strategy
4) sustainment phase, i.e., the structures relating to outer and inner context (meaning within and out-
side of the implementing organisations), processes and supports are applied in order to assure im-
plementation, with adaptation as necessary
These phases are described by the EPIS (Exploration, Preparation, Implementation, Sustainment) frame-
work that was developed to assist and structure large-scale interventions in healthcare. Other frameworks
describing implementation phases apply different names of the phases and accentuate slightly different chal-
lenges that are faced in implementation projects. For example, the scale-up framework used by the IHI de-
scribes similarly the implementation process without clearly differentiating does it work? and make it work
phases but highlighting the need for developing a theory of change (for more background on the importance
of establishing a theory of change, or a logic model of the intervention, please consider section 2.7).
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 21
2.4 Recommended for supporting implementation: ImpRes tool
In order to identify useful and tested tools to support large-scale implementation efforts, we conducted a
search with renown implementation institutes and research groups. From that search, a highly useful tool
was identified, which is designed to support efforts to sustainably implement evidence-based interventions:
ImpRes (Implementation Science Research Development Tool) [37,38]. There is no specific level of exper-
tise needed in order to be able to use the tool, which is a great advantage; it leads the user through the
implementation process in forcing him or her to systematically work on important issues and aspects of
implementation projects, such as defining and operationalizing implementation success indicators. While
recommendations, guidelines and frameworks exist for specific aspects for every step of an implementation
project, there is no other tool bringing them together in one place and formatted and prepared in an easy-to-
use manner. Thus, the ImpRes tool
- provides step-by-step guidance in implementation projects, reducing the burden of synthesizing prior
evidence or searching existing specific guidelines;
- acts as a comprehensive resource for implementation research efforts, supporting the systematic as-
sessment of important indicators and the definition of important questions and their answers;
- supports implementation researchers and practitioners to not forget to integrate or evaluate important
aspects in their projects;
- is specifically directed to implementation in healthcare;
- is intended to be used if some evidence of effectiveness, or at least efficacy has priorly been estab-
lished. Thus, it accentuates the make it work phase (see section 2.1) of a programme, while still also
covering aspects relevant to effectiveness research (hybrid designs).
Recommendation: To guide and support an implementation project, use the Im-
pRes tool. It supports many important decisions and concentrates guidance and
tools to take informed decisions along the implementation process. Precondition: An
effective intervention has to be identified beforehand.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 22
The ImpRes tool is designed in the form of a document that can be filled in. As a side note, please be aware
that although the tool uses the word research pretty often, it denominates activities necessary to conduct
evidence-based and sound implementation projects. ImpRes covers core principles and methods relevant
to implementation projects in healthcare, and splits them up in ten domains:
- Project basics and phases. First, the implementation project needs to be described and the imple-
mentation questions specified, as well as the implementation phase needs to be identified (based on
the phase model presented in section 2.2).
Hybrid design appropriate? The users need to consider whether they need to do a pure imple-
mentation project, or whether and to which extent a hybrid design is appropriate. Three types
of hybrid designs are differentiated, depending on whether effectiveness research or implemen-
tation research are the primary aims, or whether they are equally important.
- Rationale for choosing framework. Second, an implementation framework/theory/model needs to
be chosen and the choice needs to be motivated, or, if none is chosen, the rationale behind this deci-
sion needs to be explained, and the customised approach described. Important: No single framework
is proposed in the ImpRes; therefore, we chose one, see following section 2.5.
- Contextual factors. Third, contextual factors that my foster or hinder implementation need to be listed
and potential local adaptations of the intervention to be implemented and their rationale need to be
laid out. This step is extremely helpful, because it urges the user to early think about the need for
potential adaptations that may come up during implementation, and thus to develop solutions and
limitations for that.
- Implementation strategies. Fourth, the implementation strategies that are intended to be used need
to be listed and the rationale for their use explained. There is also a list of implementation strategies
provided in the accompanying guide [36]. For example, the programme may rely on local support
teams providing guidance, training, and financial incentives to implement their targeted intervention.
- Service and patient outcomes. Fifth, the outcomes and their operationalisation, i.e., the way they
are measured, need to be defined. Therefore, the measurement method (e.g., observations or survey
data), the unit of measurement (e.g., individual healthcare worker, or patient, or healthcare institution),
as well as the planned points in time to assess the intended measures. This step forces the imple-
mentation project team to deeply consider the effectiveness part of the implementation project and to
deliberately take decisions on what effects to measure and how to be able to document actual change.
A logic model (see section 2.7) is very helpful in this step, as it forces the implementation team to
reflect upon and concisely define the expected effects of the intervention.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 23
- Implementation outcomes. In the sixth part of the tool, the implementation outcomes are to be de-
fined, as well as their operationalisations, congruent to the service and patient outcomes. The final
aim of any improvement programme is to sustainably integrate the intervention into the actual work of
the healthcare providers. So, every patient or service that may benefit from the intervention should
receive it after implementation in the way it was planned (fidelity). Additionally, the intervention should
be implemented in all institutions it applies to (penetration). Furthermore, for the implementation to be
successful, the intervention should become standard and be sustained on a long-term perspective
(sustainability). ImpRes proposes eight implementation outcomes (relying on Proctor et al.’s estab-
lished taxonomy [40]):
Acceptability: satisfaction with various aspects of the innovation
Adoption: intention, initial decision, or action to try or employ an innovation
Appropriateness: perceived fit, relevance, or compatibility of the innovation
Feasibility: extent to which an innovation can be successfully used
Fidelity: degree to which an intervention was implemented as planned
Cost: cost impact of an implementation effort
Penetration: integration of a practice within a service setting and its subsystems
Sustainability: extent to which a newly implemented treatment is maintained
Furthermore, the measurement methods need to be defined, at which level the measurement is gathered
(i.e., on patient, provider or unit level), and at what time points the measurement shall be conducted. Fur-
thermore, the user needs to decide whether any relationship between the measured implementation out-
comes and effectiveness outcomes will be investigated.
- Unintended consequences. Seventh, potential unintended consequences need to be anticipated,
and specified how they will be explored. These consequences may be both, positive, such as a de-
crease of time needed to fulfil a certain task, and negative, such as a higher stress level in clinicians.
This step is helpful to diligently keep track of the effects of the intervention on the healthcare institution
as a system, to reduce potential negative side effects and maximise potential unintended benefits of
the implementation of an intervention.
- Economic evaluation. In the eighth step, the financial aspects of the intervention and the costs of
the implementation are considered. Therefore, different kinds of costs are listed (e.g., net intervention
cost, net implementation cost, health benefits, etc.) and the user needs to decide whether and how to
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 24
measure them. This step is extremely helpful in the early planning stages of an implementation project
to make sound predictions of costs and benefits expected. This way, the ImpRes tool would also help
implementation teams to meet the requirement presented in the NR [2] demanding economic evalua-
tion of costs and benefits for regional and national programmes.
- Stakeholder involvement and engagement. Ninth, all stakeholders of the implementation project or
programme need to be identified and listed. These may be policy makers or frontline clinicians. For
each of the stakeholder groups, the users need to decide in which of the stages of the implementation
project they will be involved (i.e., take actively part, consult or collaborate) or engaged (informed at a
conference for example).
- Patient and public involvement and engagement. As a tenth step, the extent to which patients and
the public play a role in the programme is decided upon, again along the different steps of the pro-
gramme, from identifying topics, over managing the programme, dissemination, to finally evaluation
its impact. Again, involvement and engagement are differentiated.
Recommendation: As the challenges of an implementation project vary with the
phase it finds itself in, the ImpRes tool should be used more than once, for example
in each of the implementation phases (see section 2.2).
The guide has been developed following a review of implementation science literature and including inter-
national experts on implementation science, and has been evaluated as improving the quality of implemen-
tation research, and containing the key components to consider in implementation research by those apply-
ing the tool [37].
As developing implementation outcomes is an important and challenging step in the implementation process,
we want to highlight two additionally helpful resources to support this step: to advance the use of established,
sound assessment instruments, the society for implementation research collaboration (SIRC) established a
repository with assessment instruments, rated by their assessment quality. As a useful resource, we want to
mention here the D&I toolkits published by the Washington University in St. Louis. Their webpage provides
brief overviews about all aspects of an implementation programme, as well as about implementation out-
comes. The site can be found under: https://sites.wustl.edu/wudandi/di-toolkits/.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 25
2.5 Choosing an implementation framework: recommendation and selection criteria
The ImpRes tool does not propose a specific implementation framework to work with. In order to support
future patient safety improvement programmes in Switzerland, we reviewed the literature to identify a frame-
work most suitable and practicable. The resulting recommendation of this review is presented in the follow-
ing.
A large variety of frameworks, models, theories exist that are applied to guide and structure implementation
efforts. They were developed for various projects or initiatives and factor diverse implementation phases,
influences, and relationships of influences on implementation success. In order to structure the large array
of implementation models, frameworks, and theories, Nilsen [41] proposed to differentiate five kinds of ap-
proaches, with the 1) process models laying out a temporal sequence of implementation projects in order to
guide the user in the implementation process. The 2) determinant frameworks, 3) classical theories and 4)
implementation theories contain influencing and explaining factors to better understand and predict the out-
comes of implementation projects. In the final category, 5) evaluation frameworks are described which are
used to evaluate the implementation success (see also Figure 5).
Figure 5. Three aims of the use of theoretical approaches in implementation science and the five categories of theories,
models and frameworks, from Nilsen P. Making sense of implementation theories, models and frameworks. Implement
Sci 2015;10:53. doi:10.1186/s13012-015-0242-0, published under the terms of the Creative Commons Attribution Li-
cense (CC BY 4.0; https://creativecommons.org/licenses/by/4.0/).
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 26
In structuring the theoretical approaches (i.e., implementation theories, models, and frameworks; will be
called frameworks hereafter) into five categories and attributing them to three aims of implementation pro-
jects, Nilsen made an important contribution for practitioners and researchers tasked with implementing new
practices in healthcare.
As choosing a theoretical underpinning for a new implementation project can be a challenging task consid-
ering the manifold existing approaches, we decided to recommend a specific framework for our future work
implementing patient safety interventions in Swiss healthcare, as the aim of this project was to provide guid-
ance. It is shortly presented in the following before criteria to choose implementation frameworks are de-
scribed motivating our recommendation in more depth.
As a suitable and flexible framework, we recommend EPIS (Exploration, Preparation, Implementation,
Sustainment) which was developed to guide the implementation of evidence-based practices in the public
sector. Going back to Nilsen’s [41] differentiation, with EPIS we chose a process model that at the same time
works as a determinant framework and includes many relevant influencing factors for implementation out-
comes.
The EPIS framework covers a large array of important factors for considering in implementing evidence-
based practices:
- Factors of the outer context (environment of the service and interorganisational context, client/pa-
tient characteristics, etc.)
- Factors of the inner context (organisational culture/climate, leadership, quality and fidelity monitor-
ing, staffing, workers’ attitudes towards the intervention, etc.) [42]
- Innovation factors covering a) the developers of the intervention, e.g., whether they are engaged in
the implementation process and whether they take part in testing the intervention, adaptation of the
intervention, b) intervention characteristics, e.g., how complex it is, how much needs to be learnt to
apply it etc., and c) the fit of the intervention to the needs of the population and/or the context it is
implemented in
- Bridging factors recognizing the interconnectedness between outer and inner context of the organi-
sation (such as community-academic partnerships, or lobbyists impacting pharmacy legislation, or
governmental policies influencing hospital certification or reporting requirements)
EPIS helps govern the understanding of the implementation process, its potentially relevant influences (de-
terminants), and its evaluation. It is therefore applicable to a large variety of purposes and covers many
relevant aspects to account for in implementation and scale-up projects. It explicitly highlights the engage-
ment of stakeholders from multiple system levels [43], making it particularly valuable for large-scale imple-
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 27
mentation efforts. Along the different stages of an implementation project, different factors may be of rele-
vance. For example, funding as a factor of the outer context may be a relevant factor across all phases,
while fidelity monitoring (inner context factor) may only be important during implementation and sustainment
phase [44] of a programme. In order to better identify the relevant factors, it is recommended to develop a
theory of change or logic model of the intervention and its implementation strategy [44] (see section 2.7).
Additionally, a large variety of quantitative tools to assess EPIS-related factors has been developed [44]. As
a resource to be informed about latest developments, e.g., about operationalisations of factors of the frame-
work, worksheets, guides and other tools, there exists a website: https://episframework.com/.
Recommendation: Use a framework in future programmes/implementation projects
to structure the implementation process, factor important concepts, and to gain an
understanding of the processes and influences crucial to implementation efforts. We
recommend using EPIS in applying the ImpRes tool.
As ImpRes integrates Proctor’s framework of evaluation dimensions [40], we do not recommend a specific
framework for evaluation purposes here.
Selecting a framework for implementation projects can be a daunting task due to the large variety of existing
approaches. Therefore, several criteria have been proposed to support the selection process. In the follow-
ing, we shortly discuss these criteria and their influence on our choice.
In an influential review, Tabak et al. [31] developed categories to sort frameworks and proposed the following
criteria for selecting one for a specific project:
- The model or framework’s socio-ecological level (e.g., individual, organisation, community, system,
see also [43]) should match the level targeted with the implementation project. With EPIS spanning
across many levels of the healthcare system, it was considered particularly appropriate for national
projects focusing on implementation in healthcare institutions.
- The contents of the constructs that are factored in a framework should fit to the relevant issues in
the planned implementation project. Construct definitions may be flexible to varying extents: Broadly
flexible constructs are rather loosely outlined and allow an application to a large array of projects,
while more operationally defined constructs have a more narrowly defined area of application in an
implementation project. With its multitude of relevant factors, EPIS covers a broad range of poten-
tially relevant constructs for national programmes.
- Whether the framework or model is rather dissemination and/or implementation-oriented, should
also be considered. The EPIS framework supports both orientations.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 28
Another relevant criterion for framework choice is the prior use of a model or framework, i.e., the context it
was originally developed for and has been used for so far [41]. EPIS was evaluated in a recent review as
being applied across a multitude of settings and projects [44] and has priorly been used in healthcare, e.g.,
in child welfare and mental health sectors; targeting health problems such as maltreatment, and behavioural
problems [44]. Additionally, EPIS explicitly takes stakeholder engagement on different system levels into
account as one of the key elements of implementation projects [43], which makes it valuable for scaling up
on a national level. Apart from that, an international survey reported that EPIS was among the quite fre-
quently used frameworks [45].
Since each project comes with new and individual challenges, it may appear to make sense to develop for
each implementation effort a new model or framework. Even more radically, it has also been argued that the
use of implementation frameworks is not necessarily better than applying common sense as long as the
effectiveness of the frameworks has not been empirically tested and compared [46]. However, the ad-
vantages of using established models or frameworks are the possibility to base own projects on prior
knowledge, to support the implementation process with established frameworks and to be able to use previ-
ously developed tools for assessment and analysis of implementation success. As implementation endeav-
ours usually result in large projects spanning over a long timeframe, it may be necessary or appropriate to
use more than one framework [37]. However, the use of ImpRes and our decision to recommend EPIS
should be able to cover a large variety of challenges, reducing the need to look for other frameworks. Fur-
thermore, existing frameworks may be tailored and adapted to specific use cases in order to match the
challenges faced in a particular project [31].
Recommendation: Rather than self-developing a new framework for a programme,
there are many advantages to using established frameworks. Previously developed
assessment tools can be applied, and the results compared across different pro-
grammes, e.g., in using standardized measures.
A plethora of tools and online resources has been developed to support researchers and practitioners in the
process of selecting the right implementation model and in the whole implementation project workflow (e.g.,
[47]). One resource was considered particularly useful and therefore is mentioned here: https://dissemina-
tion-implementation.org/. A site that is supposed to help in selecting an implementation model, but also to
designed to compare models and frameworks. Out of that, we do not list further resources, as the goal of
this report is to make specific recommendations to spare the work of searching through the literature for
future programmes taking place in the Swiss healthcare setting.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 29
2.6 Developing implementation strategies
In the fourth step of the ImpRes tool, the implementation strategies need to be described and their application
motivated. Implementation strategies are required to bring evidence-based practices into healthcare prac-
tice, with the aim that the new, safer, or better way of doing things is taken up by all relevant personnel and
institutions, and that the intervention’s effects reach all affected patients. Generally, implementation strate-
gies are understood to be the methods used to enhance the adoption and sustainable implementation of a
practice [48].
Implementation strategies are usually designed to address certain implementation barriers. However, there
is no consensus or consistent relationship established between certain strategies and certain barriers so far
[49], highlighting the role of context and need for further research. Stakeholders, such as clinicians can
contribute to identify implementation strategies that are feasible and tailored to overcome anticipated imple-
mentation barriers [50,51]. Implementation strategies are often interconnected in taking evidence to scale
[52]. For example, several strategies may be interlinked to enhance clinician knowledge, for example, de-
velop educational materials, conduct educational outreach visits, or create a learning collaborative [52].
In an effort to compile and define implementation strategies, the Expert Recommendations for Implementing
Change (ERIC) study gathered input from implementation experts and iteratively developed a list of 73 dif-
ferent implementation strategies [4,53,54]. In the second chapter of this report, Waltz et al. categorisation of
these implementation strategies [4] was used to classify studies on patient safety interventions. Table 2 lists
these implementation strategy categories along with some examples from Perry et al. [52] to illustrate what
the implementation strategies may look like.
An implementation strategy defines which actor, does what kind of action, to which target, in which dose, at
what kind of temporality, using which justification and expecting what kind of outcome [48,52]. So for exam-
ple, if the intended implementation strategy is clinical supervision, the actor is the supervisor who is an expert
in the implemented practice; the action is the provision of the clinical supervision in answering questions,
reviewing implementation, making suggestions and encouraging; the target of the actions are the clinicians
who are newly trained in the intervention; the temporality may be specified as the supervision needing to
start one week after training; the dose could for example be to supervise once per week for 15 minutes
during 12 weeks; the implementation outcomes affected are expected to be the uptake of the intervention,
the penetration among eligible patients and the fidelity to the interventions’ practices; the justification for this
strategy could be that research suggests that post-training clinical supervision is very important (example
from [48]). It is recommended to specify the implementation strategy of a planned intervention along these
dimensions.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 30
Table 2. Implementation strategy examples (based on Perry CK, Damschroder LJ, Hemler JR, Woodson TT, Ono SS,
Cohen DJ. Specifying and comparing implementation strategies across seven large implementation interventions: a
practical application of theory. Implement Sci. 2019;14(1):32. doi:10.1186/s13012-019-0876-4)
Implementation strategy
category
Strategy example 1
Strategy example 2
Strategy example 3
Use evaluative and iterative
strategies
Obtain and use patients and
family feedback, e.g., patient
focus group, patient and family
advisory councils
Audit and provide feedback
Develop and organise quality
monitoring systems
Provide interactive assis-
tance
Provide clinical supervision
Provide local technical assis-
tance, e.g., audit charts to val-
idate data reports
Implementation facilitation,
e.g., review workflows and
create actionable plans to put
best practice in place
Adapt and tailor to context
Tailor strategies
Use data experts
Promote adaptability
Develop stakeholder interre-
lationships
Identify and prepare champi-
ons, e.g., assist in creating
and engaging practice leaders
and others in promoting effi-
cient care strategies
Use an implementation advi-
sor, e.g., support and educa-
tion for practice facilitators by
data experts and/or consultant
or physician faculty hired as
expert consultant
Identify early adopters
Train and educate stake-
holders
Conduct ongoing training
Provide ongoing consultation,
e.g., provide expert advice by
e-mail or in person
Distribute educational materi-
als
Support clinicians
Revise professional roles
Facilitate relay of clinical data
to providers
Remind clinicians, e.g., assist
the change process through
use of clinical reminders and
decision-support tools
Engage consumers
Involve patients/consumers to
be active participants, e.g., in-
volve patients in QI teams
intervene with patients/con-
sumers to enhance uptake
and adherence
Use mass media
Utilize financial strategies
Alter incentive/ allowance
structures
Make billing easier
Start a dissemination organi-
sation
Change infrastructure
Mandate change
Change accreditation or mem-
bership requirements
Change records systems, e.g.,
help practices upgrade to new
EHR and/or modify existing
EHRs for efficiency and accu-
racy
Additionally, the strategy needs to be defined in specifying potentially discrete components. These specifi-
cations allow for a systematic application of strategies and for developing evidence-based or evidence-in-
formed implementation strategies in the preparation phase of a national programme. Thus, the best way of
increasing and sustaining a high adoption of a practice can be empirically identified [25]. While this may be
costly and methodically challenging, it might be valuable to test and justify certain implementation strategies
before a large, national scale-up. As the evidence of the effectiveness of implementation strategies in Swit-
zerland is very scarce, it would make sense to develop a kind of database of implementation strategies. The
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 31
strategies could be systematically described in their above mentioned definitory characteristics and their
effects reported.
Recommendation: Systematically reporting and gathering implementation strate-
gies could be a promising avenue for systematising future national improvement pro-
grammes in Swiss healthcare.
2.7 Logic model of the intervention and its implementation
Foundational for all efforts to implement new practices into healthcare is an assumption about how it will
affect care: This underlying theory of change [55] is also often called logic model of the intervention. It is
particularly important to make the logic model visible and elaborate it at two points in the implementation
project: a) The logic model about how the intervention should reach the intended aims, and b) the logic model
about how the selected implementation strategies will be effective in bringing about a high adoption rate.
Logic models describe the anticipated relationships between activities, measurements, implementation ef-
forts and patient outcomes. Explicitly formulating key assumptions of the project is a very important and
valuable step. For example, being aware about the expected effects and modes of action supports the im-
plementation project team in deliberating whether and which local adaptations to allow for and to keep track
of their effects on the original intentions of the intervention and their implementation. Thus, developing and
regularly updating the logic model of the intervention and its implementation strategies, helps reach a better
understanding of the key mechanisms of change in the implementation effort. The before mentioned website
may support the development of a logic model: https://dissemination-implementation.org/content/plan.aspx
Recommendation: Generate a logic model for the intervention and its implementa-
tion, featuring all relevant factors, effects, and their interrelations write down the key
assumptions behind the intervention and its implementation.
Formulating the assumptions of cause and effect behind the project also will support in dealing with the
emerging need for adapting an intervention to local conditions. With the logic model, it is easier to realise
when a change to an intervention alters its potential to cause the intended effect, i.e., if it interferes with
assumptions of the logic model. Changing certain aspects of an intervention may fundamentally change its
effectiveness, in altering its inner workings. For example, a study conducted at the Patient Safety Foundation
investigated the safe surgery checklists of 18 hospitals and revealed vast differences between them [56].
Additionally, essential items of the original WHO recommendations often were omitted from the lists in use.
Changes that may appear small in everyday work may be quite consequential for the effectiveness of an
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 32
intervention. Therefore, keeping a close eye on how changes affect an intervention is important. Diagnosing
interferences will help to find ways to deal with the need for local adaptation that may not alter the basic
functioning of an intervention or, to change the intervention being aware of how its basic mechanism is
affected. The same applies to implementation strategies.
Particularly for large-scale implementation programmes, covering multiple sites, settings, or regions, local
adaptations are an often-discussed issue. The more complex an intervention is, the quality of the implemen-
tation strategies, and the context of the setting of implementation influence whether adaptation is perceived
as necessary or not [57]. The need to implement exactly the same intervention needs to be balanced with
the desire to facilitate local adaptations, small changes that are expected to make uptake easier. Using the
logic model and additionally the categories to characterize modifications and adaptations of interventions
developed by Stirman et al. [58], will help gaining a deeper understanding of their potential effects: for ex-
ample in specifying the content, the context of the modification, and identifying who is making it, defining its
nature, i.e., whether elements are added, or removed etc. Instead of understanding adaptations as a dis-
turbance to the implementation process, they can also be conceptualized as a process to foster implemen-
tation fidelity in meeting local needs. The challenge is to change the intervention in its effectivity as little as
possible in order to assure that the desired effects proven before the roll out can be kept up. Within the EPIS
framework, a process for planning for adaptations rather than dealing with them ad hoc has been tested, it
is called the Dynamic Adaptation Process and involves multiple stakeholders, is data-informed and collabo-
rative [59]. In taking outer and inner contexts into account, the EPIS framework is specifically helpful to
understand changes to implementation strategies. Goodrich and colleagues [32] also suggest a systematic
approach to adapt interventions in order to find a balance between achieving a high fidelity rate and allowing
adaptations for frontline stakeholders at the same time.
2.8 The role of evidence in implementing evidence-based practices
Evidence is called for on three levels in large improvement projects: evidence of the intervention itself, of its
implementation, and of its sustainable use in the work settingsin the form of an evaluation. The basic idea
of informing and guiding practice in healthcare based on evidence goes back to the origins of evidence-
based medicine [60], which stipulates the combination of individual clinical experts with external clinical evi-
dence in order to ground clinical decisions and actions on sound empirical evidence about the best ways to
diagnose and treat patients and to evaluate own performance. This basic idea has been translated to
healthcare improvement in general: More and more new interventions are demanded to rely on evidence-
based practices, counteracting the tendency to optimistically believe an improvement works [23], because
of its persuading face validity. Rather, research shows that implementations of programmes and interven-
tions that have been successful in one setting may fail in more rigorous testing [23]. Thus, post-implemen-
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 33
tation efforts to understand what, why, and how it worked or not are highly necessary [55,61]. Also, improve-
ment efforts cost resources and the effective use of these scarce resources needs to be justified. Often, it
becomes only clear that an intervention did not work, but the reasons remain unclear. Particularly, in evalu-
ating large projects, realizing and admitting that the implemented intervention does not have the planned
effects, is rather uncomfortable for all stakeholders of the project [61].
While the motivation to call for evidence in healthcare improvement is well justified as considerable resources
are being invested in those projects, there are also pitfalls to be considered in the approach of claiming
evidence to justify every part of an improvement programme. It is very costly and time-intensive to develop
the assessments, to analyse them and to draw conclusions from them that inform the implementation. Due
to the complexity of the work being done in the healthcare system, changing a seemingly simple work prac-
tice might have unintended consequences in the institutions and in the healthcare system more broadly (e.g.,
people being interrupted to carry out a certain task, shifting priorities away from other important tasks, or
consequences in financial aspects, etc.). Grasping this complexity with effectiveness and implementation
measures is a challenging endeavour, specifically because the systems are constantly changing and adapt-
ing to new demands which may also affect the intervention to be implemented. Additionally, many interven-
tions and established tools of healthcare improvement are hard to measure in their outcomes on patient
level. This is often true for interventions addressing meta-competencies to learn from errors, or simulation
trainings to improve communication skills. Providing clear and unequivocal evidence on the effects of such
programmes is sheer impossible. Their effect may show in the kind and frequency of adverse events tracked
over time, and, also, in the safety culture. However, other measures defining intervention effects and imple-
mentation outcomes are possible and should be assessed. For example, the individual self-reports of the
participants learnings in a simulation training.
Learning from positive deviance is a more recently emerging form of developing evidence-based
practices. Instead of developing an intervention and then generating evidence corroborating it, the process
starts with the identification of a successful practice through its evidence. Positively deviant institutions or
units can be identified through registry data or other outcome measures, for example low infection rates.
After investigating the factors facilitating the success of the unit or institution, they need to be tested in rep-
resentative samples [62,63] before a larger implementation is rolled out. For example, in the UK a large
project studied the factors driving the success of a high performing maternity care unit in order to identify
factors that could be used for dissemination to other maternity units [64]. In focusing on solutions for patient
safety that are already existent and have been developed within the community, interventions based on
positive deviance identifications tend to be better accepted by practitioners [63].
High-performing healthcare institutions or units can be identified in using routine data. Data depth, quality
and availability plays an important role. Developing a learning from positive deviance programme in
which other units or institutions learn iteratively from high performers represents a promising future
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 34
opportunity for Swiss healthcare. In continuously assessing the targeted outcome and iteratively emulat-
ing the approaches of the high-performing organisational units, the participating units are expected to im-
prove. It is crucial however, that the high performers are willing to share their knowledge and expertise and
usually have to be compensated for devoting their time and resources to support others. This kind of ap-
proach relies on the co-production of knowledge between practitioners, which is facilitated by external sup-
port from a programme. Thus, providing useful tools, identifying and communicating successful strategies in
collaboration with the high performers is expected to encourage the less well performing units to implement
practices to achieve better results.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 35
3 Review and assessment of the progress! programmes
Since 2012, the Swiss Patient Safety Foundation has been mandated by the Federal Office of Public Health
to develop and conduct pilot programmes with the aim of improving patient safety in healthcare. These pilot
programmes are carried out under the label progress!. The overall goal of the progress! programmes is to
support healthcare organisations in enhancing patient safety and reducing adverse events by implementing
evidence-based interventions in areas with significant need for action and an established potential for
improvement.
To date, three programmes are completed and two are ongoing. The goals of each programme are listed
below. More details on the content of the programmes can be found under the follwowing weblink:
https://www.patientensicherheit.ch/programme-progress/
- No.1: progress! Safe surgery (20122015): Implement surgical safety checklists in hospitals to reduce
surgical errors.
- No.2: progress! Medication reconciliation (20142017): Introduce medication reconciliation in
hospitals as a strategy to reduce the risks for medication discrepancies at hospital admission, transfer
and discharge.
- No.3: progress! Safe urinary catheterisation (20152018): Implement an evidence-based intervention
bundle in hospitals to reduce the use of indwelling urinary catheters and promote safe catheter
insertion and maintenance.
- No.4: progress! Medication safety in nursing homes (20162021): Implement standardised processes
for reviewing and monitoring medication in nursing homes to reduce polypharmacy and the use of
potentially inadequate medication.
- No.5: progress! COM-CheckSafe surgery (20192021): Build internal hospital competencies to
conduct direct observation and feedback to improve compliance with surgical checklists.
3.1 Basis and structure of the progress! programmes
According to the project proposal to the Federal Office of Public Health from 16th May 2012 the pilot
programmes shall respond to a veritable need, be realistic and adapted to Swiss conditions, and be models
for similar, more extensive programmes.
In particular, they should rely on a sound evidence base, whereby this evidence may originate from within
or outside Switzerland. The focus is not on a renewed effectiveness review based on a before-and-after
measurement of clinical outcomes, but on the implementation of already proven interventions. Furthermore,
the progress! programmes are explicitly defined as pilot programmes, i.e., the implementation in several
hospitals is to be tested within the scope of the progress! programmes. They are thus intended to serve as
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 36
models for later scale-up at the national level. However, the national roll-out is not part of the pilot projects.
According to the structure of the EPIS model (see section 2.3), the progess! programmes can be thus
assigned to the preparation phase of a larger national implementation programme.
The first three progress! programmes were based on two axes: Axis one consisted of a campaign that aimed
to raise awareness and increase visibility through widespread dissemination of information, tools and
materials throughout Switzerland. Axis two consisted of an in-depth improvement project with the aim to
accompany and support a small group of motivated pilot organisations throughout implementation of a
standardised intervention. Participation was voluntary and adaptations to local needs were possible. The
activities on both axes were intended to reinforce each other. Whereas the first three programmes followed
a very similar structure, some important changes were made in the two subsequent programmes: The fourth
programme (Medication safety in nursing homes) is the first programme that does not take place in acute
care hospitals, but in nursing homes. This long-term care setting is entirely different from the hospital setting
and is characterised by great variety of types of organisations and models of collaboration between nursing
home staff, physicians and pharmacists, in part rendered by regional law. Furthermore, there are also fewer
evidence-based interventions available for nursing home contexts. For this reasons, the two-axis structure
was abolished and the programme was split into two sequential projects: (1) analysing the situation and
identifying primary fields of action and possible improvement measures (completed); (2) improvement project
with nine pilot nursing homes, accompanied by a pre/post outcome study. The fifth programme is a follow-
up of the first programme and is still ongoing. The main aim is to build competencies and skills among
interprofessional clinical teams in 17 pilot hospitals, so that they themselves can monitor compliance with
the checklist in their hospital and provide constructive feedback to their peers to improve checklist
application.
3.2 Reflection
In order to learn from the previous progress! programmes for future improvement projects, the results and
findings from the three completed pilot programmes progress! have been reviewed and classified;
experiences from the two current programmes have been integrated if appropriate. Furthermore, findings
from the literature, the NR [2] as well as personal experiences of the internal project managers of the
progress! programmes served as information source for this section. In order to find out the current status of
implementation in the pilot hospitals of the first three progress! programmes, the pilot hospitals were
contacted again by email/telephone.
Basis of the progress! programmes
To date, the principal goal of the progress! programmes was to improve the quality and safety of healthcare
by promoting and supporting the implementation of evidence-based interventions. However, the evidence
base for Switzerland is rather poor. There are some interventions that have already proven effective in other
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 37
countries and settings. However, for the Swiss context, which has some specific characteristics (such as
three national languages), this evidence is rare. For example, there is no single intervention that has been
proven effective to improve polypharmacy and potentially inadequate medication in nursing homes and that
could easily be transferred to a Swiss context withfor examplea limited number of clinical pharmacists
available. Thus, for the fourth programme, there was a need to develop and test new interventions rather
than implement existing ones. The fifth programme was developed because even after successful
implementation, there was still need to improve compliance with surgical safety checklists. Therefore, the
fifth programme uses a different project design focusing on building up internal competencies and structures
to observe and monitor compliance with surgical safety checklists. This is in line with the recommendations
of the NR [2], which states that education and training of health professionals in quality management, human
factors, and implementation science is essential, especially for those involved in national programmes.
Structure
The two axes (campaign and improvement project) implemented in the first three progress! programmes
were intended to interact and reinforce each other (rising tide). This model was adopted from other national
strategies like the National Programme for HIV and Other Sexually Transmitted Infections (NPHS). The
effects of the two axes were not evaluated independently. It is therefore not possible to distinguish the effects
of each of the two axes on the overall programme outcomes.
Evaluation of the progress! programmes
The evaluation of the progress! programmes focused on process measures in order to examine
implementation fidelity, facilitators and challenges on the one hand and changes in safety culture and staff
attitudes over time on the other. Over time and with the development of new progress! programmes,
stakeholders increasingly asked for data on clinical outcomes in order to assess the effects on adverse
events and to better promote uptake of the intervention on a national level. With regard to the evaluation of
the progress! programmes, the authors of the NR write that no evaluation of clinical outcomes was possible
and that it was difficult to prove the clinical effect [2]. At this point, however, it must be clearly stated that the
evaluation and verification of the clinical effect has never been the aim of the progress! programmes and
they were not equipped with resources to address effectiveness. For the third programme, data on clinical
outcomes was collected (catheter utilization, catheter-associated urinary tract infection and non-infections
complications) in a before/after surveillance. Due to the study design, the programme was classified as
research project and therefore subject to approval by the cantonal ethics committee. It may have been a
misconception that a programme can be implemented without local evidence of effectiveness. At the same
time, however, it became clear that it is very demanding to provide evidence of effectiveness across several
hospitals.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 38
Recruitment
The improvement projects were conducted with a selected group of pilot organisations. In the first three
programmes, the number of pilot organisations ranged from seven to ten. In several programmes,
recruitment of a sufficient number of pilot organisations was difficult and took longer than originally planned.
Reasons for this were, among others, the expected additional expenses from participating in the project,
restrictions as to which hospitals could participate (e.g., could not have participated in a progress!
programme before; had to be from a certain region) and time consuming administrative processes to agree
and sign project agreements.
Duration
The overall time frame for the previous progress! programmes (two to three years for the programme in its
entirety) has been too tight. Delays were caused, for example, by difficulties in recruiting pilot hospitals, by
the requirement that all events and documents (including all revisions) had to be translated into two or three
national languages and the fact that the participating pilot organisations needed sufficient preparation time,
e.g., to obtain the consent of the hospital management and to make resources available. In addition, the
improvement process on axis two had to be restricted to a relatively short period of time. In the first three
progress! programmes, the intervention was implemented in the pilot hospitals over the duration of 6 to 12
months. The time span was too short to show changes, especially when it comes to aspects targeting the
patient safety culture.
National scope
Progress! programmes were required to be national in scope and consider the different language regions.
Great efforts were made in all previous and current programmes to ensure that they could be conducted in
several languages, by providing all materials, workshops, evaluation instruments, consultations, etc. in two
or three languages. This has led to a considerable investment in time and resources and has oftentimes
made project management complex, inert and demanding. In particular, flexibility in adaption was limited.
Local adaptation
Pilot organisations were intentionally selected to represent a wide range of organisational types (e.g., small,
regional and university hospitals) and geographic regions (e.g., urban, suburban, rural). The heterogeneity
of the pilot organisations reflected the Swiss reality well and allowed to gather implementation experiences
from different types of organisations. But it also made it impossible to implement the same measures in every
organisation. Rather, every organisation was free to adapt the improvement measures to their local needs
and capacities. The possibility of adapting measures to local conditions may have encouraged some
hospitals to participate in a pilot programme. However, as a result, implementation differed greatly among
the pilot organisations and evaluation results did not allow to draw generalisable conclusions for hospitals
throughout Switzerland. Moreover, not all adaptations to local conditions are beneficial; some of them may
even compromise the quality of the original intervention [56].
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 39
National roll-out
Each progress! programme included activities to make materials, tools and lessons learned available to all
hospitals in Switzerland after programme completion. In the first two programmes, a declaration was signed
by stakeholders proclaiming the use of surgical safety checklists and medication reconciliation as standards
of care in Swiss hospitals. In the third programme, a follow-up project to establish a catheter device
stewardship was recommended to the Federal Office of Public Health. The feedback we have received from
individual pilot hospitals upon request indicates that the progress! programme materials are used and
appreciated by healthcare organisations throughout Switzerland; however, without further resources it is not
possible to analyse systematically if and to which extent other hospitals implemented the intervention
concepts. In particular, there were no national regulations or incentives to adopt the practices. Assuming a
linear progression from improvement programme planning to pilot testing to implementation at the national
level ignores the dynamic context in which implementation projects are embedded.
The progress! programmes were conceived to be models for larger scale national programmes. However,
to our knowledge, no action has been taken to establish national programmes that are based on the
experiences of the progress! programmes. Due to the fact that the progress! programmes are sometimes
unfortunately referred to as national programmes instead of national pilot programmes (as in the NR [2], for
example), there may have evolved a discrepancy between the expectation and the actual intention of the
programmes. This may also explain why the authors of the NR criticise that reach and scope of the
programmes was limited and that they have been “only haphazardly implemented nationally” [2, p.33],
although roll-out at national level has never been the aim of the pilot programmes and funding of the acitvities
ended with the end of the pilot programm.
Sustainability
In all progress! programmes, the time frame for the implementation projects was limited to one year or less.
It was therefore not possible to determine whether changes observed could be sustained over time. The
authors of the NR criticise that there were hardly any systematically conducted follow-up evaluations [2].
Considering the predetermined time frame of the progress! programmes and the fact that changes in culture
need time to develop and become fully evident, an evaluation of sustainability would have to take place after
programme completion, requiring additional funding.
Insights into sustainability of specific programmes
No.1 progress! Safe surgery: As part of the follow-up programme COM-Check, 16 interviews were conducted
with various health institutions (hospitals, special clinics, outpatient surgical practices, etc.) from all three
language regions, in which the current status of checklist use was surveyed. All the institutions surveyed
stated that the checklist is now widely accepted and has become an integral part of everyday surgical
practice. Frequently mentioned difficulties in the use of the surgical checklist concerned the dependence on
people, the loss in quality of checklist application through routine, and the checklist application at sign-out,
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 40
because often not the entire surgical team is present. Outside the operating theatre, i.e., in the interventional
area, the checklist culture seems to be less pronounced. In outpatient surgical practices, the surgical safety
checklist still seems to be less widespread.
No.2 progress! Medication reconciliation: It is only partially known whether and how the pilot hospitals
continue to carry out the best possible medication history. In January 2020, 5 out of 8 project leaders
responded to a corresponding email enquiry. Two hospitals confirmed the continuation of the best possible
medication history, albeit with certain restrictions in one facility. Another hospital is testing the effectiveness
of the best possible medication history again in a study in order to convince the hospital management to
extend the processes to the entire hospital network if the results are positive. Overall, the large amount of
work involved in medication review, together with the lack of financial support for the additional work, seems
to be a major barrier why medication reconciliation is not widely implemented.
No.3 progress! Safe urinary catheterisation: A survey of the seven participating pilot hospitals in January
2020 showed that most of the activities of the programme were continued. Whether and how further hospitals
were encouraged to implement the intervention bundle cannot be assessed. To our knowledge, the recom-
mendation for catheter device stewardship resulting from the programme has not yet been implemented.
Project management
All three completed programmes were managed in a professional manner, implemented according to plan
and successfully completed within the agreed time frame. However, they also faced some challenges that
influenced project management. Time constraints were tight for all phases of the programme and imposed
administrative challenges. In several programmes, recruitment of pilot organisations proved particularly time-
consuming. Some pilot organisations found it difficult to implement the measures within the given time frame,
especially if there were competing projects such as the implementation of a new clinical information system
that demanded their attention. The language diversity and the possibility for local adaptation increased the
administrative burden of the project management team, since documentation, implementation and data
collection tools, etc. all had to be translated and individually tailored. This is most remarkably demonstrated
by the fifth programme, where data collection sheets were programmed for each participating hospital
individually in accordance with their local surgical checklist, which resulted in a total of 68 different data
collection sheets. Since participation is voluntary, the project lead does not have the possibility to enforce
programme requirements or sanction deviant pilot organisations. Instead, the programme was dependent
on each pilot organisation, since losing one or more would have meant the requirements of the overall
programme could not be met and the required sample size for the evaluation would not have been achieved.
The programme lead is thus in a poor position to demand anything from hospitals.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 41
3.3 Conclusions and recommendations for future pilot programmes
Overall, it seems that the expectations and demands placed on the progress! programmes have moved
further and further away from the initial mandate and intention of the programmes over time. While it was
defined at the beginning that the progress! programmes should rely on evidence-based interventions and
therefore focus on implementation outcomes rather than effectiveness outcomes, it became increasingly
clear during the course of the programmes that there is not enough evidence for patient safety interventions
in the Swiss setting and therefore both implementation and effectiveness need to be considered in such pilot
programmes. Simply assuming that an intervention with proven international evidence would be accepted
as effective in Switzerland proved wrong and limited public perception. In addition, a significant conceptual
gap emerged: Namely, that it was not defined how and by whom the findings from the pilot programmes
were to be implemented at national level. Some stakeholders interpreted these discrepancies as
weaknesses of the progress! programmes, claiming that the programmes did not deliver what they were
supposed to deliver. Of course, one can criticise that this is due to misconceptions in design and
methodology or a lack of insisting on a national roll-out phase. However, these aspects were only brought to
the surface through the implementation of the progess! programmes and in this context, it must be
emphasised that pilot programmes are there precisely to test things. So instead of focusing on what the
progress! programmes have not achieved, the focus should be on what they have achieved and what lessons
have been learnt from the implementation of the progress! programmes with regard to future national (pilot)
programmes. We are convinced that the conceptual weaknesses of the progress! programmes can be
overcome in future programmes through a revised and methodologically sound programme concept and the
appropriate time and financial resources.
From the critical reflection of the progress! programmes, the following recommendations can be drawn which
should be incorporated into the design of future improvement programmes:
- In order to measure clinical outcomes and assess the effects of the intervention on outcomes, a
much stronger study design is required. see Evaluation of the progress! programmes
- Interventions to be implemented sustainably at national level must be simple and clear. Insights
into sustainability of specific programmes
- Participating in a pilot programme, and sustaining the improvement efforts after programme
completion, requires a substantial amount of resources from organisations. In the future, financial
compensation should therefore be considered as an additional incentive for programme participation,
in particular during testing and piloting phases in which no other incentives or regulations are yet in
place. see Recruitment
- For the effects of an intervention to become fully evident, a longer implementation period and more
long-term data are required. see Duration
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 42
- For the preparation phase of a programme, it should be evaluated whether it is necessary to test an
intervention in two or three language regions due to the associated burden and inertia. In some cases,
it may be more appropriate for financial and cultural reasons to pilot an intervention in one language
region only, and then make adaptions before roll out to other regions. see National Scope
- The possibility to adapt measures to local conditions does offer many advantages to participating
organisations including higher acceptance and motivation. Depending on the goals and design of a
programme, local adaptation may need to be restricted in order to ensure the quality of
implememented interventions see Local Adaptation
- More data on implementation outcomes should be collected already in the course of the programme
to monitor the current implementation progress. see Project Management
- Pilot programmes may reveal findings that lead to an adjustment of the original intervention and/or
implementation strategy, up to and including the possibility that the findings of the pilot
implementation reveal that the planned intervention cannot be implemented in the Swiss context or
at the national level. see National roll-out
- If pilot programmes are to be continued, it will be important to define the necessary steps and
responsibilities to ensure that successful pilot programmes are rolled out nationally. see National
roll-out
- Sustainability must be given greater priority in future programmes, which must be reflected in the
design (e.g., follow-up measurements) and the financial resources dedicated to the programme.
see Sustainability
- In order to implement measures sustainably, continuous monitoring is needed to show progress or
regression, to identify areas that still need to improvement and to remind participants of the importance
of the safety measures. see Sustainability
- The position of the project lead should be strengthened, and adequate mechanism should be put into
place to ensure that participating organisations meet programme requirements and have strong
incentives not to withdraw from current programmes. see Project Management
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 43
4 Directions for future improvement programmes
4.1 The current state of improvement efforts in healthcare
More than twenty years ago, the publication of the report To Err is Human of the Institute of Medicine in the
USA [65] helped spark the patient safety movement. Its publication not only provoked concern about the
extent of medical harm, but also enthusiasm to initiate change and improve the safety of healthcare. Subse-
quently, large improvement projects and programmes were conducted in many countries, and initiatives and
institutions were founded to advance patient safety. After the initial eagerness to make healthcare safer, a
certain disillusion grappled the movement: Positive effects on local or regional level were often reported.
However, on national level, only little or no improvements of patient safety were found.
Key message: In the last two decades, there has been a great effort to improve
healthcare quality. However, there is only little evidence that preventable patient
harm has been reduced. Robust and significant effects still need to be achieved to
make healthcare safer.
Two problem areas may be the reason for this sobering diagnosis of missing effects from numerous im-
provement programmes and projects:
1) The effects are hard to measure. Current established measurements may not be sensitive enough to
show the improvements achieved, particularly on a larger, for example a national, scale.
2) The effects have not been large enough to produce associated positive evidence.
Both explanations will be explored in the following. First, data on the success of improvement programmes
is scarce and often of poor quality (see chapter 1 of this report, there are ambivalent findings, lack of meth-
odological rigor and of data on cost-effectiveness). Therefore, the effectiveness of large-scale implementa-
tions of interventions has hitherto not been sufficiently and systematically assessed. Still, large programmes
and projects are costly and need a thorough evaluation [66], as elaborated in chapter 3. To assess the extent
of preventable patient harm that is occurring in healthcare settings, studies investigating adverse even rates
are often conducted. These studies commonly use retrospective chart review for detection and thus the
identified rates of events mirror the rate to which events are documentable [67]. While such studies are
important and each country should conduct AE studies regularly, using them as main indicator of improve-
ment is difficult and dangerous. Other measures, for example those obtained by routine measurements like
patient safety indicators, often lack validity [68].
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 44
Thus, detecting improvement over time using common measures of adverse events is difficult on methodo-
logical grounds. This leads to healthcare improvement basically flying blind [67] as long as preventable ad-
verse event rates are the main outcome targeted. Rather, valid measures tracking the incidence of spe-
cific events rigorously over time should be established, i.e., targeted surveillance for the type of adverse
events that are addressed by a specific intervention programme [69,70]. Currently, there exist only few
valid metrics that could be used to identify a change in patient safety in Swiss healthcare. A more
complex set of indicators of harm produced by healthcare should be developed and established, including
programme-specific indicators and leading indicators of past harm and future risks [71]. This system of
measures of harm should be accompanied by indicators of implementation of safety measures.
So, healthcare institutions implementing certain interventions should be able to track their own effort in as-
sessing outcome measures. A recent publication described the use of Statistical Process Control charts for
realizing this measurement [72] by clinical teams. They conclude that measuring improvement is a highly
technical task requiring specific expertise and recommend capacity-building programmes to improve meas-
urement capabilities, offering support and the development of repositories of validated measures that can
be assessed and used by practitioners. The programme progress! COM-Check Safe surgery of the foun-
dation actually is such a capacity-building effort to improve measurement capabilities.
Key message: Sound measurement and evaluation schemes need to be estab-
lished on national and institutional level to be able to generate evidence on whether
quality and safety improvement programmes actually reduce medical harm.
Second, large studies assessing the incidence of adverse events show that patient safety issues are still
pervasive [69,73]. Although such evidence on adverse events in Swiss healthcare is currently missing, we
expect that results from other countries are indicative for the situation in Switzerland. A recent meta-analysis
showed that large-scale patient safety initiatives only minimally reduced adverse event rates [66]. Thus,
despite the local improvement efforts, widespread reductions in rates of harm in medical care were not
achieved [69]. It is concluded that considering the large amounts of financial resources being invested in
improvements, the healthcare systems did not get much safer and key issues not appropriately addressed
[74]. There are several causes explaining the limited success of past improvement efforts. One of them lays
in the contradiction between the need to design an effective intervention that works as a whole in reducing
harm and the emerging need to change its parts in rolling it out to healthcare institutions. As described above,
an intervention needs to be tested before large-scale implementation in its fit to various contexts. However,
the demand for making changes to facilitate implementation to local contexts is still likely to emerge in an
improvement project. Thus, there is a dilemma and a fine line to be balanced between a dogmatic one size
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 45
fits all approach and a laisser-faire approach which may water down the initial idea and potential effects of
an intervention. While rapid cycles of local adaptations may certainly be conducive to implementation and
foster acceptance and participation of stakeholders, they also likely lead to adaptations. Major dangers are
related to allowing for extensive local adaptations: The logic model of the intervention may be inadvertently
changed in the process of adapting an intervention, which may be impacting is potential for effectiveness
[56]. Whether this is the case may be hard to realise, because, very often, the actual mechanisms that make
an intervention effective are not known. For example, for the surgical checklist, positive effects have been
shown, but it is still discussed whether the major driving forces behind these effects lay in the fact that each
team introducing the checklist needs to organise their work processes more mindfully, or whether the check-
list mainly acts as a cognitive aid or even as a communication tool.
Often, the mechanisms of change are not clear even after a programme has been shown to be effec-
tive for certain contexts. This makes it hard to reproduce success in other contexts: For example, an
intervention to reduce catheter-associated infections that was successful in the US could not be replicated
for the UK healthcare setting [75]. A study by Dixon-Woods and colleagues concluded that one of the reasons
for this was that the mechanisms of bringing the positive change in the US hospitals had not been fully
understood [55,76] by the implementation teams: Initially, it was concluded, that a technical checklist had
reduced the central line catheter infections in improving the care. However, in post-hoc interviews, it was
revealed that a major factor was that clinicians and decision-makers saw the programme as something a
good intensive care unit needed to participate in [55]. The failing of the implementation of the programme in
the UK also was explained by the fact that the US programme was rather seen as a professional-led initiative,
while the UK-programme was seen more as a top-down government-driven initiative [75,76].
To conclude, in the current situation we cannot disentangle what the main problem is a) that we do not have
sufficient and robust measurement of the effects of programmes and interventions, or b) that the interven-
tions and programmes implemented are not sufficiently effective. It therefore is necessary to develop
measures indicating that a change has happened that complement national adverse event rates. Measure-
ment and tracking of improvements should be a part of each programme on every level of the
healthcare system targeted by the programme. To conclude, assessing and measuring progress in patient
safety on various levels, e.g., national, or institutional, is an important undertaking that urgently needs to be
tackled in order to provide the necessary resources for successful improvement programmes and to avoid
the flying blind mode [67]. Furthermore, the implementation of interventions in healthcare practice needs to
be organised in using the resources and tools discussed in chapter 2. Successful implementation needs to
rely on a systematic process, addressing barriers and facilitators to reach adoption of the intervention at the
front-end of care. To this end, expertise and resources in patient safety and implementation research needs
to be developed.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 46
4.2 Designing large-scale programmes: scenarios for three kinds of interventions
Besides the above outlined methodological aspects, interventions differ in how they actually work in address-
ing patient safety. Actions modifying healthcare workers’ behaviour to improve patient safety can be located
on a dimension from weak to strong [21]. Weak actions are for example issuing protocols or adding warning
labels, for example to avoid confusion errors in administering look-alike medicines. Accordingly, training
healthcare workers also represents a weak improvement action, as it is person- rather than system-centred
in its approach. Standardizing or automating processes is considered a more effective improvement action
as it addresses work on the system-level. The more integrated within the actual functioning of a system a
certain measure, i.e., the less dependent from individual decision-making and motivation it is, the stronger it
is considered. Design changes, as for example placing hand disinfectant in spaces that are close and visible,
or creating forcing functions represent strong forms of improvement action. While designing a new training
or educational intervention is rather easy and quick to do, it is more resource- as well as knowledge-intensive
to develop interventions that lead to change at the system-level. The most effective form of protecting pa-
tients from a specific hazard is eliminating it completely out of the system [77]. Eliminating the possibility of
misconnecting Luer-Lock connectors in providing connectors that are connectable solely with the intended
medical applications would reduce the risk of harm coming from misconnectingfor example wrongly ad-
ministering intravenous drugs via feeding tubeto zero. Traditionally, strong systemic changes eliminating
or drastically reducing the potential for patient harm are not realised in Swiss healthcare despite their good
chances of success. For example, the probability of inadvertent intrathecal application of vincristine can be
enormously reduced in packaging it in minibags instead of syringes. This is a well-known recommendation
[78], that would avoid tragic, usually fatal errors. But the recommendation is not implemented at large scale
in Switzerland: Our recent survey showed that in 16 of 21 hospitals administering vincristine, it was still
delivered in syringes [79]. To conclude, the main focus is currently on changing the healthcare workers
behaviour to increase patient safety, while a great potential of systemic changes lays largely dormant in
Swiss healthcare.
Recommendation: Current approaches to improve patient safety in Swiss
healthcare focus mainly on worker behaviour, ignoring the large gains to be made in
changing systemically work or tool design. Strong measures eliminate or drastically
reduce the potential for error and thus patient harm. Structures need to be developed
that foster such systemic change.
One important aspect we want to bring up is that not only implementing new and safer ways of doing things
into healthcare is an important part of improving patient safety, but also getting rid of harmful or wasteful
practices. De-implementing practices that are unproven, inappropriate or even harmful is also an important
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 47
undertaking [80] that offers great potential for improving patient safety. For example, certain double check
procedures may rather be substituted by one technical check and a plausibility check in order to increase
the identification of medication errors [81], which would reduce interruptions and time resources allocated to
checking in certain contexts. De-implementation is expected to work differently from implementation [82] and
thus will not be discussed in depth here. However, from a long-term perspective, some implemented inter-
ventions will need to be scrutinised regularly to evaluate whether they are still having their intended effects
or whether the contexts have changed so that the practices are not useful anymore. This will particularly be
important for all information technology-related practices that are implemented. Additionally, the potential
unintended consequences of improvement interventions need to be accounted for in order to sustain the
positive effect of an implementation programme [83]. Therefore, the recommended ImpRes [84] tool features
the unintended effects to make programme managers and implementation teams aware of them.
In the following, we will describe the implementation process of three different kinds of interventions to illus-
trate how potential programmes for patient safety improvements could be designed (see Table 3). Patient
safety interventions targeting actual work practices at the front end (e.g., the safe administration of drugs)
can be differentiated from interventions that aim at increasing competencies of the workforce relevant for
safety (e.g., the ability to learn from errors as an individual and as an organisation). As a third kind of inter-
vention, we will illustrate a strong measure to improve patient safety. Please note that this a very brief presen-
tation of very complex projects that is necessarily passing over many aspects.
One of the important key questions in planning a programme is how much evidence an intervention has for
the context it should be implemented in. Is there empirical evidence corroborating the effectiveness of an
intervention from Swiss healthcare institutions? Or do we know from other countries and health systems that
an intervention worked well there? It may also be that there is only anecdotal evidence or a high face validity.
All these aspects define how the programme is designed, and how much of effectiveness and implementa-
tion research needs to be incorporated. The examples reported in Table 3 are special for Swiss healthcare
in the sense that they rely on a good evidence base regarding their effectiveness (specifically scenarios
1 and 3). In contrast, many of the pressing issues and available ideas for interventions to improve patient
safety have only little evidence and often no evidence from the Swiss healthcare context.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 48
Table 3. The implementation process of three different kinds of interventions
Intervention
Steps
Scenario 1:
Improving a clinical practice
Scenario 2:
Improving a patient safety-rele-
vant competency
Scenario 3:
Improve the work system to
avoid errors
Reducing catheter-associ-
ated urinary tract infec-
tions
Improve learning through
Morbidity-Mortality-confer-
ences (Mo-Mos)
Avoid inadvertent intrathe-
cal vincristine administra-
tion using minibags
Situation at start
of programme
progress! safe urinary cathe-
terisation (20152018) de-
veloped an evidence-based
intervention bundle in hospi-
tals to reduce the use of in-
dwelling urinary catheters
and promote safe catheter
insertion and maintenance.
- Measures on clinical out-
comes have been devel-
oped (catheter utilization,
catheter-associated urinary
tract infection and non-in-
fections complications).
- As the intervention bundle
has been tested before, a
hybrid setting (type 3) is
chosen: Implementation
and effectiveness are as-
sessed concurrently.
Patient Safety Foundation de-
veloped a manual for improving
Mo-Mos in prior project.
The manual is based on evi-
dence about Mo-Mos in Swiss
hospitals, as well as on interna-
tional evidence. It has been
adopted and is being used by
leaders who are conducting
Mo-Mos in their units on a vol-
untary basis.
- Implementation will be as-
sessed simultaneously with
effectiveness (hybrid type 2).
Patient Safety Foundation is-
sued a factsheet and Q&A ex-
plaining the risk of inadvertent
intrathecal vincristine applica-
tion and recommending the
use of infusion minibags for
risk reduction. Our survey
study had shown low use of
minibags in Switzerland.
Setting
Acute care hospitals
All acute care hospitals
All acute care institutions that
administer vincristine
Step 1: Efficacy
(could it work?)
Efficacy established in prior
pilot programme
Efficacy established in prior
project.
Efficacy is clear, packaging in
minibag is feasible and effec-
tive.
Step 2: Effec-
tiveness
(does it work?)
Effectiveness is assessed
during implementation pro-
cess.
Measures of improved learning
from errors in terms of patient
outcomes are developed:
- Perceived psychological
safety during Mo-Mos will be
assessed (using a survey af-
ter the Mo-Mo);
- the improvement actions will
be evaluated in their actual
effect on the safe provision of
care (using document analy-
sis and observation).
Effectiveness is established,
minibags for infusion reduce
the chance of error.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 49
Step 3a: Imple-
mentation
(make it work)
- Implementation measures
need to be developed.
- Best implementation strate-
gies are used for scale-up
to national level.
- Implementation tools are
adapted to different lan-
guages. A network region-
ally organised of imple-
mentation consultants sup-
ports implementation.
Measuring implementation suc-
cess: The use of tools and pro-
cedures recommended by the
manual are assessed (self-re-
ports, protocol analysis). Man-
ual exists in German, Italian
and French. It may be adapted
after the exploration phase.
Support structures for health-
care institutions adopting the
bundle need to be provided.
In the implementation phase,
facilitators and barriers for
adoption will be addressed.
Implementation is assessed in
using fidelity measures: How
often patients receiving vin-
cristine therapy still receive it
in syringes will be estimated
in surveying the hospital phar-
macies.
Step 3b: Na-
tional scale-up
(mandatory vs.
voluntary)
As the intervention bundle
will be well-tested, it is pro-
posed to choose a directive
implementation strategy, i.e.,
to make the implementation
of the intervention bundle
mandatory and a part of the
contract between providers
and assurers.
Leading physicians as im-
portant facilitators/stakeholders
play a key role. A dissemina-
tion strategy is used, with a
communication campaign to
enhance the (voluntary) use of
the manual, combined with of-
fering expert support for im-
proving Mo-Mos locally
Mandatory scale-up: regula-
tors and pharmacists could
make it a rule that vincristine
is only administered via mini-
bag in order to reduce wrong-
route administrations and en-
hance chemotherapy safety.
Step 4: Sustain
& evaluate
(keep it working)
Assess urinary catheter in-
fections and potentially inju-
ries at national level and
continue assessing effective-
ness and implementation
measures in selected set-
tings periodically for evalua-
tion purposes.
Continue assessing fidelity
measures (use of recom-
mended tools; in some institu-
tions track the improvements
and psychological safety). As-
sess status of dissemination of
recommendations (survey with
leading physicians). Cooperate
with professional organisations
to include Mo-Mo expertise in
curricula.
Adoption of minibag use is
evaluated in surveying the
number of patients that re-
ceive vincristine in syringes
with hospital pharmacies
(penetration).
Other examples
for interventions
- Surgical safety checklist
- Intervention bundle to re-
duce interruptions in medi-
cation preparation and/or
administration
- Intervention minimising the
risk for RFO, retained for-
eign objects (surgical count
procedures)
- Avoiding overdoses and
drug interactions in improv-
ing medication ordering
(digitally supported)
- Establish and use institutional
error tracking and learning
systems
- Develop institutional capacity
to perform a root cause anal-
ysis in order to learn from
very serious events
- Learning from errors in pri-
mary care settings
- Improve medication prepa-
ration rooms (light, space,
air, noise) to reduce errors
- Avoid methotrexat over-
doses (daily instead of
weekly) in developing safer
order processes using IT /
providing packaging design
- Improve connector safety in
designing connectors that
only connect with desig-
nated application route
- Enhance the design and the
use of Smart Pumps to re-
duce infusion errors
- Reduce sound a-like and
look-alike medicines to
avoid confusion errors
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 50
4.3 Suggested topics for future national programmes in Switzerland
As described in section 1.6, several fields of action relevant to Switzerland could be identified, but only a
limited number of solutions with good evidence to improve in these fieldsin particular with good evidence
for the Swiss context. Contrary to what is sometimes suggested, there is no drawer filled with ready-made
solutions to big problems that only need to be implemented. From this diagnosis, different strategies could
be concluded.
First, one approach would be to first conduct comprehensive research and development in Switzerland to
develop solutions for the fields with large potential for improvement. This would mean, however, that for the
time being, little improvements with regard to patient safety would be pursued. As a few measures and
bundles of measures already exist (see Table 3) that have also proven to be relevant and effective for the
Swiss context, this approach would mean delaying their implementation.
The second approach, to the other extreme, would be to first focus on solutions, i.e., improvement interven-
tions, that have already proven effective somewhere, not taking into account the relevance of the problem.
This strategy would ignore the relevance of the problem: This means that measures might be implemented
to solve a problem that is actually not that big locally. For example, there are well-tested programmes for
early sepsis recognition and treatment; however, official incident rates for sepsis in Switzerland seem to
suggest that sepsis is not one of the most pressing patient safety problems in Switzerland. Without prior
study into the extent to which sepsis management can be improved nationally, it seems that implementing
the solution cannot be justified. The situation is similar with pressure ulcers: There are already intervention
bundles with good evidence. However, the data collected by the Swiss National Association for Quality De-
velopment in Hospitals and Clinics (ANQ) indicate that pressure ulcers are not an urgent patient safety prob-
lem in Swiss hospitals [85].
The third strategy from this would be to look for appropriate matches, i.e., what fields of action are there in
Switzerland for which effective solutions already exist? However, only following existing problem-solution
matches would ignore urgent problem areas where there are simply no good solutions available yet, as well
as innovative approaches that may have face validity but have not yet been well enough empirically studied
for large scale implementation. Only focusing on the combination relevant problemgood solution
available is therefore not an adequate strategy from our perspective. We therefore propose to develop
a) a suitable combination of fields of action and available bundles of measures, to quickly achieve
improvements with specific, well-researched and feasible measures; and at the same time b) to con-
duct research and develop improvement activities on urgent fields of action to explore viable, new
solutions. In designing the future topics of programmes, additional to the combinations of improvement
interventions and targeted problems it should be considered, which professional groups will be affected by
the improvement activities (nursing, physicians, administrative-technical staff, etc.), which healthcare sectors
(inpatient, outpatient, rehabilitation, etc.) and target populations (children, older adults, etc.).
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 51
Table 4 shows a range of relevant problem areas for Switzerland, categorised by the relevance of the prob-
lem (major vs. moderate) and the availability and feasibility of solutions (no or little vs. practicable solutions
available) resulting in three problem-solution-combinations: (1) Major problems with no or little practicable so-
lution yet; (2) Moderate problems with practicable solution; (3) Major problems with practicable solution. Moderate
problems with no or little practicable solutions are not listed as they are the lowest priority. Based on the
relevance of the problem and the availability, feasibility, and scope of solutions we give recommendations
for action.
Table 4. Range of relevant problem areas for Switzerland categorised by the relevance of the problem (major vs. mod-
erate) and the availability and feasibility of solutions (no or little vs. practicable solutions available).
Major problems with no or little practicable solution
Problem area
Solutions: availability, feasibility, and scope
Recommendation
Alarm fatigue
Little promising solutions available (mostly indirect
approaches that aim to improve the safety culture in
general and train nurses in setting individual alarm
limits)
Adapt/extend existing approaches
and explore new solutions
Care of people with
mental health problems
No / little promising solutions available (e.g., physi-
cal restraint, suicide prevention)
Conduct research to identify main
problem areas and explore new
solutions; test new instruments for
adverse event detection in mental
care
Diagnostic errors
Some promising findings for clinical decision sup-
port and staff education and training available but
results are ambiguous and methodological quality of
most studies is low
Adapt existing approaches to the
Swiss context and test them with
rigorous study designs; explore
new solutions
Delirium prevention
Some multicomponent interventions with multidisci-
plinary teams show promising effects, but studies
are very different and hard to compare; quality of
evidence is low to moderate; focus on older adults
Explore which components of
bundles are most effective for de-
lirium prevention and test which
are effective for the Swiss context
Early identification of
clinical deterioration in
hospital patients
Validated instruments (early warning scores) in Eng-
lish-speaking countries available, recommendations
and early experiences available for Switzerland, but
little evidence on effectiveness for Switzerland
Adapt existing approaches to the
Swiss context and test their effec-
tiveness
Information transfer at
care process interfaces
Some promising solutions available (e.g., staff train-
ing; written care plans for patients) but mixed results
and mostly very specific for one interface (e.g., hos-
pital discharge; transition from emergency depart-
ment to acute care) or for one professional group
Adapt and test existing ap-
proaches for the Swiss context;
explore new solutions that target
all involved professional groups
(nurses, doctors, etc.) and differ-
ent interfaces (e.g., between
emergency department, acute
care, rehabilitation services)
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 52
Medication: Administra-
tion and Preparation er-
rors
Some promising solutions available (e.g., smart
pumps, separate preparation rooms) but overall
mixed results (e.g., for staff training); little re-
search/evidence on effectiveness of structured pro-
cess changes and redesigned workflows
Test effectiveness and feasibility
of structural changes in the Swiss
context; adapt and test promising
solutions that aim at behaviour
change (staff training)
Medication: Prescribing
errors
Some promising solutions available (e.g., detec-
tion/alert systems; pharmacist involvement) but
most approaches are very hospital-specific, diffi-
cult/resource-intensive to implement in other con-
texts and limited to the inpatient setting.
Adapt existing approaches to the
Swiss context and test their effec-
tiveness; explore new solutions
(including ambulatory care)
Polypharmacy
Some promising solutions available with improve-
ments for patient safety and economic outcomes
(Geriatrician and clinical pharmacist reviews, patient
education, STOPP criteria) but long-term clinical uti-
lization is rarely reported; transition between care
facilities is considered rarely; focus on older adults
Adapt existing approaches to the
Swiss context and test their effec-
tiveness; explore new solutions
(with focus on transition between
care facilities)
Moderate problems with practicable solution
Problem area
Availability, feasibility, and scope of solutions
Recommendation
Catheter Associated
Urinary Tract Infection
Effective intervention bundles available, already
tested in Swiss hospitals (see section 3.2), feasible
to be implemented in Swiss hospitals
Prepare intervention bundle for
national scale-up
Major problems with practicable solutions
Problem area
Availability, feasibility, and scope of solutions
Recommendation
Errors in surgery
(wrong patient, wrong
site, retained foreign
objects)
Effective intervention bundle for surgical safety
checklist available, already tested in Swiss hospi-
tals, feasible to be implemented in hospitals and fur-
ther sectors
Prepare intervention bundle for
national scale-up; extend and test
for other sectors (e.g., ambulant
setting)
Inadvertent intrathecal
administration of Vin-
cristin
Effective intervention available; change of packag-
ing most promising
Check how safe packaging can be
made mandatory; support the im-
plementation process in providing
information
Nutritional support
Effective intervention bundles available, already ap-
plied in Swiss hospitals with high level evidence
Test and prepare intervention
bundle for national scale-up
Patient Blood Manage-
ment
Effective intervention bundles available, already ap-
plied in Swiss hospitals with high level evidence
Test and prepare intervention
bundle for national scale-up
Eliminating Luer-Lock
misconnections
Promising solutions available (connectors that are
only connectable with intended medical applica-
tions)
Support use of switch to connect-
ors with a national programme tar-
geting implementation, provision
of training, recommendations for
purchasing departments, etc.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 53
We recommend approaching several problem areas from all three problem-solution-combinations at the
same time to advance patient safety on a broad scale. The list presented in Table 4 is not comprehensive
and serves as example. It covers interventions that aim at making clinical processes safer. For sustainably
changing healthcare, investing in the development of a better safety culture throughout the whole system is
crucial. Therefore, programmes targeting meta-competencies should also be conducted. For example, learn-
ing from morbidity-mortality conferences, from incident reports, or from the analysis of very serious events
are such organisational processes that could be enhanced with targeted programmes. Furthermore, com-
munication in emergency situations can be trained for example, enhancing the teamwork competencies of
the healthcare workforce.
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 54
Literature
1 National Cancer Institute. Implementation Science at a Glance. 2019.
https://cancercontrol.cancer.gov/sites/default/files/2020-07/NCI-ISaaG-Workbook.pdf
2 Vincent C, Staines A. Enhancing the Quality and Safety of Swiss Healthcare. Report for the Federal
Office of Public Health (FOPH). Bern: 2019.
https://www.bag.admin.ch/bag/en/home/versicherungen/krankenversicherung/krankenversicherung-
qualitaetssicherung.html
3 World Health Organization. The Conceptual Framework for the International Classification for Patient
Safety; Version 1.1. 2009. https://www.who.int/patientsafety/taxonomy/icps_full_report.pdf
4 Waltz TJ, Powell BJ, Matthieu MM, et al. Use of concept mapping to characterize relationships among
implementation strategies and assess their feasibility and importance: results from the Expert
Recommendations for Implementing Change (ERIC) study. Implement Sci 2015;10:109.
doi:10.1186/s13012-015-0295-0
5 Althoff FC, Neb H, Herrmann E, et al. Multimodal Patient Blood Management Program Based on a
Three-pillar Strategy. Ann Surg 2019;269:794804. doi:10.1097/SLA.0000000000003095
6 Kaserer A, Rössler J, Braun J, et al. Impact of a Patient Blood Management monitoring and feedback
programme on allogeneic blood transfusions and related costs. Anaesthesia 2019;74:153441.
doi:10.1111/anae.14816
7 Schuetz P, Fehr R, Baechli V, et al. Individualised nutritional support in medical inpatients at
nutritional risk: a randomised clinical trial. Lancet 2019;393:231221. doi:10.1016/S0140-
6736(18)32776-4
8 Westbrook JI, Coiera E, Dunsmuir WTMM, et al. The impact of interruptions on clinical task
completion. Qual Saf Heal Care 2010;19:2849. doi:10.1136/qshc.2009.039255
9 Agency for Healthcare Research and Quality (AHRQ). Strategies to Improve Patient Safety: Draft
Report to Congress for Public Comment and Review by the National Academy of Medicine. 2020.
https://pso.ahrq.gov/sites/default/files/wysiwyg/strategies-to-improve-patient-safety_draft-report.pdf
10 Mekonnen AB, McLachlan AJ, Brien JAE. Effectiveness of pharmacist-led medication reconciliation
programmes on clinical outcomes at hospital transitions: A systematic review and meta-analysis. BMJ
Open. 2016;6. doi:10.1136/bmjopen-2015-010003
11 Drovandi A, Robertson K, Tucker M, et al. A systematic review of clinical pharmacist interventions in
paediatric hospital patients. Eur. J. Pediatr. 2018;177:113948. doi:10.1007/s00431-018-3187-x
12 Lee JK, Slack MK, Martin J, et al. Geriatric patient care by U.S. pharmacists in healthcare teams:
Systematic review and meta-analyses. J. Am. Geriatr. Soc. 2013;61:111927. doi:10.1111/jgs.12323
13 Tan ECK, Stewart K, Elliott RA, et al. Pharmacist services provided in general practice clinics: A
systematic review and meta-analysis. Res Soc Adm Pharm 2013;:116.
doi:10.1016/j.sapharm.2013.08.006
14 Mekonnen AB, Abebe TB, McLachlan AJ, et al. Impact of electronic medication reconciliation
interventions on medication discrepancies at hospital transitions: a systematic review and meta-
analysis. BMC Med. Inform. Decis. Mak. 2016;16:114.http://dx.doi.org/10.1186/s12911-016-0353-9
15 Marien S, Krug B, Spinewine A. Electronic tools to support medication reconciliation: A systematic
review. J. Am. Med. Informatics Assoc. 2017;24:22740. doi:10.1093/jamia/ocw068
16 Baysari MT, Lehnbom EC, Li L, et al. The effectiveness of information technology to improve
antimicrobial prescribing in hospitals: A systematic review and meta-analysis. Int. J. Med. Inform.
2016;92:1534. doi:10.1016/j.ijmedinf.2016.04.008
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 55
17 Härkänen M, Voutilainen A, Turunen E, et al. Systematic review and meta-analysis of educational
interventions designed to improve medication administration skills and safety of registered nurses.
Nurse EducToday 2016;41:3643.
18 Raban MZ, Westbrook JI. Are interventions to reduce interruptions and errors during medication
administration effective?: A systematic review. BMJ Qual Saf 2014;23:41421. doi:10.1136/bmjqs-
2013-002118
19 World Health Organization. Patient safety: global action on patient safety: report by the Director-
General. Exec Board 2018;144.https://apps.who.int/iris/bitstream/handle/10665/327526/B144_29-
en.pdf?sequence=1&isAllowed=y
20 Mascherek AC, Schwappach DLB. Patient safety priorities in mental healthcare in Switzerland: a
modified Delphi study. BMJ Open 2016;6:e011494. doi:10.1136/bmjopen-2016-011494
21 Trbovich P, Shojania KG. Root-cause analysis: swatting at mosquitoes versus draining the swamp.
BMJ Qual Saf 2017;:bmjqs-2016-006229. doi:10.1136/bmjqs-2016-006229
22 Reames BN, Krell RW, Campbell DA, et al. A checklist-based intervention to improve surgical
outcomes in michigan: Evaluation of the keystone surgery program. JAMA Surg.
http://dx.doi.org/10.1001/jamasurg.2014.2873
23 Dixon-Woods M. How to improve healthcare improvementan essay by Mary Dixon-Woods. Bmj
2019;:l5514. doi:10.1136/bmj.l5514
24 Stephens TJ, Peden CJ, Haines R, et al. Hospital-level evaluation of the effect of a national quality
improvement programme: time-series analysis of registry data. BMJ Qual Saf 2020;29:62335.
doi:10.1136/bmjqs-2019-009537
25 Brown CH, Curran G, Palinkas LA, et al. An Overview of Research and Evaluation Designs for
Dissemination and Implementation. Annu Rev Public Health 2017;38:122. doi:10.1146/annurev-
publhealth-031816-044215
26 Curran GM, Bauer M, Mittman B, et al. Effectiveness-implementation Hybrid Designs: Combining
Elements of Clinical Effectiveness and Implementation Research to Enhance Public Health Impact.
Account Patient Saf Policy Reform 2012;50:21726. doi:10.1097/MLR.0b013e3182408812
27 The Breakthrough Series: IHI’s Collaborative Model for Achieving Breakthrough Improvement. IHI
Innova. Boston: : Institute for Healthcare Improvement 2003. www.IHI.org
28 Ebert L, Amaya-Jackson L, Markiewicz JM, et al. Use of the Breakthrough Series Collaborative to
Support Broad and Sustained Use of Evidence-Based Trauma Treatment for Children in Community
Practice Settings. Adm Policy Ment Heal Ment Heal Serv Res 2012;39:18799. doi:10.1007/s10488-
011-0347-y
29 Hempel S, O’Hanlon C, Lim YW, et al. Spread tools: a systematic review of components, uptake, and
effectiveness of quality improvement toolkits. Implement Sci 2019;14:83. doi:10.1186/s13012-019-
0929-8
30 Glasgow RE, Vinson C, Chambers D, et al. National Institutes of Health Approaches to Dissemination
and Implementation Science: Current and Future Directions. Am J Public Health 2012;102:127481.
doi:10.2105/AJPH.2012.300755
31 Tabak RG, Khoong EC, Chambers DA, et al. Bridging Research and Practice Models for
Dissemination and Implementation Research. Am J Prev Med 2012;43:337.
doi:10.1016/j.amepre.2012.05.024
32 Goodrich DE, Miake-Lye I, Braganza MZ, Wawrin N KA. Quality Enhancement Research Initiative.
QUERI Roadmap for Implementation and Quality Improvement. 2020.
www.queri.research.va.gov/tools/roadmap/
Stiftung Patientensicherheit Schweiz · Asylstrasse 77 · CH-8032 Zürich
T +41 43 244 14 80 · info@patientensicherheit.ch · www.patientensicherheit.ch 56
33 Sackett DL, Rosenberg WMC, Gray JAM, et al. Evidence based medicine: what it is and what it isn’t.
BMJ 1996;312:712. doi:10.1136/bmj.312.7023.71
34 Wade A, Toumi I, Hemels M. A pharmacoeconomic evaluation of escitalopram versus citalopram in
the treatment of severe depression in the United Kingdom. Clin Ther 2005;27:48696.
doi:10.1016/j.clinthera.2005.03.007
35 Moullin JC, Sabater-Hernández D, Fernandez-Llimos F, et al. A systematic review of implementation
frameworks of innovations in healthcare and resulting generic implementation framework. Heal Res
policy Syst 2015;13:16. doi:10.1186/s12961-015-0005-z
36 King’s Improvement Science. Implementation science research development (ImpRes) tool. A
practical guide to using the ImpRes tool. 2018. http://www.kingsimprovementscience.org/ImpRes
37 Hull L, Goulding L, Khadjesari Z, et al. Designing high-quality implementation research: development,
application, feasibility and preliminary evaluation of the implementation science research
development (ImpRes) tool and guide. Implement Sci 2019;14:80. doi:10.1186/s13012-019-0897-z
38 King’s Improvement Science. Implementation science research development (ImpRes) tool. A tool to
improve the quality of implementation projects. 2018.
http://www.kingsimprovementscience.org/ImpRes
39 Barker PM, Reid A, Schall MW. A framework for scaling up health interventions: lessons from large-
scale improvement initiatives in Africa. Implement Sci 2015;11:12. doi:10.1186/s13012-016-0374-x
40 Proctor E, Silmere H, Raghavan R, et al. Outcomes for Implementation Research: Conceptual
Distinctions, Measurement Challenges, and Research Agenda. Adm Policy Ment Heal Ment Heal
Serv Res 2011;38:6576. doi:10.1007/s10488-010-0319-7
41 Nilsen P. Making sense of implementation theories, models and frameworks. Implement Sci
2015;10:53. doi:10.1186/s13012-015-0242-0
42 Aarons GA, Hurlburt M, Horwitz SM. Advancing a Conceptual Model of Evidence-Based Practice
Implementation in Public Service Sectors. Adm Policy Ment Heal Ment Heal Serv Res 2011;38:423.
doi:10.1007/s10488-010-0327-7
43 Moullin JC, Dickson KS, Stadnick NA, et al. Ten recommendations for using implementation
frameworks in research and practice. Implement Sci Commun 2020;1:42. doi:10.1186/s43058-020-
00023-7
44 Moullin JC, Dickson KS, Stadnick NA, et al. Systematic review of the Exploration, Preparation,
Implementation, Sustainment (EPIS) framework. Implement Sci 2019;14<