ArticlePDF Available

Improving In-Hospital Care For Older Adults: A Mixed Methods Study Protocol to Evaluate a System-Wide Sub-Acute Care Intervention in Canada



Introduction: Acute care hospitals often inadequately prepare older adults to transition back to the community. Interventions that seek to improve this transition process are usually evaluated using healthcare use outcomes (e.g., hospital re-visit rates) only, and do not gather provider and patient perspectives about strategies to better integrate care. This protocol describes how we will use complementary research approaches to evaluate an in-hospital sub-acute care (SAC) intervention, designed to better prepare and transition older adults home. Methods: In three sequential research phases, we will assess (1) SAC transition pathways and effectiveness using administrative data, (2) provider fidelity to SAC core practices using chart audits, and (3) SAC implementation outcomes (e.g., facilitators and barriers to success, strategies to better integrate care) using provider and patient interviews. Results: Findings from each phase will be combined to determine SAC effectiveness and efficiency; to assess intervention components and implementation processes that 'work' or require modification; and to identify provider and patient suggestions for improving care integration, both while patients are hospitalized and to some extent after they transition back home. Discussion: This protocol helps to establish a blueprint for comprehensively evaluating interventions conducted in complex care settings using complementary research approaches and data sources.
Improving In-Hospital Care
For Older Adults: A Mixed
Methods Study Protocol to
Evaluate a System-Wide
Sub-Acute Care Intervention
in Canada
Dr. Malcolm Doupe
Manitoba Centre for Health
Policy, University of Manitoba,
408–727 McDermot Ave,
Winnipeg, Manitoba, Canada
sub-acute care; mixed
methods; program evaluation;
administrative data;
implementation measures
Doupe MB, Enns JE, Kreindler S,
Brunkert T, Chateau D, Beaudin
P, Halas G, Katz A, Stewart T.
Improving In-Hospital Care For
Older Adults: A Mixed Methods
Study Protocol to Evaluate
a System-Wide Sub-Acute
Care Intervention in Canada.
International Journal of
Integrated Care, 2022; 22(1):
25, 1–14. DOI: https://doi.
*Author affiliations can be found in the back matter of this article
Introduction: Acute care hospitals often inadequately prepare older adults to
transition back to the community. Interventions that seek to improve this transition
process are usually evaluated using healthcare use outcomes (e.g., hospital re-visit
rates) only, and do not gather provider and patient perspectives about strategies to
better integrate care. This protocol describes how we will use complementary research
approaches to evaluate an in-hospital sub-acute care (SAC) intervention, designed to
better prepare and transition older adults home.
Methods: In three sequential research phases, we will assess (1) SAC transition
pathways and effectiveness using administrative data, (2) provider fidelity to SAC core
practices using chart audits, and (3) SAC implementation outcomes (e.g., facilitators
and barriers to success, strategies to better integrate care) using provider and patient
Results: Findings from each phase will be combined to determine SAC effectiveness
and efficiency; to assess intervention components and implementation processes that
‘work’ or require modification; and to identify provider and patient suggestions for
improving care integration, both while patients are hospitalized and to some extent
after they transition back home.
Discussion: This protocol helps to establish a blueprint for comprehensively evaluating
interventions conducted in complex care settings using complementary research
approaches and data sources.
2Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
Older adults are the fastest growing segment of our
population worldwide [1]. As life expectancy increases, so
does the number of people who have complex functional
challenges, chronic physical diseases, and mental illness
[2–6]. Nationally and internationally, 8% of community-
dwelling older adults 75-84 years old and 20% of those
85+ years old report challenges completing activities of
daily living tasks such as walking unassisted or preparing
meals [7, 8]. Three-quarters of older adults are chronically
ill [9] and one-quarter have three or more chronic diseases
[7]. Compared to their healthy peers, older people with
multi-morbidity are more likely to visit emergency
departments (EDs) [10–13], and when hospitalized, they
are at higher risk of experiencing infection [14], pressure
ulcers [15], delirium [16–18] and medication-related
errors [19, 20]. A recent systematic review reports that
nearly 20% of older people are re-admitted to hospital
within 30 days of their index separation [21], further
perpetuating this cycle of adverse events. The need for
hospital re-admission is at least partially attributed to
care integration challenges across providers and settings
(e.g., poor communication, and inadequate discharge
planning and implementation) [22–26]. Healthcare
administrators, providers and researchers have for
almost two decades been seeking to improve hospital-
to-home transitional care for vulnerable older people
According to Holland et al. (2003) [30], improving
hospital-to-home transitional care requires strategies to
enhance in-patient hospital care and discharge planning
processes [31–35], and also to improve follow-up support
once patients have relocated to other settings and/or
care environments [36–41]. Multiple terms are used to
describe the different components of this transitional
care continuum (e.g., sub-acute care, intermediate care,
post-acute care) [42], and in this protocol paper we
describe an approach to evaluate a sub-acute care (SAC)
program that was implemented in Winnipeg, Canada.
SAC programs typically occur in units or facilities that
are dedicated to provide time-limited care to medically
stable (sub-acute) hospital inpatients, are generally
designed to be integrated with and extend the traditional
acute care model so that patients have shorter lengths of
acute care hospital stay, involve multidisciplinary teams
who collaborate to better prepare patients for discharge,
and actively facilitate the patient transition process as
part of hospital discharge [29, 42, 43]. It is important to
note that (1) a range of SAC structures and processes
exist, and that no single model has been shown to have
clear and consistent advantages over others [29]; and (2)
most of these interventions have been evaluated using
metrics such as hospital length of stay [31–33], changes
in patient function [31, 32, 34], and healthcare costs [31,
35], collectively with mixed results. Recent reviews [21,
27, 44, 45] and critiques [46] of this literature conclude
that more diverse and synergistic evaluation approaches
are needed to assess not only intervention effectiveness,
but also to assess the ways in which transitional care
interventions can be better integrated across providers
and settings.
‘Transitional care’ is an umbrella term that describes how
patients are transferred from one healthcare setting or
care provider to another [24, 42]. While several researchers
have sought to enhance transitional care based on the
seminal ideas of Coleman [28, 47] and Naylor [39, 48] (e.g.,
having multidisciplinary teams prepare and enact plans),
ongoing improvements are required to better integrate
care transitions from a health system perspective (e.g.,
improving coordination across sectors and providers)
[49], from a ‘whole systems’ perspective (e.g., improving
linkages between hospitals and communities) [50], and
by more effectively engaging with providers (e.g., gaining
their perspectives on how to reduce care fragmentation)
[51] and patients (e.g., ensuring that strategies to
improve transitional care meet their personalized needs)
[52]. These perspectives have clear parallels to the
evaluation framework proposed by Proctor et al. (2011)
[53], who emphasize the importance of evaluating
complex healthcare interventions using traditional
Institute of Medicine [54] metrics (e.g., determining if
transitional care interventions reduce hospital re-visit
rates) combined with both implementation outcomes
and client perspectives. Proctor et al. (2011) purport that
implementation outcomes such as provider acceptability
and perceived appropriateness (i.e., perceptions that
a new approach has value over status quo and is
compatible with provider beliefs and organizational
culture), incremental cost and sustainability (e.g.,
opinions whether added time demands are worth it
and can be sustained), and fidelity to the intervention
(whether stakeholders can actually conduct core facets
of the intervention as planned) are collectively essential
to help differentiate between ineffective interventions
and promising practices that have been poorly deployed
(a termed call ‘implementation failure’ by Proctor et al.
[2011] [53]). Similarly, Kreindler’s Population, Capacity
and Process framework [55] proposes that effective
system redesign requires a clear understanding and
agreement about the intended target populations (e.g.,
if stakeholders feel that SAC patient eligibility criteria are
clearly defined and correct), if the right kind and adequate
number of providers is available to effectively care for
these patients, and if processes have been developed to
effectively link the two (e.g., properly matching provider
care to patient need).
3Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
Guided by this knowledge, our study aims to evaluate
an in-hospital SAC intervention implemented across
the Winnipeg Health Region in the city of Winnipeg,
Canada. We will evaluate the SAC intervention using
a mixed-methods and sequentially phased research
approach, first to evaluate intervention efficiency and
effectiveness (e.g., use administrative data to report on
the number and type of care transitions that SAC patients
experienced preceding, during, and post discharge from
SAC care), second to examine provider fidelity to the
intervention (e.g., use chart audits to help define the
type and frequency of care that SAC patients received,
including the preparation of hospital-to-home discharge
plans), third to investigate through provider interviews
SAC implementation processes (e.g., how providers
view the acceptability, perceived appropriateness, and
implementation costs of SAC), and forth to determine
through patient interviews additional perspectives
about program success and/or the need for further
modification. We will triangulate this complementary
evidence to determine the extent to which and the
reasons why the SAC intervention is succeeding or failing,
and to understand the strategies required to better
integrate this care transition program across providers
and settings. Lessons learned from this research will
help researchers and health system planners develop
a roadmap for more comprehensively evaluating large-
scale interventions conducted in complex healthcare
Canada has a publicly-funded universal healthcare
system that is governed by the Canada Health Act
[56] and delivered provincially. Winnipeg is located
in Manitoba, one of ten Canadian provinces with a
population of 1.4 million people [57]. The province has
five geographically diverse health regions; four of these
regions are rural or remote and the Winnipeg Health
Region is the only large metropolitan area (population
817,000). Most tertiary care specialized services in
Manitoba (e.g., cardiac surgery, neurology, and intensive
care) are provided in Winnipeg through six hospitals
comprising 2,085 inpatient beds.
In recent years, Winnipeg has had one of the longest
ED wait-times in Canada [58], above average lengths
of inpatient stay [59], and large numbers of alternate
level of care hospital patients [60]. Three recent reviews
have also stated that Manitoba’s health system is
both fragmented and inefficient [61–63]. In 2017, the
Winnipeg Health Region responded by launching a
major system transformation (called Healing our Health
System) designed to consolidate clinical services and
staff resources [64]. While three Winnipeg Health Region
hospitals (1,506 beds) have remained as traditional acute
care facilities, the remaining three hospitals (579 beds)
were re-purposed to provide sub-acute care, specialized
services (e.g., orthopaedic surgery, dialysis, geriatric
care), and to a lesser extent transitional care to nursing
homes. EDs at the three sub-acute sites were converted
to urgent care departments, intensive care units were
consolidated to the three acute hospital sites, and revised
EMS-ambulance routing algorithms and public media
campaigns were developed to help stream patients to
the appropriate site based on acuity. As well, a ‘Home is
Best’ policy was created, from which enhanced hospital-
to-home care transition services (Priority Home [65] and
Rapid Response Nursing Teams [66] were developed.
Winnipeg implemented SAC in a phased manner; 89
acute care beds in one hospital were converted to SAC in
October 2017, followed by 222 beds from two additional
hospitals in June/July 2019. As part of the Winnipeg
Health Region transformation strategy, SAC was
developed to help improve the efficiency of the overall
hospital system, first by concentrating resources for
higher acuity patients at a smaller number of sites, and
second by allowing SAC patients to be managed in units
with lower physician and nursing ratios but with more
allied health staff. Potential benefits of SAC are therefore
at both the system level (e.g., more efficient care for both
high and lower acuity patients) and the patient level (e.g.,
enabling SAC patients to receive more patient-centred
interprofessional care, and better preparing them to
return home).
Sub-acute care is a subservice of the Winnipeg
Health Region general medicine program. Similar to the
criteria used by others [31, 33], patients are eligible for
SAC if they (a) still require general but not acute levels
of medical care, (b) have stable vital signs and no
oxygen requirements, (c) are unlikely to decompensate
medically, and (d) do not require ongoing special care
(e.g., rehydration) or significant behaviour therapy. While
SAC patients are admitted via three major pathways
(on-site urgent care, off-site EDs, and off-site acute
care units), ambulance routing procedures and public
information campaigns have helped to ensure that most
patients with sub-acute care needs are admitted from
on-site urgent care departments. A Central Bed Access
service monitors system-wide capacity and flow and
helps to prioritize and coordinate off-site transfers into
To support SAC implementation, interprofessional
teams were redeployed from existing hospital units;
these staff received additional training in dementia
and gerontological care, and in SAC best practices and
care plan procedures. SAC care is initiated by registered
nurses and is authorized/adjusted by the attending
4Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
family physician within one day of patient admission.
Interprofessional teams participate in daily rounds to
create and amend care plans as needed, and to discuss
barriers to patient progress and transition. All aspects
of patient care (e.g., team meeting dates, decisions
made) are documented using clinical decision software.
Patient transition is a team-consensus decision based on
medical (e.g., stable vitals) and additional patient (e.g.,
functional status) criteria. Additional details about the
SAC intervention are provided in
Table 1
Guided by the aforementioned frameworks proposed by
both Proctor et al. (2011) [53] and Kreindler (2017) [55],
we have developed six research questions to evaluate
the SAC intervention in Winnipeg:
1. Who is the population of SAC patients and did this
change with time (from early to later stages of the
2. Does SAC operate efficiently, e.g., how many and
what type of transitions did SAC patients experience,
what were their hospital lengths of stay, what
type and intensity of care (priority home and rapid
response nursing home care services, follow-up visits
with primary care physicians) did SAC patients get
following hospital discharge?
3. How effective is SAC, e.g., did it result in prolonged
community living, fewer emergency department
visits and lower hospital re-admission rates?
4. To what extent did providers deliver the types of care
intended for patients while in a SAC bed, and did this
vary by select patient group?
5. Do providers feel that SAC is a valued and suitable
intervention to enhance hospital care? What factors
Service Purpose To (1) provide quality care to patients who require daily individual assessment, general medical care, and
interventions to enhance their functional, cognitive, psychosocial, and spiritual well-being; and (2) liaise with
various community-based and institutional programs to facilitate out-of-hospital patient transitions.
Patient Profile Adult patients 18+ years old with a general medical diagnosis and who have stable vitals; low/stable oxygen
requirements; are unlikely to decompensate, and; do not require acute specialised hospitalized services. The
target length of stay for SAC patients is 14–16 days.
Admission Pathways &
SAC patients are admitted (1) directly from the onsite urgent care departments (primary pathway); (2) offsite
from one of three emergency departments, or; (3) via transfer from an acute care medicine bed. A Central Bed
Access service provides a gate keeping function, usually prioritizing direct onsite (urgent care) admissions.
Team Composition,
Recruitment & Training
SAC patients are visited by an attending physician at least once daily and nursing care is provided by a mix
of registered and licensed practical nurses. SAC teams are comprised of an extensive complement of allied
health disciplines including clinical nutrition, speech-language pathology, occupational therapy, physiotherapy,
pharmacy, respiratory therapy, social work, spiritual health and therapeutic recreation. Specialist consults from
off-site programs such as psychiatry, geriatrics, and orthopedics are available as-needed. Hospital-based nurse
case coordinators liaise with a range of community and institutional (e.g., nursing home) staff to facilitate out-
of-hospital care transitions.
SAC staff were redeployed from previously-undifferentiated (acute/sub-acute mix) hospital units. Teams received
specialized training in dementia care and in use of the National Early Warning System (identifies patients at risk
of clinical deterioration). Teams also received a patient flow guide, and an operations manual that defines roles,
accountability, best practice procedures, and tools to help support & evaluate patient progress.
Care Planning &
Care plans are initiated by the attending family physician within one day of patient admission. Four processes
are used to support care planning, delivery & inter-professional collaboration. These include:
1) Daily Action Rounds: The entire care team meets daily to review care plans & to address barriers to patient
2) Complex Case Rounds: Teams have dedicated time to develop care plans (e.g., engaging with off-site staff)
for particularly complex patients.
3) Bedside White Boards: These tools are used to communicate important information to the patient/family
about care goals, to provide an estimated discharge date, and to name the care team members.
4) Patient Flow & Clinical Decision Software: All staff have access to real-time data on wait times, patient
admissions, discharges and bed availability. Care plans, staff meeting dates, patient progress and barriers to
discharge are updated continuously using clinical decision software.
Discharge Process &
Discharge planning begins at the time of patient admission and is supported by clinical decision software.
Patients are eligible for discharge when:
1) Vital signs are stable, nausea/vomiting is controlled, pain is appropriately managed, oxygen saturation
is above 90%, lab values are in an acceptable range, and patients are able to void sufficiently and
independently (with or without support), AND;
2) Team members agree that the patient is ready for discharge from a functional, psychosocial and cognitive
Table 1 Sub-Acute Care* Intervention Components.
* Termed “Lower Acuity Care” in the Winnipeg Health Region.
5Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
facilitated and/or impeded SAC implementation, and
how can SAC be better integrated across providers
and settings?
6. What do patients and their family/friend caregivers
think of SAC and what are important measures of
success and failure from their perspective? Which
aspects of SAC do they feel worked well, and what
recommendations do they have for change?
An overview of our evaluation strategy is provided in
Table 2
. We will use an explanatory mixed methods
sequential design, in part so that qualitative findings will
help to explain, elaborate and contextualize quantitative
results [67]. In Phase 1 (months 1–12), administrative
healthcare use records will be linked, first to define the
profile of SAC patients and to assess whether these
patient characteristics changed during the intervention,
and second to evaluate SAC using measures of efficiency
and effectiveness. The latter analyses will be conducted
overall and across time periods (earlier versus later
stages of SAC) and patient subgroups. These findings will
be used in part to guide audits of SAC patient medical
charts in Phase 2 (months 12–15), designed to assess
provider fidelity to SAC standard operating procedures.
Qualitative interviews will be conducted in Phase 3
(months 16–24), designed to explore the experiences
that providers and patients (as well as their family/friend
carers, where possible) have had with SAC, and to identify
their recommendations for change.
Data Sources and Study Variables. Manitoba’s population-
based healthcare system data repository is housed at
the Manitoba Centre for Health Policy. The Repository
contains information on every registered Manitoban
(>99% of the population) since 1970. While Repository
data are de-identified (names and addresses removed),
files are linked using a scrambled 9-digit personal health
identification number attached to each record using a
secure standard process [68].
Key repository files to be used in this study are listed
Table 3
. These data were selected given our teams’
experience measuring care transitions [69–71], to reflect
the ‘whole systems’ perspective of care transitions (e.g.,
ensuring that we describe key healthcare transitions
leading into, during, and post hospital care), and to reflect
the major outcomes assessed in the academic literature
[21, 72]. These data will be used for two purposes. First,
the Admission, Discharge and Transfer file provides date-
stamped and bed-specific durations of hospital stay,
from which we can define the SAC cohort. These patients
will be characterized by (a) socio-demographic factors
including patient age, sex, marital status and income
quintile; (b) the presence of chronic diseases (e.g.,
arthritis, COPD, diabetes, ischemic heart disease, stroke,
Alzheimer’s disease/dementia, and a measure of multi-
morbidity) using validated algorithms [73] based on past
hospitalizations, ambulatory care physician visits, and in
some instances, prescription drug dispensations; and (c)
their index hospitalization (i.e., hospital patient service
[74] and procedural codes [75] will be used to help define
patients’ acuity status during this hospital stay).
Administrative healthcare use records will also be
used to examine the following healthcare use outcomes
as discussed by Proctor (2010) [53] and in keeping with
the Institute of Medicine [54]:
• Efficiency. This includes transition pathways into SAC
(e.g., directly from home or after multiple ED visits);
intra-hospital transition patterns (e.g., one versus
multiple transitions preceding SAC, repetitive bouts
of SAC and non-SAC care, the proportion of total
hospital time spent in SAC); and continued support
post hospital discharge (e.g., how many patients
Phase 1 Healthcare Service Outcomes
- efficiency, effectiveness of the intervention
Linked person-level administrative
healthcare use records
Months 1–12
Phase 2 Implementation Outcomes
- fidelity of the provider to the intervention
Medical chart audits Months 12–16
Phase 3 Implementation Outcomes
- acceptability and feasibility of the intervention
- barriers and facilitators, strategies to enhance integrated care
Provider interviews Months 16–24
Patient/Informal Caregiver Experiences
- measures of success and failure of the intervention, strategies to
enhance integrated care
Patient/informal caregiver interviews
Phase 4 Integration of the Results Phases 1–3 Months 24–36
Table 2 Overview of Evaluation Domains, Data Sources and Study Timeline.
6Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
received home care post hospital discharge, the
length of time between discharge and their first
home care visit; and the type, intensity, and duration
of home care services that they received).
• Effectiveness. We will define SAC patients’ post
discharge rates of ED visits and hospital re-
admissions; and time prior to death and nursing
home admission. As per the literature [21, 72], these
outcomes will be measured at 30, 90 and 180 days
after hospital separation.
Cohort Development & Quantitative Analysis Plan. We
will allow for a three-month ‘settling in’ period during
which healthcare use data will not be used in the study.
While SAC commenced in October 2017, healthcare use
data will be analyzed from January 2018 to April 2020
(26 months; the study period).
Using the methods defined by Schneeweiss et al.
(2009) [76], we will use a machine learning algorithm
to generate a propensity score of SAC patients by their
existing chronic disease, previous healthcare use, and
index hospital visit profile. This propensity score will be
used to match truly exposed (SAC) patients to a cohort
of unexposed patients with similar profiles (hereafter
referred to as ‘SAC look-alikes’). This approach has been
shown to produce results similar to RCTs by identifying
maximally important confounders [77, 78]. A random
sample of conventional acute care patients (i.e., those
who had a negligible likelihood of being exposed) will
also be developed based on this knowledge. These
approaches will also be used to define a pre-SAC (April 1,
2015 to September 30, 2017) comparison group for the
combined SAC and SAC look-alike patients, and a separate
pre-SAC group for the conventional acute care patients.
In keeping with the overall goals of SAC (e.g., to enhance
quality of care for both SAC and acute care patients),
we hypothesize that (pre-SAC to SAC) improvements in
effectiveness and efficiency will be significantly greater
for each of the SAC and conventional acute care patient
groups versus the SAC look-alike group.
Three complementary stages of quantitative analyses
will occur. First, as per Kreindler’s Population, Capacity and
Process framework [55], descriptive and multivariable
modelling will occur to define the main sociodemographic
and chronic disease profile of SAC patients, to compare
this profile to the SAC look-alike and conventional study
groups, and to assess if the predominant profile of
SAC patients changed during the course of the study
period. Second, process control charts [79] will be used
to determine the pattern of healthcare use outcomes
during the study intervention (e.g., if outcomes improved
suddenly and dramatically versus slowly over time),
overall and by the major sub-groups of SAC patients. This
strategy will use historical (pre-SAC) data to create ‘usual
care’ sigma values, and 26 data points (duration of SAC in
months) to apply the Western Electric Rules for detecting
non-random deviations from this historical data [80].
Third, these findings will be used to guide multivariable
statistical modelling to determine study group-specific
improvements in healthcare use pre versus during the
SAC intervention, and whether the numerical size of these
improvements varied by key predetermined factors (e.g.,
by major patient groups, from early to later stages of the
intervention). Healthcare use outcomes will follow either a
Population Repository This file defines registered Manitobans by key socio-demographic factors (age, sex, marital status, income
quintile) and death date (using the Repository cancellation code).
Admission, Discharge, and
Transfer File
This file provides date-stamped and bed-level hospital use data and will be used to (1) identify (using bed
identifiers) SAC patients, and (2) define detailed hospital transitions pathways leading to and from SAC units.
Hospital Discharge Abstract
This file provides date- and site-stamped data on hospital use parameters, and up to 26 international
classification of disease (ICD-10-CA) codes to define patient’s admitting diagnosis and complications that
arise after hospital admission.
Emergency Department
Information System
This file provides date- and time-stamped records of emergency department visits by site and patient
Medical Claims This file provides date-stamped record on ambulatory care physician visits. One ICD-9-CM (clinical
modification) code is provided per visit.
Home Care This file provides the start- and end-date, volume and type of home care services received by each
registered Manitoban (e.g., to identify prevalence [before SAC] and incidence [after SAC] home care users).
Use of the Priority Home and Rapid Response Nursing programs are included in this overall file.
Nursing Home This file provides the admission and exit date of nursing home use (to determine SAC disposition status).
Supported Living This file provides the admission and exit date of congregate community housing use (to determine SAC
dispositions status).
Drug Program Information
This file provides dispensation-level data on prescription drugs dispensed from retail (not in hospital)
pharmacies (i.e., by their anatomical, therapeutic & chemical classification system).
Table 3 Administrative Datasets from the Data Repository used in this Study.
7Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
binomial distribution (e.g., hospital re-admission) in which
case we will use logistic regression, or a count distribution
(e.g., physician & ED visits) in which case we will use a
Poisson or negative binomial regression. Because there
are a small finite number of hospitals (N = 3) where the
SAC program operates, multilevel modelling techniques
will be used to account for data clustering.
Sample Size Estimation. From discussions with planners,
we anticipate having data on ∼~2,340 SAC patients during
the study period (on average, three patients are admitted
daily into SAC). Sample size estimates were calculated
using 90-day hospital re-admission rates as the outcome.
Using data from the systematic review of Le Berre et al.
(2017) [21], we anticipate secular re-admission rates
to range between 25% and 35%, and about an 18%
reduction in re-admission rates associated with SAC.
Based on these estimates, we will require at least 1,000
patients to detect statistically significant events. Given this
sample size calculation relative to our projected cohort
size, we anticipate conducting all phase 1 data analyses
across hospital sites combined. Descriptive results will
first be compared across hospitals, to justify this decision.
Sample size calculations are provided in Additional File 1.
Patient Selection. SAC care strategies are documented in a
clinical decision software program (see
Table 1
). Following
the 10% rule of thumb proposed by Gregory et al. (2008)
[81], these data will be analyzed for ~200 SAC patients
selected randomly within pre-defined strata defined by
Phase 1 results (e.g., by dementia status if Phase 1 results
show that SAC effectiveness differs significantly by this
versus other patient groups), thus permitting sub-group
comparisons. Our selection of patient strata will be made
by team consensus after reviewing phase 1 results, and
with consideration of the existing academic literature
(e.g., Le Berre et al. [2017] conclude that transitional
care interventions may be less effective for people with
congestive heart failure [21]). Combining Phase 1 and
2 results in this way, along with the interviews in Phase
3, will help us to understand the extent to which SAC
‘works’ overall and for sub-groups of patients (i.e., the
concept of intervention failure as defined by Proctor et
al. [2011] [53]), or conversely, whether provider fidelity
to core SAC practices varies across patient groups (the
concept of implementation failure as defined by Proctor
et al. [2011]).
Data Collection. SAC requires providers to undertake
specific activities within defined time limits (e.g., conduct
a full patient assessment within 24 hours, immediately
establish a rehab plan, conduct daily team meetings,
provide regular rehab therapy; see
Table 1
). Clinical
decision software will be reviewed, and for each criterion
teams will receive a score of ‘no compliance’ (i.e., care
activities did not occur), ‘partial compliance’ (i.e., activities
were documented but occurred less frequently than
required), and ‘full compliance’ (frequency of activities
documented as per guidelines). We will present results
descriptively and explore any associations between
patient characteristics and provider compliance.
Analysis Plan. A draft of the audit tool is provided in
Additional File 2. To refine this tool, team members will
first conduct an informal focus group with 4-6 providers,
asking how care is recorded and the meaning of the
terminology used. Two auditors will then independently
review the data for 15 patients, compare results and
adjust the audit tool as necessary. Additional strategies
will be used to optimize rigor. First, auditors will not be
shown the administrative data results and will also be
blinded to the different strata (e.g., patients with and
without dementia) of charts selected. Second, auditors
will review ~20 of the same charts early in data collection
to measure inter-rater reliability. Re-training will occur if
kappa values [82] from this comparison are below 0.6.
We will interview 4-6 key informants (planners and
providers involved in the program) to verify that we
properly understand how the SAC intervention should
ideally work, eligibility criteria for SAC, and how these
processes have changed with time. These key informants
will also help us to refine provider interview questions and
to identify important groups of providers to interview.
Interviews will then be conducted with providers
who refer patients to SAC units and also with providers
comprising the interprofessional SAC care team
Table 1
). Using purposive sampling to ensure
representation of both referring and receiving staff,
participants will be recruited via e-mail, word of mouth
and staff meetings, and through snowball sampling.
Sampling will proceed until data have reached thematic
saturation. Based on the experiences of others [83], to
reach saturation we anticipate conducting interviews on
about 20 SAC-referrers and 20 SAC-providers.
Data Collection & Analysis Plan. Interviews will be
conducted using a semi-structured guide. Interview
questions will inquire about the SAC population (e.g., if
respondents feel that the ‘right’ patients are currently
being served by SAC, and if not, who should/should
not be accepted), provider knowledge and awareness
(e.g., about the overall SAC purpose and individual
responsibilities), the SAC intervention (e.g., if people feel
that SAC is an acceptable and feasible alternative to status
quo, suggestions for improvement), implementation
facilitators and barriers (e.g., whether clear duties and
operational guidelines are defined; if staff feel they were
adequately trained, are sufficient in number, and as a
team have the right complement of expertise to care for
SAC patients), and strategies to enhance integrated care
(e.g., between referring and receiving staff, amongst the
SAC multidisciplinary team, between the hospital and
8Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
community sectors). Additional interview questions will
be formulated depending on the results of Phase 1 (e.g.,
asking providers to help explain why SAC ‘worked better’
for some patient groups than others, to provide potential
solutions to demonstrated challenges).
Provider interviews will be audiotaped and transcribed
verbatim. Qualitative analyses, proceeding from a
interpretivist paradigm, will focus on describing the
semantic content of participants’ responses. The analysis
will have deductive and inductive phases. First, content
analysis, guided by a preliminary coding scheme informed
by Proctor et al. (2011) [53] and Kreindler (2017) [55] will
provide a descriptive account of participants’ responses.
The data will then be revisited to identify new themes
inductively, paying special attention to previously
uncoded text. Differences in themes across participant
groups (e.g., referrers vs receivers) will be explored to
determine the extent to which understandings are shared
or divergent. Each analysis phase will be undertaken by
two independent coders, who will meet frequently to
discuss discrepancies and reach consensus. Preliminary
findings will be shared with our diverse stakeholder
team to enable participant validation of themes and key
Selection of Interviewees and Data Collection. Team
members will randomly select about 30 charts of SAC
patients discharged from hospital in the last six months.
While this timeline will not permit us to compare
the perspectives of patients who received care in the
earlier versus latter stages of SAC, it was chosen to
help minimize recall bias. To help minimize selection
bias, we will attempt to recruit patient participants in
proportion to SAC cohort by age group (<65, 65–84, 85+
years old) and sex (about five participants per age- and
sex-category) and more generally by socio-economic
status (e.g., lower versus higher income quintiles) and
hospital (three sites). Patient names will be sent to the
Manitoba Government, who will in turn send patients an
information letter inviting them (and by extension, their
family/friend caregivers) to contact researchers directly
for more information and an interview. Patients within
each stratum will be oversampled by 50% to account
for non-responds and those who do not consent to
participate in the study. Interviewers and analysists will
also be blinded to select strata criteria (e.g., hospital
status) to help further minimize bias. Lastly, at the
beginning of each interview, participants will be asked
to recall some basic details about their SAC care (e.g.,
the approximate dates that they were hospitalized and
their duration of stay, the type of services they received
upon returning to the community). These participant
responses will be compared to our phase 1 records,
and if similar, we will conclude that participants have
sufficient recall to participate in the interview. People
who cannot accurately recall these basic details (e.g.,
that they were hospitalized or received home care post
return the community), will not be eligible to complete
the interview.
Eligible interview participants, and where possible,
their family/friend carers, will be jointly asked about
their awareness of SAC (e.g., how they were introduced
to the program), their perceived effectiveness of the
intervention, and important outcomes and suggestions
they have for making improvements. A more detailed
interview guide will be developed in consultation with
our patient partners, who will in turn gain feedback from
a larger patient and caregiver advisory group. Patient and
caregiver interviews will be audiotaped and transcribed
verbatim. Deductive and inductive analysis will be carried
out in the same manner as described above.
Quantitative and qualitative methods will be integrated
at several junctures. First, as research phases are
sequential, findings from earlier phases will guide the
sampling frame and questions posed in latter phases.
Second, towards the end of provider interviews in
Phase 3, we will give respondents the opportunity to
react to specific Phase 1 and 2 findings, inviting them
to offer potential explanations of these results. Third,
preliminary quantitative and qualitative findings will
be combined into a matrix to help identify convergent
and divergent results. Following the strategies proposed
by Rossman and Wilson (1985)[84], the quantitative
and qualitative findings will be synthesized to leverage
analytical corroboration and elaboration, which will
be discussed among the team to help initiate new or
modified interpretations of existing study results, and/or
to suggest potential areas of follow-up analysis.
This research uses a mixed method design to evaluate a
sub-acute care intervention that has been implemented
across the Winnipeg Health Region in Manitoba, Canada,
designed to improve in-hospital and hospital-to-
home care transitions for older adults. Care transitions
involve multiple players (e.g., patients, providers,
decision makers), cultures and settings, and are often
complicated by suboptimal communication and
coordination processes that create bottlenecks and result
in fragmented and unsafe care [85, 86]. More patient-
centred and integrated care approaches are needed to
help ensure that patients receive a seamless continuum
of services that respond to their changing needs as
they transition from hospital to home. The literature
on care transition interventions highlights a number of
9Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
challenges related to the outcomes measured, limited
descriptions of implementation processes, and a lack of
knowledge about the ways in which care setting contexts
influence intervention success [27]. This protocol
describes a mixed methods research approach that aims
to mitigate some of these challenges by triangulating
the knowledge generated from complementary
sources. Administrative healthcare use records, medical
chart review data, and provider and patient/caregiver
interviews will be used to describe the intervention
successes and challenges from different perspectives,
to identify intervention components that work and
that require change, and to examine implementation
processes (e.g., additional training) required to optimize
integrated care approaches. The lessons learned from
this research will help healthcare planners to further
adapt the SAC intervention, and will help to produce a
methodological roadmap for evaluating large-scale
interventions conducted in complex healthcare settings.
Our research team consists of seven multidisciplinary
health services researchers; five decision makers
representing the Manitoba Government, Winnipeg
Health Region and Shared Health; care providers; and
two patient representatives. Guided by the Canadian
Institutes for Health Research’s integrated knowledge
translation approach, this team will meet at the
beginning, mid-point, and end of each study phase
to interpret and contextualize findings, and to discuss
how the new knowledge should be used to help refine
subsequent research questions and methods. We have
created an Operations Committee consisting of SAC
program planners and providers, to help ensure that the
detailed research questions posed in each study phase
are appropriately contextualized and hold meaning to
program providers. As well, one of our team members
(GH) will work with community advisors to help ensure
that we effectively engage with these individuals. We
have also allotted funds to pursue various non-traditional
knowledge translation activities including creating an
infographic, writing op-eds and media releases, and
hosting a public discussion forum to reach a diverse
audience. An end-of-grant workshop will be held with
key decision makers from across Western Canada to
share learnings and to facilitate future research (e.g.,
prospective evaluation) endeavours.
The major strengths of this research include the rich
quantitative and qualitative data sources combined
with our synergistic analysis plan, and our integrated
knowledge translation activities involving local and
national partners. While Phases 1 and 2 of the research
will assess, in part, post-hospital discharge processes
(i.e., in Phase 1 we will determine the proportion of
SAC patients who received home care post hospital
discharge; in Phase 2 we will assess the extent to which
providers pre-emptively made these transitional care
plans), due to budget constraints (a) provider interviews
will be confided to those giving in-hospital care; and
(b) we have limited ourselves to conducting 30 patient
interviews, which is a small in comparisons in the total
number of patients that we expect in our cohort (~2340).
During these patient interviews, we will ask respondents
to provide suggestions to improve both in-hospital and
hospital-to-home transitions.
The proposed research will comprehensively evaluate an
intervention designed to improve hospital-to-home care
transitions for older adults. The lessons learned from
our evaluation approach have application for evaluating
additional interventions conducted in complex healthcare
SAC: sub-acute care; ED: emergency department;
EMS: emergency medical services; ICD: International
Classification of Disease; HIPC: Health Information
Privacy Committee.
The additional files for this article can be found as follows:
• Additional File 1. Sample Size Calculation. DOI:
• Additional File 2. Draft Capture Sheet for Subacute
Care Chart Audit. DOI:
Ethics approval was obtained from the University
of Manitoba Health Research Ethics Board
(HS23411-H2019:447) and the Winnipeg Health Region
(RAAC 2020-002). The Manitoba Government’s Health
Information Privacy Committee (HIPC) has also approved
the study to ensure that Manitoba resident privacy will
be protected during our analysis of healthcare utilization
data (HIPC No. 2019/2020 – 46). Consent from study
subjects in Phase 1 was not required as permitted under
section 24(3)c of the Personal Health Information Act
All data used for this phase are contained in the Population
Health Data Repository at the Manitoba Centre for
Health Policy, and undergoes a process of removing any
10Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
identifying information by Manitoba Health Seniors and
Active Living prior to being placed within the Repository.
All results will be presented publicly as aggregate data.
All study participants (providers, patients and informal
carers) in Phases 2 and 3 of the study will provide written
consent prior to participation.
We would like to acknowledge Krista Allan (Chief Nursing
Officer and Chief Health Operations Officer, Winnipeg
Regional Health Authority), Brenda Comte (Director,
Operational Analysis and Reporting, Winnipeg Regional
Health Authority), Jeanette Edwards (Community Health,
Quality and Learning, Shared Health) and Lanette
Siragusa (Health System Integration and Chief Nursing
Officer, Shared Health) for their strong contributions and
commitment to this project. We are grateful to Thomas
Beaudry and Lorie Deda, community partners on the study,
for providing critical context from their lived experiences in
the healthcare system. We also thank our administrative
data providers, Manitoba Health, Seniors and Active Living
and the Winnipeg Regional Health Authority.
Dr. rer. medic. Jörn Kiselev, Physiotherapist/MSc in
Physiotherapy, Senior Researcher/Project Management
Project Präp-Go, Department of Anesthesiology and
Operative Intensive Care Medicine, Campus Charité Mitte,
Berlin, Germany.
Dr. Julie MacInnes, Senior Research Fellow, Integrated
Care and ICAP Programme Lead, Centre for Health
Services Studies, University of Kent, Canterbury, UK.
This work is supported by an operating grant from
the Canadian Institutes of Health Research (funding
reference number 107897).
The authors have no competing interests to declare.
MBD conceived of the research idea, and transformed
it into a successful grant application with input from
TS, AK, DC, JEE, PB, GH and SK. This manuscript was first
drafted by JEE and MBD, additional content and draft
reviews were provided by TS, SK, and TB. All other authors
reviewed and approved the final version.
Dr. Malcolm B. Doupe
Manitoba Centre for Health Policy, Department of Community
Health Sciences, Rady Faculty of Health Sciences, University of
Manitoba, Winnipeg, Canada; Rady Faculty of Health Sciences,
University of Manitoba, Winnipeg, Canada
Dr. Jennifer E. Enns
Manitoba Centre for Health Policy, Department of Community
Health Sciences, Rady Faculty of Health Sciences, University of
Manitoba, Winnipeg, Canada
Dr. Sara Kreindler
George and Fay Yee Centre for Healthcare Innovation,
Department of Community Health Sciences, Rady Faculty of
Health Sciences, University of Manitoba, Winnipeg, Canada
Dr. Thekla Brunkert
Manitoba Centre for Health Policy, Department of Community
Health Sciences, Rady Faculty of Health Sciences, University of
Manitoba, Winnipeg, Canada
Dr. Dan Chateau
Manitoba Centre for Health Policy, Department of Community
Health Sciences, Rady Faculty of Health Sciences, University of
Manitoba, Winnipeg, Canada
Dr. Paul Beaudin
George and Fay Yee Centre for Healthcare Innovation,
Department of Community Health Sciences, Rady Faculty of
Health Sciences, University of Manitoba, Winnipeg, Canada
Dr. Gayle Halas
Rady Faculty of Health Sciences, University of Manitoba,
Winnipeg, Canada
Dr. Alan Katz
Manitoba Centre for Health Policy, Department of Community
Health Sciences, Rady Faculty of Health Sciences, University of
Manitoba, Winnipeg, Canada
Dr. Tara Stewart
George and Fay Yee Centre for Healthcare Innovation,
Department of Community Health Sciences, Rady Faculty of
Health Sciences, University of Manitoba, Winnipeg, Canada
1. United Nations. 2019 Revision of World Population
Prospects 2019. Available from:
2. Stephan BCM, Birdi R, Tang EYH, et al. Secular Trends
in Dementia Prevalence and Incidence Worldwide: A
Systematic Review. Journal of Alzheimer’s disease: JAD.
2018; 66(2): 653–80. DOI:
3. World Health Organization. Dementia: Fact sheet
2017. Available from:
4. 2020 Alzheimer’s disease facts and figures. Alzheimer’s
11Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
& dementia: the journal of the Alzheimer’s Association.
5. Chatterji S, Byles J, Cutler D, Seeman T, Verdes E. Health,
functioning, and disability in older adults-present status
and future implications. Lancet. 2015; 385(9967): 563-75.
6. Prince MJ, Wu F, Guo Y, et al. The burden of disease
in older people and implications for health policy and
practice. Lancet. 2015; 385(9967): 549–62. DOI: https://
7. Canadian Institutes of Health Information. Health Care
in Canada, 2011: A Focus on Seniors and Aging. https://
enpdf. 2011.
8. Edemekong PF, Bomgaars DL, Sukumaran S, Levy SB.
Activities of Daily Living (ADLs). StatPearls. Treasure Island,
FL: StatPearls Publishing; 2020.
9. National Council on Aging. Healthy Aging Facts. Available
10. Griffith LE, Gruneir A, Fisher K, et al. Insights on
multimorbidity and associated health service use and
costs from three population-based studies of older adults
in Ontario with diabetes, dementia and stroke. BMC Health
Serv Res. 2019; 19(1): 313. DOI:
11. Fisher K, Griffith L, Gruneir A, et al. Comorbidity and
its relationship with health service use and cost in
community-living older adults with diabetes: A population-
based study in Ontario, Canada. Diabetes Research and
Clinical Practice. 2016; 122: 113–23. DOI: https://doi.
12. Greenwald PW, Stern ME, Rosen T, Clark S, Flomenbaum
N. Trends in short-stay hospitalizations for older adults
from 1990 to 2010: Implications for geriatric emergency
care. The American Journal of Emergency Medicine.
2014; 32(4): 311–4. DOI:
13. Gruneir A, Griffith L, Fisher K, et al. Increasing
comorbidity and health services utilization in older adults
with prior stroke. Neurology. 2016; 87(20): 2091-. DOI:
14. Quach C, McArthur M, McGeer A, et al. Risk of infection
following a visit to the emergency department: A cohort
study. CMAJ. 2012; 184(4): E232–9. DOI: https://doi.
15. Gardiner JC, Reed PL, Bonner JD, H4aggerty DK, Hale
DG. Incidence of hospital-acquired pressure ulcers - a
population-based cohort study. International wound
journal. 2016; 13(5): 809–20. DOI:
16. Cole MG, Ciampi A, Belzile E, Zhong L. Persistent delirium
in older hospital patients: a systematic review of frequency
and prognosis. Age Ageing. 2009; 38(1): 19–26. DOI:
17. McCabe JJ, Kennelly SP. Acute care of older patients in
the emergency department: strategies to improve patient
outcomes. Open Access Emergency Medicine. 2015; 7:
45–54. DOI:
18. Hirschman KB, Paik HH, Pines JM, et al. Cognitive
impairment among older adults in the emergency
department. West J Emerg Med. 2011; 12(1): 56–62.
19. Tam VC, Knowles SR, Cornish PL, et al. Frequency, type
and clinical importance of medication history errors
at admission to hospital: a systematic review. Cmaj.
2005; 173(5): 510–5. DOI:
20. Armor BL, Wight AJ, Carter SM. Evaluation of Adverse
Drug Events and Medication Discrepancies in Transitions of
Care Between Hospital Discharge and Primary Care Follow-
Up. Journal of pharmacy practice. 2016; 29(2): 132–7. DOI:
21. Le Berre M, Maimon G, Sourial N, Guériton M, Vedel I.
Impact of Transitional Care Services for Chronically Ill
Older Patients: A Systematic Evidence Review. Journal of
the American Geriatrics Society. 2017; 65(7): 1597–608.
22. Zurlo A, Zuliani G. Management of care transition and
hospital discharge. Aging Clin Exp Res. 2018; 30(3): 263–70.
23. Moore C, McGinn T, Halm E. Tying up loose ends:
discharging patients with unresolved medical issues. Arch
Intern Med. 2007; 167(12): 1305–11. DOI: https://doi.
24. Naylor MD. Transitional care of older adults. Annual review
of nursing research. 2002; 20: 127–47. DOI: https://doi.
25. Sheikh F, Gathecha E, Bellantoni M, et al. A Call to Bridge
Across Silos during Care Transitions. Joint Commission
journal on quality and patient safety. 2018; 44(5): 270–8.
26. Li JY, Yong TY, Hakendorf P, Ben-Tovim D, Thompson
CH. Timeliness in discharge summary dissemination is
associated with patients’ clinical outcomes. Journal of
evaluation in clinical practice. 2013; 19(1): 76–9. DOI:
27. Rennke S, Nguyen OK, Shoeb MH, et al. Hospital-initiated
transitional care interventions as a patient safety strategy:
a systematic review. Ann Intern Med. 2013; 158(5 Pt 2):
433–40. DOI:
28. Coleman EA, Boult C. Improving the quality of transitional
care for persons with complex care needs. Journal of
the American Geriatrics Society. 2003; 51(4): 556–7. DOI:
29. Dawda P, Russell L. Sub-acute care An international
literature review 2014. Available from: https://rsph.anu.
30. Holland DE, Harris MR. Discharge planning, transitional
care, coordination of care, and continuity of care: clarifying
concepts and terms from the hospital perspective. Home
Health Care Serv Q. 2007; 26(4): 3-19. DOI: https://doi.
12Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
31. Manville M, Klein MC, Bainbridge L. Improved outcomes
for elderly patients who received care on a transitional
care unit. Can Fam Physician. 2014; 60(5): e263–71.
32. Kushner DS, Peters KM, Johnson-Greene D. Evaluating
Siebens Domain Management Model for inpatient
rehabilitation to increase functional independence and
discharge rate to home in geriatric patients. Archives of
physical medicine and rehabilitation. 2015; 96(7): 1310–8.
33. Del Giudice E, Ferretti E, Omiciuolo C, et al. The
hospital-based, post-acute geriatric evaluation and
management unit: the experience of the acute geriatric
unit in Trieste. Archives of gerontology and geriatrics.
2009; 49 Suppl 1: 49–60. DOI:
34. Young J, Green J, Forster A, et al. Postacute care for older
people in community hospitals: a multicenter randomized,
controlled trial. Journal of the American Geriatrics Society.
2007; 55(12): 1995–2002. DOI:
35. O’Reilly J, Lowson K, Green J, Young JB, Forster A.
Post-acute care for older people in community hospitals-
-a cost-effectiveness analysis within a multi-centre
randomised controlled trial. Age Ageing. 2008; 37(5):
513–20. DOI:
36. Hitch B, Parlier AB, Reed L, et al. Evaluation of a Team-
Based, Transition-of-Care Management Service on 30-Day
Readmission Rates. North Carolina medical journal. 2016;
77(2): 87–92. DOI:
37. Coleman EA, Parry C, Chalmers S, Min SJ. The care
transitions intervention: results of a randomized controlled
trial. Arch Intern Med. 2006; 166(17): 1822–8. DOI: https://
38. Robinson TE, Zhou L, Kerse N, et al. Evaluation of a New
Zealand program to improve transition of care for older
high risk adults. Australasian journal on ageing. 2015;
34(4): 269–74. DOI:
39. Naylor MD, Bowles KH, McCauley KM, et al. High-value
transitional care: translation of research into practice.
Journal of evaluation in clinical practice. 2013; 19(5):
40. Buurman BM, Parlevliet JL, Allore HG, et al.
Comprehensive Geriatric Assessment and Transitional
Care in Acutely Hospitalized Patients: The Transitional Care
Bridge Randomized Clinical Trial. JAMA internal medicine.
2016; 176(3): 302–9. DOI:
41. Heim N, Rolden H, van Fenema EM, et al. The
development, implementation and evaluation of a
transitional care programme to improve outcomes of frail
older patients after hospitalisation. Age Ageing. 2016;
45(5): 643–51. DOI:
42. Wang YC, Chou M-Y, Liang C-k, et al. Post-Acute Care as
a Key Component in a Healthcare System for Older Adults.
Annals of Geriatric Medicine and Research. 2019; 23. DOI:
43. Gledhill K, Hanna L, Nicks R, Lannin NA. Defining
discharge-readiness from subacute care from all
stakeholders’ perspectives: a systematic review. Disability
and rehabilitation. 2020: 1-8. DOI:
44. Kim H, Thyer BA. Does transitional care prevent older
adults from rehospitalization? A review. Journal of
evidence-informed social work. 2015; 12(3): 261–71. DOI:
45. Davis J, Morgans A, Stewart J. Developing an Australian
health and aged care research agenda: a systematic
review of evidence at the subacute interface. Aust Health
Rev. 2016; 40(4): 420–7. DOI:
46. Wee SL, Vrijhoef HJ. A conceptual framework for
evaluating the conceptualization, implementation and
performance of transitional care programmes. Journal
of evaluation in clinical practice. 2015; 21(2): 221–8. DOI:
47. Coleman EA. Falling through the cracks: challenges and
opportunities for improving transitional care for persons
with continuous complex care needs. Journal of the
American Geriatrics Society. 2003; 51(4): 549–55. DOI:
48. Naylor MD, Sochalski JA. Scaling up: bringing the
transitional care model into the mainstream. Issue brief
(Commonwealth Fund). 2010; 103: 1–12.
49. World Health Organization. WHO global strategy on
people-centred and integrated health services: Interim
report Geneva 2015. Available from:
50. Leutz W. Five laws for integrating medical and social
services: lessons from the United States and the United
Kingdom. Milbank Quarterly. 1999; 77(1). DOI: https://doi.
51. Nolte E, Pitchforth E. What is the evidence on the
economic impacts of integrated care? Copenhagen:
European Observatory on Health Systems and Policies;
52. National Voices. Think Local Act Personal. A narrative for
person-centred coordinated care 2013. Available from:
53. Proctor E, Silmere H, Raghavan R, et al. Outcomes
for implementation research: conceptual distinctions,
measurement challenges, and research agenda.
Administration and policy in mental health. 2011; 38(2):
65–76. DOI:
54. Institute of Medicine Committee on Crossing the Quality
Chasm: Adaptation to Mental H, Addictive D. The National
Academies Collection: Reports funded by National
Institutes of Health. Improving the Quality of Health Care
for Mental and Substance-Use Conditions: Quality Chasm
Series. Washington, DC: National Academies Press (US);
13Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
55. Kreindler SA. Six ways not to improve patient flow: a
qualitative study. BMJ Quality &amp; Safety. 2017; 26:
56. Madore O. The Canada Health Act: Overview and Options.
Parliament of Canada; 2012.
57. Manitoba Health. Annual Statistics 2018-2019 Winnipeg:
Manitoba Health; 2019. Available from:
58. Canadian Institute for Health Information. NACRS
Emergency Department Visits and Length of Stay:
Quickstats Ottawa, Ontario2018. Available from: https://
59. Canadian Institute for Health Information. Inpatient
Hospitalization, Surgeries, Newborns and Childbirth
Indicators. Quickstats 2015-2016 Ottawa, Ontario2017.
Available from:
60. Sivananthan S, Doupe M, McGregor M. Exploring the
Ecology of Canada’s Publicly Funded Residential Long-Term
Care Bed Supply. Canadian Journal on Aging. 2015; 34(1):
60–74. DOI:
61. Wait Times Reduction Task Force. Wait Times Reduction
Task Force: Final Report Winnipeg, Manitoba2017.
Available from:
62. Peachey D, Tait N, Adams O, Croson W. Provincial Clinical
and Preventive Services Planning for Manitoba: Doing
Things Differently and Better Halibut Bay, Nova Scotia2017
[cited 2020 February 20]. Available from:
63. KPMG. Health System Sustainability & Innovation Review:
Phase 1 Report Winnipeg, Manitoba2017. Available from:
64. Winnipeg Health Region. Healing our Health System.
Winnipeg, Manitoba2019. Available from: http://
65. WInnipeg Health Region. Priority Home 2017. Available
66. Winnipeg Health Region. Rapid Response Nursing 2017.
Available from:
67. Creswell JW, Plano Clark VL. Designing and Conducting
Mixed Methods Research. Third Edition ed: SAGE; 2018.
68. Roos LL, Gupta S, Soodeen RA, Jebamani L. Data
quality in an information-rich environment: Canada as
an example. CanJ Aging. 2005; 24 Suppl 1: 153–70. DOI:
69. Doupe M, Finlayson G, Khan S, et al. Supportive Housing
for Seniors: Reform Implications for Manitoba’s Older Adult
Continuum of Care Winnipeg, MB: Manitoba Centre for
Health Policy; 2016.
70. Doupe MB, Day S, McGregor MJ, et al. Pressure Ulcers
Among Newly Admitted Nursing Home Residents:
Measuring the Impact of Transferring From Hospital. Med
Care. 2016; 54(6): 584–91. DOI:
71. Doupe MB, Palatnick W, Day S, et al. Frequent users
of emergency departments: developing standard
definitions and defining prominent risk factors. Annals of
Emergency Medicine. 2012; Jul(1): 24–32. DOI: https://doi.
72. Weeks LE, Macdonald M, Martin-Misener R, et al. The
impact of transitional care programs on health services
utilization in community-dwelling older adults: a
systematic review. JBI database of systematic reviews and
implementation reports. 2018; 16(2): 345–84. DOI: https://
73. Lix L, Yogendran M, Burchill C, et al. Defining and
Validating Chronic Diseases: An Administrative Data
Approach. http://mchp-appservcpeumanitobaca/reference/
chronicdiseasepdf. 2006.
74. Canadian Institute of Health Information. Acute
and Ambulatory Care Data Content Standard 2017.
Available from:
75. Canadian Institute of Health Information. A Guide to
Intervention Code Assignment 2018. Available from:
76. Schneeweiss S, Rassen JA, Glynn RJ, et al. High-
dimensional propensity score adjustment in studies
of treatment effects using health care claims data.
Epidemiology (Cambridge, Mass). 2009; 20(4): 512–22. DOI:
77. Schneeweiss S. Automated data-adaptive analytics for
electronic healthcare data to study causal treatment
effects. Clinical epidemiology. 2018; 10: 771–88. DOI:
78. Austin PC. An Introduction to Propensity Score Methods
for Reducing the Effects of Confounding in Observational
Studies. Multivariate Behav Res. 2011; 46(3): 399–424. DOI:
79. Diaz M, Neuhauser D. Pasteur and parachutes: when
statistical process control is better than a randomized
controlled trial. Qual Saf Health Care. 2005; 14(2): 140–3.
80. Western Electric Company. Statistical quality
control handbook 1958. Available from: https://www.
81. Gregory BH, Van Horn C, Kaprielian VS. 8 steps to a chart
audit for quality. Family practice management. 2008;
15(7): A3–8.
82. Cohen S, Syme L. Issues in the Study and Application of
Social Support. In: Cohen S, Syme L (eds.), Social Support
and Health. 3–22. New York: Academic Press; 1985.
83. Koppitz AL, Dreizler J, Altherr J, et al. Relocation
14Doupe et al. International Journal of Integrated Care DOI: 10.5334/ijic.5953
Doupe MB, Enns JE, Kreindler S, Brunkert T, Chateau D, Beaudin P, Halas G, Katz A, Stewart T. Improving In-Hospital Care For Older
Adults: A Mixed Methods Study Protocol to Evaluate a System-Wide Sub-Acute Care Intervention in Canada. International Journal of
Integrated Care, 2022; 22(1): 25, 1–14. DOI:
Submitted: 06 April 2021 Accepted: 16 March 2022 Published: 28 March 2022
© 2022 The Author(s). This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0
International License (CC-BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original
author and source are credited. See
International Journal of Integrated Care is a peer-reviewed open access journal published by Ubiquity Press.
experiences with unplanned admission to a nursing
home: a qualitative study. Int Psychogeriatr.
2017; 29(3): 517–27. DOI:
84. Rossman GB, Wilson BL. Numbers and words:
Combining quantitative and qualitative methods
in a single large-scale evaluation study. Evaluation
Review. 1985; 9: 627–43. DOI: https://doi.
85. Aase K, Waring J. Crossing boundaries: Establishing a
framework for researching quality and safety in care
transitions. Appl Ergon. 2020; 89: 103228. DOI: https://doi.
86. Aase K, Schibevaag L, Waring J. Crossing boundaries:
quality in care transition. In: Aase K, Waring J, Schibevaag
L (eds.), Researching Quality in Care Transitions:
International Perspectives. 3–29. Cham: Palgrave
MacMillan; 2017. DOI:
Background Older adults frequently experience deconditioning following acute illnesses and require discharge from acute-care facilities to post-acute care facilities, which are limited. Our study aimed to explore predictors and outcomes associated with elongated length of stay (LOS) among older adults awaiting discharge to skilled nursing facility (SNF).Methods Retrospective cohort study was conducted at Shamir Medical Center, Israel, among adults (> 65 years) eligible for SNF. ROC curve analysis was used to determine prolonged LOS based on the risk to fall. Logistic and Cox regressions were used to analyze predictors and outcomes.ResultsAmong 659 older adults awaiting transfer to SNF, 127 patients (24% among survivors of the index hospitalization) had prolonged LOS (> 12 days). The median age of patients was 82 years and 51% were females. The independent predictors for prolonged LOS were lower Norton index, higher MUST score, and admission from home. Prolonged LOS was independently associated with hospital-acquired infections, device related infections, and acquisitions of multidrug-resistant organisms.Conclusion Prolonged LOS among older adults, awaiting transfer to SNF, should be suspected among non-institutionalized older adults with lower nutritional status and higher risk of pressure ulcers. The burden associated with establishing additional SNF beds, must be outweighed vs. the substantial infectious complications among awaiting older adults.
Full-text available
Background Despite the breadth and diversity of research and policies on care transitions, research studies often report similar components that affect the quality and safety of care, including communication across professional groups and care settings, transfer of information, coordination of resources or training of healthcare personnel. In this article, we aim to deepen our understanding of care transitions by proposing a heuristic research framework that takes into account the components and factors influencing the quality and safety of care transitions in diverse settings. Methodology Using a pragmatic qualitative narrative meta-synthesis of empirically grounded research studies (N = 13) involving 31 researchers from seven countries (Australia, Canada, Denmark, Germany, the Netherlands, Norway and the UK), we conducted a thematic analysis to identify the components analysed in the included studies. We then used these components to create a framework for researching care transitions. Results Our narrative synthesis found that the quality and safety of care transitions are influenced by a range of patient-centred, communicative, collaborative, cultural, competency-based, accountability-based and spatial components. These components are encompassed within a broader set of dimensions that require careful consideration: (1) the conceptualising of the care transition notion, (2) the methodology for researching care transitions, (3) the role of patients and carers in care transitions, (4) the complexity surrounding care transitions, (5) the boundaries intertwined in care transitions and (6) care transition improvement interventions. These six dimensions constitute an analytical framework for planning and conducting research on care transitions in diverse settings. Conclusion The proposed six-dimensional framework for researching quality and safety in care transitions offers a roadmap for future practice and policy interventions and provides a starting point for planning and designing future research.
Full-text available
Background Most studies that examine comorbidity and its impact on health service utilization focus on a single index-condition and are published in disease-specific journals, which limit opportunities to identify patterns across conditions/disciplines. These comparisons are further complicated by the impact of using different study designs, multimorbidity definitions and data sources. The aim of this paper is to share insights on multimorbidity and associated health services use and costs by reflecting on the common patterns across 3 parallel studies in distinct disease cohorts (diabetes, dementia, and stroke) that used the same study design and were conducted in the same health jurisdiction over the same time period. Methods We present findings that lend to broader Insights regarding multimorbidity based on the relationship between comorbidity and health service use and costs seen across three distinct disease cohorts. These cohorts were originally created using multiple linked administrative databases to identify community-dwelling residents of Ontario, Canada with one of diabetes, dementia, or stroke in 2008 and each was followed for health service use and associated costs. Results We identified 376,434 indviduals wtih diabetes, 95,399 wtih dementia, and 29,671 with stroke. Four broad insights were identified from considering the similarity in comorbidity, utilization and cost patterns across the three cohorts: 1) the most prevalent comorbidity types were hypertension and arthritis, which accounted for over 75% of comorbidity in each cohort; 2) overall utilization increased consistently with the number of comorbidities, with the vast majority of services attributed to comorbidity rather than the index conditions; 3) the biggest driver of costs for those with lower levels of comorbidity was community-based care, e.g., home care, GP visits, but at higher levels of comorbidity the driver was acute care services; 4) service-specific comorbidity and age patterns were consistent across the three cohorts. Conclusions Despite the differences in population demographics and prevalence of the three index conditions, there are common patterns with respect to comorbidity, utilization, and costs. These common patterns may illustrate underlying needs of people with multimorbidity that are often obscured in literature that is still single disease-focused. Electronic supplementary material The online version of this article (10.1186/s12913-019-4149-3) contains supplementary material, which is available to authorized users.
Full-text available
Older adults often experience functional decline following acute medical care. This functional decline may lead to permanent disability, which will increase the burden on the medical and long-term care systems, families, and society as a whole. Post-acute care aims to promote the functional recovery of older adults, prevent unnecessary hospital readmission, and avoid premature admission to a long-term care facility. Research has shown that post-acute care is a cost-effective service model, with both the hospital-at-home and community hospital post-acute care models being highly effective. This paper describes the post-acute care models of the United States and the United Kingdom and uses the example of Taiwan’s highly effective post-acute care system to explain the benefits and importance of post-acute care. In the face of rapid demographic aging and smaller household size, a post-acute care system can lower medical costs and improve the health of older adults after hospitalization.
Full-text available
Background Decision makers in health care increasingly rely on nonrandomized database analyses to assess the effectiveness, safety, and value of medical products. Health care data scientists use data-adaptive approaches that automatically optimize confounding control to study causal treatment effects. This article summarizes relevant experiences and extensions. Methods The literature was reviewed on the uses of high-dimensional propensity score (HDPS) and related approaches for health care database analyses, including methodological articles on their performance and improvement. Articles were grouped into applications, comparative performance studies, and statistical simulation experiments. Results The HDPS algorithm has been referenced frequently with a variety of clinical applications and data sources from around the world. The appeal of HDPS for database research rests in 1) its superior performance in situations of unobserved confounding through proxy adjustment, 2) its predictable efficiency in extracting confounding information from a given data source, 3) its ability to automate estimation of causal treatment effects to the extent achievable in a given data source, and 4) its independence of data source and coding system. Extensions of the HDPS approach have focused on improving variable selection when exposure is sparse, using free text information and time-varying confounding adjustment. Conclusion Semiautomated and optimized confounding adjustment in health care database analyses has proven successful across a wide range of settings. Machine-learning extensions further automate its use in estimating causal treatment effects across a range of data scenarios.
Full-text available
Current demographic and epidemiological trends highlight a growing task in surgical departments by elderly patients, characterized by high prevalence of comorbidity, complexity, and functional disability. Of consequence, discharge of an elderly patient must be considered in a new cultural perspective and should be imagined as a well-structured process starting from admission to surgical department and finishing with the patient discharge in a setting able to support her/him in the best possible way. The lack of a suitable discharge planning and of a proper transition program in the elderly subjects increases the risk of quick re-admission and may negatively affect the functional and the status quality of life of patients and caregivers. To reduce the risk of negative outcome it is essential a hospital organization dedicated to the discharge of frail older patients considering: (1) adequate attention to assess the comprehensive clinical/social/care conditions; (2) respect of the expectations of the patient and her/his relatives; (3) formalization of institutional roles or teams designated to the planning and coordination of discharge; (4) good knowledge of management programs of transitional care, and (5) strong communication/information ability in patients transition between hospital, home care and community settings.
Purpose: Research to date has focused on clinicians’ views on patients’ discharge readiness from acute hospital settings.This study aims to synthesise the literature on discharge readiness from sub-acute (rehabilitation) hospital settings from all stakeholders’ perspectives. Methods: Electronic databases (MEDLINE, CINAHL, Ageline, AMED and Global Health) were systematically searched for post-2000 publications on discharge readiness of adult inpatients in sub-acute settings. After screening, quantitative and qualitative studies were assessed for bias using the Downs and Black checklist and McMaster critical assessment tool respectively, and narrative analysis conducted. Results: From the 3516 papers identified, 23 were included in the review. Overall quality of articles was rated as adequate. Narrative synthesis identified three main themes: the importance of functional outcomes; confounding factors impact on discharge destination and length of stay and barriers and facilitators to discharge. Conclusion: Despite limited literature defining sub-acute patients’ discharge-readiness from all stakeholders’ perspectives, synthesis of available findings identified major themes for consideration when determining when a patient is ready to leave hospital. Limitations include the heterogeneity of the studies located impacted on data extraction and quality appraisal. • IMPLICATIONS FOR REHABILITATION • Discharging patients from hospital is complex, discharge too early may lead to poor medical outcomes or readmission, while discharge too late may increase the risk of hospital-based adverse events. • Multiple factors need to be considered when considering the discharge readiness of an inpatient. • Ensuring adequate social support is key to maximising transition from hospital to home. • Combining the use of functional outcome measures with clinical decision-making allows for quantifying readiness for discharge.
Background: Time trends for dementia prevalence and incidence rates have been reported over the past seven decades in different countries and some have reported a decline. Objective: To undertake a systematic review to critically appraise and provide an evidence-based summary of the magnitude and direction of the global changes in dementia prevalence and incidence across time. Methods: Medline, EMBASE, and PsychINFO were searched for studies focused on secular trends in dementia prevalence and/or incidence until 18 December 2017. In total, 10,992 articles were identified and 43 retained. Results: Overall, prevalence rates are largely increasing (evidence primarily from record-based surveys and cohort studies in Japan, Canada, and France) or have remained stable (evidence primarily from cohort studies in Sweden, Spain and China). A significant decline in prevalence has however been reported in more recent studies (i.e., from 2010 onwards) from Europe (e.g., UK and Sweden) and the USA. Incidence rates have generally remained stable or decreased in China, Canada, France, Germany, Denmark, Sweden, the Netherlands, UK, and USA. An increase has only been reported in five countries: Italy, Japan, Wales, Germany, and the Netherlands. Only one study reported findings (stability in incidence) from a low and middle-income country using data from Nigeria. Conclusions: The evidence on secular trends in the prevalence and incidence of dementia is mixed including contradictory findings using different (and in some cases the same) datasets in some countries (e.g., the USA, UK, and Sweden). This making it difficult to draw concrete conclusions. However, declining trends recently observed in some high-income Western countries in the most recent two decades including the UK, USA, and Sweden are encouraging. Updated dementia prevalence and incidence estimates will inform public health and financial planning as well as development of prevention strategies.
Background: Older adults with complex medical conditions are vulnerable during care transitions. Poor care transitions can lead to poor patient outcomes and frequent readmissions to the hospital. Factors contributing to suboptimal care transitions: Key factors related to ineffective care transitions, which can lead to suboptimal patient outcomes, include poor cross-site communication and collaboration; lack of awareness of patient wishes, abilities, and goals of care; and incomplete medication reconciliation. Fundamental elements for effective care transitions put forth by The Joint Commission for effective care transitions include interdisciplinary coordination and collaboration of patient care in care transitions, shared accountability by all clinicians involved in care transitions, and provision of appropriate support and follow-up after discharge. Review of four existing models of care transitions: Consideration of four existing care transitions models representing different health care settings-Care Transitions Intervention® Guided Care, Interventions to Reduce Acute Care Transfers (INTERACT®), Home Health Model of Care Transitions-revealed that they are important but limited in their impact on transitions across health care settings. Proposal of the integrated care transitions approach: An innovative approach, Integrated Care Transitions Approach (ICTA), is proposed that incorporates the best practices of the four models discussed in this article and factors identified as essential for an effective care transition while addressing limitations of existing transitional care models. ICTA's four key characteristics and seven key elements are unique and stem from factors that help achieve effective care transitions.
Objective: The objective was to identify and synthesize the best available evidence on the impact of transitional care programs on various forms of health services utilization in community-dwelling older adults. Introduction: There is growing evidence that transitional care programs can help address important challenges facing health care systems and our increasing older adult population in many countries by reducing unnecessary health service utilization. There is a need for a systematic review of the research evaluating the impact of transitional care programs on hospital and other health service usage. Inclusion criteria: The review included studies on community-dwelling adults age 60 and over with at least one medical diagnosis, and which evaluated the outcomes of transitional care programs on health system utilization of older adults. The outcomes for this review were hospital usage including admissions and readmissions, emergency department usage, primary care/physician usage, nursing home usage, and home health care usage. The review considered experimental and epidemiological study designs including randomized controlled trials, non-randomized controlled trials, quasi-experimental studies, before and after studies, prospective and retrospective cohort studies, and case-control studies. Methods: A three-step search was utilized to find published and unpublished studies conducted in any country but reported in English. Six electronic databases were searched from inception of the database to May, 2016. A search for unpublished studies was also conducted. Methodological quality was assessed independently by two reviewers using the Joanna Briggs Institute critical appraisal checklist for systematic reviews and research synthesis. Quantitative data were extracted from included studies independently by the two reviewers using the standardized Joanna Briggs Institute data extraction tools. Due to the methodological heterogeneity of the included studies, a comprehensive meta-analysis for all outcomes was not possible. Meta-analysis was conducted for rehospitalization at 30, 90 and 180 days. A narrative summary of other quantitative findings was conducted. Results: Twenty-three studies met the inclusion criteria and were included in the review. Nineteen of the studies were randomized controlled trials and four were case control studies, involving 20,997 participants in total with a mean age of 76. Meta-analysis found that transitional care significantly reduced hospital readmission rates at 30 days (odds ratio [OR] 0.75, 95% confidence intervals [CIs] 0.62-0.91, p < 0.01), 90 days (OR 0.77, 95% CIs 0.59-1.02, p = 0.04), and 180 days (OR 0.67, 95% CIs 0.46-0.99, p < 0.01). Narrative synthesis indicated little impact of transitional care on emergency department and nursing home usage, increased use of primary care/physician usage, and decreased home health care usage. Conclusions: Based on a review of 23 studies conducted in the USA, Hong Kong, Canada, Germany, the Netherlands, Sweden and Switzerland, we identified four major conclusions. First, transitional care reduces rehospitalization rates over time, with the largest effects seen at 30 days. Second, transitional care may increase the utilization of primary care services and thus have a favourable impact on preventative care. Third, transitional care may reduce home health usage. Fourth, transitional care interventions of one month or less appear to be as effective as longer interventions in reducing hospital usage.
The organisation and accomplishment of high-quality care transitions relies upon the coordination of multiple professionals, working within and across multiple care processes, settings and organisations, each with their own distinct ways of working, profile of resources and modes of organising. In short, care transitions might easily be regarded as complex activities that take place within complex systems, which can make accomplishing high-quality care challenging. In its broadest sense, this collection is concerned with the complexities of achieving quality in care transitions.