Sustained Impact of Electronic Alerts on Rate of
Prophylaxis Against Venous Thromboembolism
Patrick Emanuel Beeler1, Nils Kucher2, Jürg Blaser1
1 Research Center for Medical Informatics, Research and Education, University Hospital
2 Venous Thromboembolism Consult Service, Clinics for Angiology & Cardiology,
University Hospital Bern, Switzerland
Keywords (MeSH Terms):
- Clinical Decision Support Systems
- Computerized Medical Records Systems
- Venous Thromboembolism/prevention & control
- Reminder Systems
Running title: Sustained Impact of eAlerts
Correspondence: Prof. Dr. Jürg Blaser
Direktion Forschung und Lehre
Sonneggstr. 6, D5
Phone. +41 44 255 3618
Fax: +41 44 634 5503
Advanced electronic alerts (eAlerts) and computerized physician order entry
(CPOE) increase adequate thromboprophylaxis orders among hospitalized medical
patients. It remains unclear whether eAlerts maintain their efficacy over time, after
withdrawal of continuing medical education (CME) on eAlerts and on thromboprophylaxis
indications from the study staff.
We analyzed 5317 hospital cases from the University Hospital Zurich during 2006-
2009: 1854 cases from a medical ward with eAlerts (intervention group) and 3463 cases
from a surgical ward without eAlerts (control group). In the intervention group, an eAlert
with hospital-specific VTE prevention guidelines was issued in the electronic patient chart
6 h after admission if no pharmacological or mechanical thromboprophylaxis had been
ordered. Data were analyzed for three phases: pre-implementation (phase 1), eAlert
implementation with CME (phase 2), and post-implementation without CME (phase 3).
The rates of thromboprophylaxis in the intervention group were 43.4% in phase 1
and 66.7% in phase 2 (p<0.001), and increased further to 73.6% in phase 3 (p=0.011).
Early thromboprophylaxis orders within 12 hours after admission were more often placed
in phase 2 and 3 as compared to phase 1 (67.1% vs. 52.1%, p<0.001). In the surgical
control group, the thromboprophylaxis rates in the three phases were 88.6%, 90.7%,
Advanced eAlerts may provide sustained efficacy over time, with stable rates of
thromboprophylaxis orders among hospitalized medical patients.
Deep vein thrombosis (DVT) and pulmonary embolism (PE) together are referred
to as venous thromboembolism (VTE) (1). The DVT attack rate (incidence of first and
recurrent DVT) has been estimated for the United States and the European Union at
128/100,000 person-years (2) and 148/100,000 person-years (3), respectively.
Approximately one quarter of all VTE events are hospital-related and at least 70% of the
VTE-associated deaths occur as a consequence of a hospital-acquired VTE (3, 4). The
case fatality rate of PE approximates 7-11% (5, 6).
Appropriate use of prophylaxis to prevent VTE is an important strategy for
improving safety among hospitalized patients (7, 8). A recent meta-analysis found
significant reductions in any PE (relative risk 0.43) and fatal PE (relative risk 0.38) due to
thromboprophylaxis without a significant increase in major bleeding complications (9).
The American College of Chest Physicians (ACCP) has established evidence-
based guidelines for the prevention of VTE, including pharmacologic treatment with
anticoagulants and mechanical prophylaxis (e.g. compression stockings) (10).
Nevertheless, many patients at risk do not receive prophylaxis, as evidenced by three
large studies (11-13). The ACCP suggests the use of various strategies for increasing
thromboprophylaxis adherence, explicitly including the use of computer decision support
systems (CDSS). Some studies investigated computerized reminders of prophylaxis and
found substantially higher thromboprophylaxis rates due to electronic alerts (eAlert) (14-
16). However, lack of physician compliance with electronically generated suggestions is a
well-known problem called ‘alert fatigue’ or ‘overriding’ (17).
Research on VTE prevention has been a focus at the University Hospital Zurich in
recent years in the context of an ongoing quality assurance program which also covered
multicenter studies (18, 19). In 2007, an advanced eAlert module was developed and
implemented into the clinical information system (KISIM, Cistec AG, Zurich, Switzerland)
of a medical ward. An eAlert was issued in the electronic chart six hours after the patient
was admitted to the ward if no pharmacologic or mechanical prophylaxis had been
ordered. All healthcare professionals involved in the patient care were reminded by a
highlighted alert button to consider thromboprophylaxis, but only the physician was
enabled to respond to the eAlert. When the physician clicked on the eAlert button, an
information screen displayed hospital-specific guidelines on VTE prevention in medical
Physician compliance with the VTE eAlerts was evaluated in a pilot study (16).
During this period, continuing medical education (CME) on prevention of VTE in
hospitalized patients and on the eAlert system was provided to the staff physicians of the
medical ward. The rate of appropriate prophylaxis during this pilot study averaged 76%,
as compared to a much lower rate of 44% observed in a cross-sectional 1-day survey
prior to the implementation of computerized physician order entry (CPOE) (16, 18). The
prospective pilot study focused on physicians’ compliance. It did not consider data on the
prophylaxis rate in patients treated in the same medical ward immediately before or after
the study period, and no data were available on patients treated in other wards during the
identical period, in the absence of eAlerts.
The purpose of the present analysis was to comparatively assess the sustainability
of the eAlert system over time.
Materials and methods
Two wards of the University Hospital of Zurich, Switzerland, introduced CPOE in
2006, i.e. a medical ward (intervention group) and a surgical ward (control group). Then,
CPOE was limited to these wards, including 22 beds in medicine and 23 beds in surgery,
respectively. The remaining wards in medicine (164 beds) and surgery (239 beds) were
not considered since they introduced CPOE not before 2009.
In the medical intervention ward, but not in the surgical ward, an eAlert was
displayed in the electronic chart of each patient who did not receive a pharmacological or
mechanical thromboprophylaxis order within six hours after admission (16). Prophylaxis
guidelines including a VTE risk score are displayed by default when the physician
acknowledges the eAlert button.
All patients admitted to the study wards from April 2006 to December 2009 were
enrolled if they stayed for at least 24 hours in these wards and (i) either entered these
wards directly or (ii) entered these wards via the emergency department; i.e. in-patients
transferred from other wards were excluded. The electronic charts of the included patients
were searched for orders of pharmacological and mechanical prophylaxis. All orders of
prophylaxis were considered for the time period from admission to discharge or transfer of
the patient to another ward.
We analyzed 5317 hospital cases during 2006-2009: 1854 cases from the medical
intervention ward with eAlerts and 3463 cases from the surgical ward without eAlerts. The
overall study period included 45 months: (a) pre-implementation phase 1, 4/2006-8/2007,
(b) eAlerts implementation phase 2, 9/2007-12/2008, and (c) post-implementation
phase 3, 1/2009-12/2009. The eAlert module was implemented in the intervention ward at
the beginning of phase 2 and included monitoring of the appropriateness of the ordered
VTE prophylaxes (16). The eAlerts remained in operation in the intervention group also
during phase 3. Only during phase 2 was specific CME on indications for prophylaxis
provided to the physicians of the intervention group. No specific CME was offered to the
physicians of the control group. This quality improvement initiative study was approved by
the local ethics committee and patient consent was waived.
The rate of prophylaxis was defined as the percentage of patients with at least one
order for VTE prophylaxis during their stay in the admitting ward. Combined prophylaxis
was defined as orders of both pharmacologic and mechanical prophylaxes.
Clinical outcome at discharge was assessed for all three phases by searching the
patient charts for diagnoses of PE and DVT (I26.0, I26.9, I82.8, I82.9; International
Classification of Diseases [ICD-10], World Health Organization, Geneva, Switzerland). All
VTE diagnoses were independently adjudicated by two physicians, classifying them into
pre-existing events and hospital-acquired complications. Hospital-acquired VTE events
were considered if they were symptomatic and objectively confirmed by an imaging test.
No imaging confirmation was required for in-hospital deaths where PE was considered as
Chi-square tests of independence were used for statistical analysis of frequencies
in 3x2 contingency tables and Fisher’s exact tests were used for 2x2 contingency tables,
using the software EpiData V18.104.22.168 (EpiData Association, Odense, Denmark). The
p-levels of ≤0.05 were considered significant.
A total of 5,317 cases with a minimal stay of 24 hours in the wards of admission
were analyzed, including 4,156 patients. The length of stay in the admitting wards
averaged 5.0 days, the median was 3.0 days (medical ward: average 7.4 days, median
5.0 days, surgical ward: average 3.7 days, median 3.0 days). Thromboprophylaxis was
prescribed in 4,232 cases during hospitalization in the two wards (79.6%). These patients
received either pharmacological thromboprophylaxis alone (2,464 of 4,232 cases; 58.2%),
a combination of pharmacological and mechanical measures (39.1%), or mechanical
prophylaxis alone (2.7%).
The rate of thromboprophylaxis increased significantly after implementing the
eAlert function in the clinical information system of the intervention group, from 43.4%
during phase 1 to 66.7% during phase 2 (p<0.001; Tab. 1). The percentage further
increased during phase 3 to 73.6% (p=0.011). No significant changes in the rate of
prophylaxis were noted over the three phases in the surgical control group (p=0.16).
During the three study phases, the number of patients with a hospital-acquired
VTE event in the medical ward was 2 (pre-implementation), 0 (eAlerts implementation)
and 3 (post-implementation); in the surgical ward 3, 2, 1, respectively. Analysis of the data
of each group revealed no statistically significant changes between the study phases.
Timing of order entry
Orders of VTE prophylaxis were frequently placed before planned admissions to
the surgical ward, i.e. in patients hospitalized for elective procedures (1,898 of
3,113 orders, 61.0%). In contrast, very few pre-admission orders were issued by
physicians of the medical ward (19 of 1,119 orders, 1.7%). Of the 2,315 orders placed
after admission, 51.8% were entered within the first 6 h, 23.7% within 6-12 h, and 24.5%
at a later stage (average time of ordering after admission 15.4 h, median 5.7 h).
After eAlert implementation, the physicians of the intervention group entered
orders for thromboprophylaxis sooner than they did in the pre-implementation phase. The
percentage of orders placed within 12h after admission of the patients increased from
52.1% (146 of 280) to 67.1% (550 of 820; p<0.001) (Fig. 1). In particular, a higher
percentage of orders were placed in the time frame 6-12 h after admission of the patients,
i.e. during the period immediately following the eAlert activation (27.1% vs.
35.9%; p=0.008). In contrast, no increase of orders placed 6-12 h after admission was
observed in the control group.
CDSS are successful in supporting preventive care (20) and offer great potential to
assist physicians in improving the quality of treatment (21, 22). In most evaluations, the
impact of interventions for quality assurance is assessed by comparing data observed
during a study period with data of historic or simultaneous controls. Some studies
documented a beneficial effect of eAlerts on thromboprophylaxis over a limited period (14-
16). One study considered whether the impact of these eAlerts on the rate of prophylaxis
could be maintained beyond the initial study period, by comparing the long-term data with
historic controls (23). The present study confirms a sustained effect of the eAlert system
observed during the eAlert implementation phase by comparing the data to both the pre-
implementation phase and the post-implementation phase.
The VTE prophylaxis rate in the medical intervention ward remained high during
the post-implementation phase despite cessation of specific CME and withdrawal of
accompanying study staff. The stability in the rates of thromboprophylaxis orders,
particularly in the rate of early orders, confirm that the physician compliance with the
eAlert system was sufficient over time. Only in the intervention group, but not in the
control group, did the percentage of orders within 6-12 h increase, i.e. immediately after
Although eAlerts for hospitalized medical patients can not and should not produce
thromboprophylaxis rates found in surgical patients, they improve adherence to guidelines
and may provide sustained efficacy without evidence of alert fatigue (17) or the
Hawthorne effect (24, 25), sometimes affecting the generalizability of clinical research.
The rates of thromboprophylaxis obtained for the identical time periods from the surgical
control ward were stable over time, with an average of 90%, confirming that
thromboprophylaxis rates are higher in surgical than in medical patients (13, 26). The
appropriateness of thromboprophylaxis was not assessed in the surgical patients, but it is
unlikely that the introduction of eAlerts on this surgical ward would further improve the use
of appropriate prophylaxis.
The present study focused on the rates of thromboprophylaxis over time,
considering neither the appropriateness of ordered prophylaxis nor the appropriateness of
omitted prophylaxis. In fact, both underuse and overuse of VTE prophylaxis occur (18).
The aim of CDSS is to improve adherence to guidelines for prescribing VTE prophylaxis
and the use of institution-tailored tools has been advocated (26, 27). The adequacy of
prophylaxis has been assessed in the intervention group during phase 2 only, resulting in
a rate of appropriate VTE prophylaxis in 76% as compared to 44% prior to the introduction
of CPOE (16). These numbers are in line with the impact of the eAlerts on the rate of
prophylaxis assessed in the present analysis, showing an increase from 43% to 67% due
to the introduction of eAlerts, up to 74% thereafter.
Limitations of our study include that the appropriateness of prophylaxis orders and
the VTE events following discharge were not monitored. Although the rate of appropriate
prophylaxis orders among medical patients was high (76%) during the eAlert
implementation phase, we cannot rule out with certainty that this rate decreased over time
in the post-implementation phase. However, eAlerts significantly reduced the rate of
inappropriate VTE prophylaxis orders in a recent management study in which our
institution participated (28). In addition, the in-hospital VTE rate was very low in both the
medical ward (5 events in 1854 cases) and the surgical ward (6 events in 3463 cases).
Therefore, our study was likely underpowered to detect significant differences in VTE
rates between the three phases.
The major increase of the thromboprophylaxis rate due to the implementation of
eAlerts illustrates the well-known gap between clinical practice and evidence-based
guidelines (29). Reasons for insufficient adherence to guidelines include complexity of
guidelines, lack of physician awareness, concerns over the risk-to-benefit ratio of
prophylaxis and lack of hospital resources (8). In particular, inexperienced interns starting
their career in a teaching hospital may feel uncertain about ordering VTE prophylaxis.
Thus, eAlerts may be considered as trainers for better adherence to guidelines in order to
improve patient safety. The ACCP explicitly recommends the use of CDSS to increase
guideline adherence (10). Linking CDSS with evidence-based risk scores and up-to-date
guidelines may help to enhance the quality of treatment (27).
However, lack of physician compliance with electronically generated suggestions is
a well-known problem. Design and quality of the user interface of a clinical information
system may contribute to its acceptance. Various strategies were proposed to limit alert
fatigue, e.g. providing concise context-sensitive online information, reducing the number
of alerts with low relevance, enforcing user response by interruptive alerts or by
implementing multi-screen alerts (30). Furthermore, the specificity of an alert may be
increased, e.g. by displaying alerts only to selected health care professionals (17, 31).
Indeed, different strategies may be required to overcome alert fatigue, also depending on
national or regional cultures and incentives.
The present eAlert module was designed to reduce the number of unnecessary
prompts. It gives the physician an adequate amount of time to place orders, and an alert is
issued only if no prophylaxis order has been placed within a predefined time frame after
admission. The eAlert module does not identify high-risk patients based on calculation of
a VTE risk score since the identification of individual risk factors by computers may be
unreliable, particularly during the initial hospital stay, owing to incomplete or missing
information in the patient database. However, the guidelines for prophylaxis indications
are displayed by default when the physician acknowledges the eAlert button.
In conclusion, advanced eAlerts for hospitalized medical patients may provide
sustained efficacy with stable rates of thromboprophylaxis over time.
N. Kucher is co-inventor of the WIPO patent application (WO/2006/052952) for a
system and method for determining whether to issue an alert to consider prophylaxis for a
risk reduction (Applicant: Brigham and Women's Hospital, Harvard Medical School,
Boston, MA, USA).
recent advances and future directions. Isr Med Assoc J 2009 Dec;11(12):753-7.
2. Spencer FA, Emery C, Lessard D, et al. The Worcester Venous Thromboembolism
study: a population-based study of the clinical epidemiology of venous thromboembolism.
J Gen Intern Med 2006 Jul;21(7):722-7.
3. Cohen AT, Agnelli G, Anderson FA, et al. Venous thromboembolism (VTE) in
Europe. The number of VTE events and associated morbidity and mortality. Thromb
Haemost 2007 Oct;98(4):756-64.
4. Francis CW. Clinical practice. Prophylaxis for thromboembolism in hospitalized
medical patients. N Engl J Med 2007 Apr 5;356(14):1438-44.
5. Cushman M, Tsai AW, White RH, et al. Deep vein thrombosis and pulmonary
embolism in two cohorts: the longitudinal investigation of thromboembolism etiology. Am J
Med 2004 Jul 1;117(1):19-25.
6. Stein PD, Kayali F, Olson RE. Estimated case fatality rate of pulmonary embolism,
1979 to 1998. Am J Cardiol 2004 May 1;93(9):1197-9.
7. Shojania KG, Duncan BW, McDonald KM, et al. Making health care safer: a critical
analysis of patient safety practices. Evid Rep Technol Assess (Summ) 2001(43):i-x, 1-
8. Cohn SL. Prophylaxis of venous thromboembolism in the US: improving hospital
performance. J Thromb Haemost 2009 Sep;7(9):1437-45.
9. Dentali F, Douketis JD, Gianni M, et al. Meta-analysis: anticoagulant prophylaxis to
prevent symptomatic venous thromboembolism in hospitalized medical patients. Ann
Intern Med 2007 Feb 20;146(4):278-88.
10. Geerts WH, Bergqvist D, Pineo GF, et al. Prevention of venous thromboembolism:
American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th
Edition). Chest 2008 Jun;133(6 Suppl):381S-453S.
11. Cohen AT, Tapson VF, Bergmann JF, et al. Venous thromboembolism risk and
prophylaxis in the acute hospital care setting (ENDORSE study): a multinational cross-
sectional study. Lancet 2008 Feb 2;371(9610):387-94.
12. Tapson VF, Decousus H, Pini M, et al. Venous thromboembolism prophylaxis in
acutely ill hospitalized medical patients: findings from the International Medical Prevention
Registry on Venous Thromboembolism. Chest 2007 Sep;132(3):936-45.
13. Goldhaber SZ, Tapson VF. A prospective registry of 5,451 patients with
ultrasound-confirmed deep vein thrombosis. Am J Cardiol 2004 Jan 15;93(2):259-62.
14. Kucher N, Koo S, Quiroz R, et al. Electronic alerts to prevent venous
thromboembolism among hospitalized patients. N Engl J Med 2005 Mar 10;352(10):969-
15. Baroletti S, Munz K, Sonis J, et al. Electronic alerts for hospitalized high-VTE risk
patients not receiving prophylaxis: a cohort study. J Thromb Thrombolysis 2008
16. Kucher N, Puck M, Blaser J, et al. Physician compliance with advanced electronic
alerts for preventing venous thromboembolism among hospitalized medical patients. J
Thromb Haemost 2009 Aug;7(8):1291-6.
17. van der Sijs H, Aarts J, Vulto A, et al. Overriding of drug safety alerts in
computerized physician order entry. J Am Med Inform Assoc 2006 Mar-Apr;13(2):138-47.
18. Chopard P, Dorffler-Melly J, Hess U, et al. Venous thromboembolism prophylaxis
in acutely ill medical patients: definite need for improvement. J Intern Med 2005
19. Spirk D, Husmann M, Willenberg T, et al. Inconsistencies in the planning of the
duration of anticoagulation among outpatients with acute deep-vein thrombosis. Results
from the OTIS-DVT Registry. Thromb Haemost 2011 Feb 1;105(2):239-44.
20. Dexter PR, Perkins S, Overhage JM, et al. A computerized reminder system to
increase the use of preventive care for hospitalized patients. N Engl J Med 2001 Sep
Hellmann I, Ellis MH. Prevention of venous thromboembolism in medical patients:
systems: a systematic review of functionality and effectiveness. J Am Med Inform Assoc
22. Damiani G, Pinnarelli L, Colosimo SC, et al. The effectiveness of computerized
clinical guidelines in the process of care: a systematic review. BMC Health Serv Res
23. Lecumberri R, Marques M, Diaz-Navarlaz MT, et al. Maintained effectiveness of an
electronic alert system to prevent venous thromboembolism among hospitalized patients.
Thromb Haemost 2008 Oct;100(4):699-704.
24. Franke RH, Kaul JD. Hawthorne Experiments - 1st Statistical Interpretation. Am
Sociol Rev 1978;43(5):623-43.
25. McCarney R, Warner J, Iliffe S, et al. The Hawthorne Effect: a randomised,
controlled trial. Bmc Med Res Methodol 2007 Jul 3;7:-.
26. Chopard P, Spirk D, Beer HJ, et al. Swiss results from a global observational study
of venous thromboembolism risk and prophylaxis use in the acute care hospital setting:
analysis from the ENDORSE study. Swiss Med Wkly 2009 Oct 31;139(43-44):630-5.
27. Tooher R, Middleton P, Pham C, et al. A systematic review of strategies to improve
prophylaxis for venous thromboembolism in hospitals. Ann Surg 2005 Mar;241(3):397-
28. Nendaz MR, Chopard P, Lovis C, et al. Adequacy of venous thromboprophylaxis in
acutely ill medical patients (IMPART): multisite comparison of different clinical decision
support systems. J Thromb Haemost 2010 Jun;8(6):1230-4.
29. Gibson NS, Douma RA, Squizzato A, et al. Application of a decision rule and a D-
dimer assay in the diagnosis of pulmonary embolism. Thromb Haemostasis 2010
30. Fiumara K, Piovella C, Hurwitz S, et al. Multi-screen electronic alerts to augment
venous thromboembolism prophylaxis. Thromb Haemostasis 2010 Feb;103(2):312-7.
31. van der Sijs H, Mulder A, van Gelder T, et al. Drug safety alert generation and
overriding in a large Dutch university medical centre. Pharmacoepidem Dr S 2009
Shiffman RN, Liaw Y, Brandt CA, et al. Computer-based guideline implementation
Legends to figures and tables
Table 1: Rate of prophylaxis during three phases in an intervention ward with eAlerts
compared to a control ward without eAlerts.
Figure 1: Timing of orders for thromboprophylaxis in the medical intervention ward.
intervention with eAlerts
286 / 659
1129 / 1274
88.6 % 1128 / 1244
449 / 67366.7 %
384 / 522
856 / 945
73.6 % <0.0001
90.6 % 0.1645
cases with thromboprophylaxis / total number of cases
pre-implementation eAlert implementationpost-implementation
What is known about this topic?
- Appropriate use of prophylaxis to prevent VTE is an important strategy for improving
safety among hospitalized patients.
- Many patients at risk do not receive prophylaxis despite evidence-based guidelines.
- Computer decision support systems can improve adherence to guidelines on
thromboprophylaxis. However, alert fatigue may compromise the long-term impact of
What does this paper add?
- An advanced eAlert system did provide sustained efficacy over time, resulting in
enhanced rates of thromboprophylaxis orders among hospitalized medical patients.
- Cessation of specific continuing medical education and withdrawal of accompanying
study staff did not reduce the prophylaxis rate.
- Orders were entered sooner as a result of good physician compliance with the eAlert
phase 1 phase 2phase 3
Prophylaxis ordered before admission
Prophylaxis ordered 0-6h after admission
Prophylaxis ordered 6-12h after admission
Prophylaxis ordered >12h after admission
Prophylaxis ordered >12h
Prophylaxis ordered 6-12h
Prophylaxis ordered 0-6h