Secondary Logo

Journal Logo

FEATURES

Optimization of Nursing-Specific Flu Alerts

Cieslowski, Bethany DNP, MA, RN; Brock, Laurie MSN, BSN, RN; Richesson, Rachel L. PhD, MPH; Silva, Susan PhD; Kim, Hyeoneui PhD, MPH, RN

Author Information
CIN: Computers, Informatics, Nursing: September 2020 - Volume 38 - Issue 9 - p 433-440
doi: 10.1097/CIN.0000000000000616
  • Free

Abstract

Clinical decision support (CDS), defined by Kawamoto and Lobach1(p146) as “the act of providing clinicians with knowledge, intelligently filtered or presented at appropriate times to enhance health and healthcare,” has significant potential to improve patient care.2–6 Electronic health records (EHRs) enhance access to clinical information, and CDS translates evidence-based knowledge into practice at the point of care, enabling clinicians to provide the recommended care often missed.1,7 Examples of CDS include smart sets (eg, predefined orders and documentation commonly used together for a visit), best practice alerts (BPAs), and reminders or recommendations for clinicians at the point of care. If implemented correctly, evidence demonstrates that CDS improves patient care outcomes and increases clinician adherence to guidelines or recommended care standards.5,8–10

Although evidence demonstrates the potential of CDS to improve patient care, alerts often trigger at inappropriate times in clinical workflows, distracting end users and ultimately diminishing clinician actions.6 Users ignore up to 96% of alerts.11 A systematic review of barriers for nurse practitioners using CDS listed nuisance alerts and inappropriate timing as significant barriers.12 Furthermore, CDS targeted at the role of nurses is understudied and requires more research to fully understand the impact on nursing workflow and patient outcomes.5,9,10,13 The Centers for Disease Control and Prevention recommends annual flu vaccination for everyone 6 months of age or older.14 Flu prevention is an organizational priority of medical centers across the country. A number of systems have been designed to monitor and improve screening and administration of the flu vaccine to eligible inpatients.15 However, hospitals continue to miss opportunities to offer the vaccine to patients prior to discharge. Alerts embedded within the EHR have been shown to improve compliance with preventive care recommendations such as flu vaccine screening and administration.7

In October 2012, nursing responsibility at the University of Virginia Medical Center (UVAMC) included the administration of flu vaccine, and alerts were implemented as part of the EHR to ensure compliance with flu vaccine screening and administration protocols. Nurses reported initial alerts as helpful, but subsequent alerts were disruptive and frustrating. Unnecessary alerts have been reported to cause alert fatigue and to disrupt nurses' ability to perform.16 Frequent interruptions and misplaced alerts caused frustration and had a negative effect on the nurse's ability to deliver high-quality care.12,17 Alerts (called “Best Practice Advisories” or BPAs in the Epic EHR system) can be used to remind clinicians to screen, order, administer, or document the flu vaccine. The UVAMC implemented and uses six institutionally defined BPAs (collectively called flu BPAs) within the Epic EHR for this purpose (Table 1). For example, BPA 1 reminds the nurse to administer the vaccine, BPA 2 (ICU) and 4 (inpatient unit) trigger to remind the nurse to order the vaccine, BPA 3 functions to prompt the nurse to document a previously administered vaccine, and BPA 5 (ICU) and 6 (inpatient unit) operate to remind the nurse to complete the flu screening.

Table 1 - BPA Description
BPA Name Triggers Action
BPA 1 Inpatient Influenza Vaccine Ordered
Not Given
• Patient admitted • Remind nurse to administer flu vaccine for
patients who have not received it and
nearing discharge
• Influenza vaccine not charted
• Patient location: Inpatient unit
BPA 2 Flu Immunization ICU Nurse Protocol • Patient admitted • Remind ICU nurse to screen patient and enter protocol order for flu vaccine
• Patient has transfer order
• Negative flu screening
• Immunization not ordered
• Patient location: ICU
BPA 3 Historical Flu Immunization Requires Documentation • Patient admitted • Remind nurse to document the vaccine
administration date in the EHR
• Positive flu screen criteria
• Historical flu immunization for current flu season not documented
• Patient location: Inpatient unit and ICU
BPA 4 Flu Immunization Acute Care Nurse
Protocol
• Patient admitted • Remind nurse to screen patient and enter protocol order for flu vaccine
• Negative flu screening
• Immunization not ordered
• Patient location: Inpatient unit
BPA 5 ICU Flu Screen Check • Patient admitted • Remind ICU nurse to complete flu screening for patient
• Patient admitted to ICU
• Patient has transfer order
• Flu screen not completed
BPA 6 Acute Care Flu Screen Check • Patient admitted • Remind acute care nurse to screen patient for flu vaccine
• Patient admitted to acute care unit
• Flu screen not completed
The six institutionally defined flu alerts designed to remind the nurse to screen, order, administer, or document the flu vaccine within the Epic EHR.

The UVAMC collects and monitors data on the activity of all EHR-integrated BPAs, including the six flu BPAs described above, using LogicStream (LogicStream Health, Minneapolis, MN), a commercial quality measure reporting system. These data show that during the previous flu season (October 1, 2017, to March 31, 2018), BPA 4 (designed to alert when a patient is admitted, screening detected that patient had not received the vaccine, and an immunization was not ordered) triggered 54 103 during this 6 months, and the desired action was performed only 8.8% of the time.18 This high rate of triggers and low rate of actions suggested that flu alerts required reevaluation and possible optimization to address nursing frustration and reduce alert fatigue.

Project Aims

The overall goal of this quality improvement project was to redesign the triggering criteria for the current flu BPAs at the UVAMC to optimize them to “trigger” less often but still remind nurses to screen, order, administer, or document the vaccine. The specific aims were to (1) reduce total number of flu BPA triggers per hospital admission and stay during the flu season; (2) maintain compliance rates at or above 90% for flu vaccine screening, documentation, and administration during flu season; and (3) reduce the rate of dismissal actions, specifically the “cancel BPA” action, for alerts 2 and 4, designed to remind the nurse to order the vaccine (Table 1).

METHODS

Design

Using a mixed-methods design, this project evaluated the flu BPA alerts that nursing staff use to screen, order, and document administration for all patients aged 6 months or older for the flu vaccine. The intervention was the implementation of the redesigned triggering criteria for the current flu BPAs at the UVAMC. The total number of flu BPA triggers per hospital admission, flu compliance at or above 90%, and dismissal actions for BPA 2 (ICU reminder to order vaccine) and 4 (inpatient reminder to order vaccine), during the flu season before and after the intervention, were compared. The preintervention flu season was October 1, 2017, through March 31, 2018, while the postintervention flu season was October 1, 2018, through March 31, 2019. The duration of each flu season was 6 months. To evaluate the impact of the revised BPA design, we used a focus group at the end of the postintervention flu season to explore nursing perceptions.

Setting

The UVAMC is a 600-bed, level I trauma center in central Virginia. This project focused on inpatient units throughout the medical center.

Sample

Intervention Samples

The intervention samples included hospital admissions for all inpatients aged 6 months or older during the preintervention and postintervention flu season. All hospital admissions from inpatient acute care settings for whom EHRs were used to admit, screen, order, and document flu immunizations were included. Inpatients younger than 6 months of age were excluded.

Focus Group

A 45-minute focus group session with 12 nursing staff, held after flu season, evaluated the refinements to the flu BPAs. Nursing staff end users were recruited throughout the medical center and invited to participate on a voluntary basis. Institutional review board (IRB) approval was obtained for the focus group in this quality improvement project.

Measures

Each hospital admission was defined as an “encounter,” and the number of flu BPA triggers was determined per encounter. Flu compliance was defined as whether or not appropriate action was taken for a patient older than 6 months who had been screened for the flu vaccine. Appropriate actions include no action (eg, previous allergy, patient refused), previous administration of the flu vaccine for the season (eg, historical administration), or administration of the vaccine during the inpatient visit. Dismissal actions for alerts 2 and 4 included the number of times the nurse pressed the “cancel BPA” button. Patient and encounter identifiers defined as “ID” allowed for multiple encounters by the same patient.

Data Collection

Data required for analysis included unique patient ID, encounter ID, date of visit, and the number of triggers and dismissals per BPA. The data query obfuscated patient ID, encounter, and date of visit (eg, preintervention/postintervention group) to maintain patient confidentiality. Compliance rate data emanated from the UVA dashboard for pre- and post-flu seasons. Data from the EHR reporting systems, end user meetings, and nursing staff focus group sessions help to evaluate the implementation.

Intervention

We talked to users and systematically applied the “five rights” of CDS to redesign the current suite of six nursing-specific flu BPAs to optimize triggering and integration into more appropriate times in the nursing workflow. Consistent with best practice on CDS implementation, users were included in the redesign process.19 End user groups were consulted about issues with current BPAs and asked for suggestions. Their suggestions were incorporated to maximize the appropriate placement of BPAs in the workflow. Workflow analyses of current and possible redesign approaches were presented for feedback.

The five rights of CDS, a defined framework for evaluating, improving, and implementing the redesigned alerts,1,5,6 were used as a model for this ideal BPA design. Using this model, the right information must be communicated (eg, nurse-driven flu protocol), to the right person (eg, nurses), in the right format (eg, alert or banner), through the right channel (eg, EHR), at the right time in the clinical workflow.1,20 All suggested or proposed changes were mapped to (or needed to address) one of the five rights. Table 2 summarizes the changes made to flu BPAs.

Table 2 - Summary of Changes to BPAs
Changes Made BPA Involvement (1–6) CDS Rights Framework Notes
Simplify text All Right information Clarified protocol instruction and intended
actions of BPA
Reorder buttons for consistency All Right information Standardized display of all BPAs
Feedback link All Right information, person,
format, channel, time
Provided feedback option for end users
regarding BPAs
Remove acknowledgement reasons 1 Right information, format, time Removed acknowledgement reasons that
should be documented elsewhere in Epic
Add acknowledgement reasons 2, 4 Right information, time Allowed RN more options for flu screening
Renamed acknowledgement reasons 1, 3, 5, 6 Right information, time Clarified action taken by nurses
Remind me in 4 h acknowledgement reason 2, 3, 4, 5 Right information, time Provided RN specified amount of time to
complete screening, ordering or documentation of historical vaccine action required by BPA
Remind me in 8 h acknowledgement reason 6 Right information, time Provided RN time to complete flu screening
action required by BPA
Remind me in 24 h acknowledgement reason 1 Right information, time Provided RN 24 h to complete administration of vaccine already ordered
Not primary RN lock out times
increased
2, 3, 4, 5, 6 Right person, time Prevented nuisance alerts
Lockouts times adjusted All Right time Allowed RN time to complete action required
by BPAs
Banner added 1, 3, 5, 6 Right format, time Provided a passive form of CDS that allows RN to compete action at more appropriate time
in workflow
Blocked BPA from triggering at night All Right person, time Allowed less interruption in RN workflow while patients sleep
Summarized changes to the six alerts using the five rights framework.

The Right Information

We standardized the display of information on all six BPAs to deliver the “right information.” This included eliminating unnecessary text and specifying clear instructions on the BPAs to include the appropriate level of detail needed to complete intended actions of the alert. Furthermore, we standardized and clarified descriptions on acknowledgment buttons (eg, changed “will document” to “remind me in 4 hours”) to reflect actions selected by nurses.

The Right Person

Alerts trigger for nursing end users, but nurses have different care roles. For instance, nursing case managers frequently review patient charts but rarely administer vaccines. The existing BPAs had the option for “not primary nurse” or “not primary team” option. We increased the lockout times, the criteria specifying the time between BPA triggers, to allow alternate nursing care roles more time in the EHR without nuisance alerts.

The Right Format

The implementation of banners, another form of CDS, facilitated the “right intervention format.” Banners use hyperlinks to access the actions encouraged by the BPAs (eg, screening, ordering, documenting) in a less disruptive manner. Deploying banners as a complement to BPAs enabled nurses to return to the actions required for vaccine administration at a time more convenient in their workflow.

The Right Channel

The EHR provided the “right channel” for flu vaccine alerts and banners.

The Right Time in Clinical Workflow

To adopt the “right time” suggested in the framework, we used feedback from end users and hospital policy to advise changes to the timing of alerts. For example, we increased the lockout times on acknowledgement buttons (eg, “not primary team” lockout time changes from 1 to 12 hours), and in accordance with hospital policy and the need for quiet at night hours, we blocked BPAs from triggering between 10:00 pm and 5:00 am, enabling patients to sleep as vaccinations usually occur during the day shift.

Testing

The redesigned BPAs were tested prior to implementation. The refined BPAs started in the proof-of-concept environment for initial testing and then migrated to the test environment after review from end users and informatics experts. Upon completion of BPA testing, alerts rolled out for hospital-wide implementation in the production environment.

Focus Group

To evaluate the impact of the revised BPA design, we conducted a focus group session at the end of the postintervention flu season to explore nursing perceptions. Participants included EHR nursing end users, and participation was voluntary. Nurses were recruited a week prior to the session by hanging information statements in break rooms on select units. All attendees consented, and IRB approval was granted prior to the session. The focus group lasted 45 minutes and occurred on March 19, 2019. The focus group was recorded using a digital voice recording device and transcribed. The format included a series of open-ended questions that explored the perceived attributes of the revised BPA implementation—specifically, the right information, right person, right format, right channel, and right time in the clinical workflow. Examples of participant questions included:

  1. Elaborate on the information provided by the alerts to the nurse in using or understanding the BPAs.
  2. Elaborate on the information provided on the buttons.
  3. Elaborate on timing of the alerts in the nursing workflow and how the alerts could be timed better.

The feedback from nursing participants was explored, and common themes contributed to the evaluation of the redesigned BPAs.

RESULTS AND ANALYSIS

Descriptive statistics were used to describe the total number of flu BPA triggers per encounter, flu compliance rates, and dismissal actions for BPA 2 (ICU reminder to order vaccine) and 4 (inpatient reminder to order vaccine) during the preintervention and postintervention flu seasons. Nonparametric Wilcoxon two-sample tests were used to test for a preintervention and postintervention flu season difference in the total number of flu BPA triggers per encounter due to the severe skewness of the data. χ2 Tests were used to examine for a flu season difference in dismissal actions for BPA 2 and BPA 4. Data were analyzed using SAS version 9.4 (SAS Institute, Cary, NC). Nondirectional statistical tests were performed with the level of significant set at .05 for each test.

Best Practice Advisory Results

Table 3 presents the total number of triggers per encounter for each flu season. For each BPA type, N represents the number of encounters in which there was at least one triggering for the BPA type specified. The median number of triggers per encounter was significantly lower for any type of BPA, BPA 2 (ICU reminder to order vaccine), BPA 3 (documentation of previously administered vaccine), BPA 5 (ICU reminder to screen), and BPA 6 (inpatient reminder to screen) in the postintervention flu season compared to preintervention flu season (all P ≤ .02). The 25th and 75th percentile scores for the total number of triggers per encounter indicate a reduction in the variability of these data during the postintervention flu season. This resulted in a significant preintervention and postintervention flu season difference for BPA 3 (documentation of previously administered vaccine) despite a median score of 2 during both flu seasons. There was no significant difference between flu seasons for total number of triggers per encounter with regard to BPA 1 (inpatient reminder to administer) or BPA 4 (inpatient reminder to order).

Table 3 - Type of BPA: Triggers Per Encounter
BPA Type and Definition BPA Type Preintervention Flu Season Postintervention Flu Season P
Any BPA Encounters (N) 10 527 10 481 <.001
Median 5 4
25th, 75th percentile 2, 15 2, 10
BPA 1
Inpatient reminder to administer
Encounters (N) 2796 2515 .21
Median 3 3
25th, 75th percentile 1, 8 2, 7
BPA2
ICU reminder to order
Encounters (N) 359 340 .01
Median 13 10
25th, 75th percentile 4, 38 4, 22
BPA 3
Reminder to document previously administered vaccine
Encounters (N) 4035 4169 .02
Median 2 2
25th, 75th percentile 1, 11 1, 7
BPA 4
Inpatient reminder to order
Encounters (N) 2998 2613 .96
Median 2.0 2.0
25th, 75th percentile 2, 15 2, 11
BPA 5
ICU reminder to screen
Encounters (N) 1437 1624 <.001
Median 4.0 3.0
25th, 75th percentile 2, 11 2, 7
BPA 6
Inpatient reminder to screen
Encounters (N) 4429 4765 <.001
Median 3 2
25th, 75th percentile 1, 5 1, 3
N represents number of encounters with at least one BPA of the specified type triggered and median number of triggers per encounter reported. P value for Wilcoxon two-sample test.

Table 4 presents the total number of alerts triggered for BPA 2 (ICU reminder to order vaccine) and 4 (inpatient reminder to order vaccine) and details the distribution of the BPA actions completed during the preintervention and postintervention flu seasons. Less than 1% of the data were categorized as null or unknown during each flu season.

Table 4 - BPA Dismissal Reasons
BPA Type and Definition BPA Action Preintervention Flu Season Postintervention Flu Season
n (%) n (%)
BPA 2
ICU reminder to order vaccine
Total alert triggers (N) 13 551 7857
Cancel BPA 10 675 (78.8) 5226 (66.5)
Accept BPA 16 (0.1) 6 (0.1)
Acknowledge/override warning 2150 (15.9) 2213 (28.2)
Open order set 664 (4.9) 386 (4.9)
Null (unknown action) 46 (0.3) 26 (0.3)
BPA 4
Inpatient reminder to order vaccine
Total alert triggers (N) 58 771 30 033
Cancel BPA 44 763 (76.2) 18 148 (60.4)
Accept BPA 127 (0.2) 48 (0.2)
Acknowledge/override warning 9049 (15.4) 8584 (28.6)
Open order set 4615 (7.9) 3160 (10.5)
Null (unknown action) 217 (0.4) 93 (0.3)
N represents the total number of alert triggers and the possible BPA action options for BPA 2 and 4. For each BPA action, the number (n) and percent (%) of the total alerts triggered (N) are reported.

Table 5 presents the total number of alerts triggered in which the BPA action was known for BPA 2 (ICU reminder to order vaccine) and 4 (inpatient reminder to order vaccine). That is, the alerts with a null or unknown action were excluded. The known BPA actions were then grouped into dismissals (cancel BPA) and nondismissals (accept BPA, acknowledge/override, open order set) actions. For BPA 2, 79.0% of BPA triggers during preintervention season were dismissed compared to 66.7% during the postintervention season (P < .001). For BPA 4, 76.5% of BPA triggers during the preintervention season were dismissed compared to 60.6% of triggers in the postintervention season (P < .001).

Table 5 - BPA 2 and BPA 4: BPA Dismissals for Alerts Triggered
BPA Definition BPA Type Preintervention Flu Season Postintervention Flu Season P
n (%) n (%)
BPA 2
ICU reminder to order vaccine
Alerts triggered with known BPA action (N) 13 505 7831 <.001
Nondismissal 2830 (21.0) 2605 (33.3)
Dismissal 10 675 (79.0) 5226 (66.7)
BPA 4
Inpatient reminder to order
vaccine
Alerts triggered with known BPA action (N) 58 554 29 940 <.001
Nondismissal 13 791 (23.6) 11 792 (39.4)
Dismissal 44 763 (76.5) 18 148 (60.6)
N represents the number of alerts triggered with a known BPA action; known actions for BPA 2 and 4 grouped into nondismissal and dismissals; number (n) and percent (%) of the alerts triggered with a known BPA action (N) that were nondismissal and dismissal actions are reported. For each BPA, a 2 × 2 χ2 test was performed to test for differences in proportions.

The flu compliance rate was 95.0% at the end of the preintervention flu season compared to 97.8% at the end of postintervention flu season, resulting in a 2.8% percentage point improvement in compliance rate. Figure 1 displays the postintervention compliance rate obtained from the UVAMC Dashboard.

FIGURE 1
FIGURE 1:
Flu vaccine dashboard. Flu compliance rates.

Focus Group Results

In a classroom at the hospital, 14 nurses from acute care and ICU settings participated in the focus group at the end of flu season. The nurses reported that the alerts delivered the right amount of information to complete the requested flu action. The lack of alert triggers during patient sleeping hours was viewed as the most significant change and was appreciated by the nurses. Additionally, staff commented positively on the addition of banners and the ability to administer or document the flu vaccine at a more convenient time. Recommendations included more specific changes to the timing of the alerts and nursing workflow. For example, alerting nurses twice a day (eg, morning and evening) during typical assessment hours was suggested. One nurse commented that there was redundancy among the banners and the alerts and felt not as many alerts were needed with the banners. Additionally, nurses communicated that they liked the reminders but the alerts still triggered too often.

DISCUSSION

The BPA redesign resulted in more encounters with fewer alerts and a decrease in the proportion of dismissal actions taken by nurses. Moreover, the changes to the flu BPAs resulted in an improvement in flu compliance rates. Focus group feedback identified important information about CDS. Nurses like an initial alert but found the subsequent alerts disruptive to workflow, even after the BPA redesign. The nurses appreciated the banners because they provided a reminder, but allowed nurses to address the actions required for flu vaccine administration at a time appropriate in their workflow. Participants commented positively on the banners, but felt they did not need as many alerts as a result. The nurses expressed gratitude for the reduction of alerts during patient sleeping hours. This feedback confirms the importance of timing in alert triggers for both the nurse and the patient.

Despite the improvement achieved with the redesign, nurses still complained that the alerts triggered too often and at inappropriate times. The nursing informatics team and the flu committee evaluated the results from this project and developed a plan for the next flu season aimed to further reduce the number of alert triggers and meet end user needs.

Using the triggering data from this project, the team reconfigured BPA 1 (Inpatient reminder to administer), BPA 2 (ICU reminder to order), BPA 3 (documentation of previously administered), and BPA 4 (inpatient reminder to order) to alert the nurse one time. In addition to the first alert, the team implemented banners for each of the four BPAs. The modifications to the alerts enable nurses to address the required flu action at a time convenient in their workflow without subsequent, disruptive alerts, and address the redundancy between banners and alerts. Triggering criteria for BPA 5 (ICU reminder to screen patient) and BPA 6 (inpatient reminder to screen) remained the same. To prevent a patient from being discharged without flu screening or administration, the informatics team included the addition of a new “discharge” BPA that triggers with discharge order entry. The BPA reminds the nurse to address flu compliance actions before the patient leaves the medical center. The new configuration of flu alerts builds on the data and the end user feedback collected in this quality improvement project.

Traditionally, most CDS evaluation studies explore the dismissal rate rather than the total number of triggers or dismissals.4,21 We chose to look at the total number of triggers per encounter and the proportion of dismissals because it shows the interruptive nature of alerts in the nursing workflow. Providing alerts at the appropriate time in the clinical workflow is one of the most challenging aspects of decision support.22 This is critically important with nursing-targeted CDS because of the amount of time nurses spend using the EHR to complete tasks.

Limitations

Length of stay and patient acuity were not included as part of the study and could not be included as covariates in the analysis. Thus, the impact of clinical status of the patient during the hospital admission on the outcomes could not be assessed. This might explain why some BPAs triggered more than 500 times during an encounter.

CONCLUSION

Nurses appreciate CDS in clinical practice, but CDS requires constant evaluation and reevaluation to appropriately meet end user needs. What made this project successful was reducing the number of triggers per encounter and introducing banners, a passive form of CDS that enables nurses more autonomy in flu compliance actions. Additionally, end users guided and evaluated the intervention throughout the process. We demonstrated that findings from this project have implications for future redesign of nursing CDS to adequately decrease unnecessary alerts, determine the appropriateness of banners to meet end user needs, and maintain flu compliance. Nursing-targeted CDS requires more research to understand its effectiveness in nursing workflow.

References

1. Kawamoto K, Lobach DF. Proposal for fulfilling strategic objectives of the U.S. roadmap for national action on clinical decision support through a service-oriented architecture leveraging HL7 services. Journal of the American Medical Informatics Association. 2007;14(2): 146–155. doi:10.1197/jamia.M2298.
2. Kawamoto K, Lobach DF. Design, implementation, use, and preliminary evaluation of SEBASTIAN, a standards-based Web service for clinical decision support. American Medical Informatics Association Annual Symposium Proceedings. 2005;380–384.
3. Cohn WF, Lyman JA, Detmer DE, Bloomrosen M. Clinical decision support: progress and opportunities. Journal of the American Medical Informatics Association. 2010;17(5): 487–492. doi:10.1136/jamia.2010.005561.
4. Nanji KC, Bates DW, Slight SP, et al. Overrides of medication-related clinical decision support alerts in outpatients. Journal of the American Medical Informatics Association. 2013;21(3): 487–491. doi:10.1136/amiajnl-2013-001813.
5. Lopez KD, Febretti A, Stifter J, Johnson A, Wilkie DJ, Keenan G. Toward a more robust and efficient usability testing method of clinical decision support for nurses derived from nursing electronic health record data. International Journal of Nursing Knowledge. 2017;28(4): 211–218. doi:10.1111/2047-3095.12146.
6. Greenes RA, Bates DW, Kawamoto K, Middleton B, Osheroff J, Shahar Y. Clinical decision support models and frameworks: seeking to address research issues underlying implementation successes and failures. Journal of Biomedical Informatics. 2018;78: 134–143. doi:10.1016/j.jbi.2017.12.005.
7. Bright TJ, Wong A, Dhurjati R, et al. Effect of clinical decision-support systems: a systematic review. Annals of Internal Medicine. 2012;157(1): 29–43. doi:10.7326/0003-4819-157-1-201207030-00450.
8. Kawamoto K, Houlihan CA, Balas EA, Lobach DF. Improving clinical practice using clinical decision support systems: a systematic review of trials to identify features critical to success. BMJ. 2005;330(7494): 765. doi:10.1136/bmj.38398.500764.8F.
9. Anderson JA, Willson P. Clinical decision support systems in nursing: synthesis of the science for evidence-based practice. CIN: Computers, Informatics, Nursing. 2008;26(3): 151–158. doi:10.1097/01.NCN.0000304783.72811.8e.
10. Lytle KS, Short NM, Richesson RL, Horvath MM. Clinical decision support for nurses: a fall risk and prevention example. CIN: Computers, Informatics, Nursing. 2015;33(12): 530–537. doi:10.1097/cin.0000000000000192.
11. Moxey A, Newby D, Robertson J, Hains I, Pearson SA, Williamson M. Computerized clinical decision support for prescribing: provision does not guarantee uptake. Journal of the American Medical Informatics Association. 2010;17(1): 25–33. doi:10.1197/jamia.M3170.
12. Borum C. Barriers for hospital-based nurse practitioners utilizing clinical decision support systems: a systematic review. CIN: Computers, Informatics, Nursing. 2018;36(4): 177–182. doi:10.1097/CIN.0000000000000413.
13. Whalen K, Bavuso K, Bouyer-Ferullo S, et al. Analysis of nursing clinical decision support requests and strategic plan in a large academic health system. Applied Clinical Informatics. 2016;7(2): 227–237. doi:10.4338/ACI-2015-10-RA-0128.
14. Centers for Disease Control and Prevention. (2019). Flu season. Available at https://www.cdc.gov/flu/about/season/current.htm. Accessed February 10, 2019.
15. Centers for Medicare & Medicaid Services and the Joint Commission. Specifications Manual for National Hospital Inpatient Quality Measures. 2018. https://www.jointcommission.org/specifications_maual_for_national_inpatient_quality_measures.aspx. Accessed February 10, 2019.
16. Sidebottom A, Collins B, Winden T, Knutson A, Britt H. Reaction of nurses to the use of electronic health record alert features in an impatient setting. CIN: Computers, Informatics, Nursing. 2012;30(4): 218–226.
17. Miller A, Moon B, Anders S, Walden R, Brown S, Montella D. Integrating computerized clinical decision support systems into clinical work: a meta-synthesis of qualitative research. International Journal of Medical Informatics. 2015;84(12): 1009–1018. https://doi.org/10.1016/j.ijmedinf.2015.09.005.
18. LogicStream Health [University of Virginia: Alerts Management Software]. Minneapolis, MN; 2018: https://cds.logic-stream.net/. Accessed, September 1, 2017.
19. Health IT. Benefits of EHRs: improved diagnostics & patient outcomes. https://www.healthit.gov/sites/default/files/3-4-5-measure-effects-and-refine-cds-interv.pdf. Accessed April 9, 2019.
20. Sirajuddin AM, Osheroff JA, Sittig DF, Chuo J, Velasco F, Collins DA. Implementation pearls from a new guidebook on improving medication use and outcomes with clinical decision support. Effective CDS is essential for addressing healthcare performance improvement imperatives. Journal of Healthcare Information Management. 2009;23(4): 38–45.
21. Nightingale PG, Adu D, Richards NT, Peters M. Implementation of rules based computerized bedside prescribing and administration: intervention study. British Medical Journal. 2000;320(7237): 750–753. doi:10.1136/bmj.320.7237.750.
22. Sittig DF, Wright A, Osheroff JA, et al. Grand challenges in clinical decision support v10. Journal of Biomedical Informatics. 2008;41(2): 387–392. doi:10.1016/j.jbi.2007.09.003.
Keywords:

Alerts; Clinical decision support systems; Electronic health records; Flu vaccine

Copyright © 2020 Wolters Kluwer Health, Inc. All rights reserved.