Skip Navigation
Innovation Attempt

Alerts and Decision Support for Abnormal Laboratory Values Do Not Improve Clinical Management or Reduce Adverse Events


Tab for The Profile
Comments
(0)
   

Snapshot

Summary

Residents and on-call physicians on two internal medicine teaching units received alerts whenever a patient had any of six abnormal laboratory values, along with a link to decision support related to the finding. The program had no impact on the clinical management of these patients or the frequency of adverse events. Program developers nonetheless believe the approach could be highly effective if the alerts and decision support were better integrated into existing workflow and the system made easier to use. They are working on these refinements and plan to test a revamped program in the future.
begin doxml

Developing Organizations

Sunnybrook Health Sciences Centre
Toronto, Canadaend do

Date First Implemented

2008
The initial trial ran from January to June.

Problem Addressed

Patients with laboratory abnormalities sometimes receive suboptimal care and may experience adverse events because physicians lack access to timely information and support in managing these patients.
  • Suboptimal clinical management: Studies of the clinical management of patients with various laboratory abnormalities have found wide variations in the degree to which physicians adhere to established guidelines, with adherence rates ranging between 40 and 76 percent.1-5
  • Unrealized potential of real-time alerts and support: Poor management may be the result of physicians not having easy access to timely information, including both the abnormal value itself and guidance on how to handle the abnormality. Implemented effectively, point-of-care, real-time alert systems and computerized decision support may help physicians better manage these situations.

What They Did

Back to Top

Description of the Innovative Activity

Residents and on-call physicians on two internal medicine teaching units received alerts whenever a patient had any of six abnormal laboratory values, along with a link to decision support related to managing patients with such an abnormality. Key program elements include the following:
  • Alerts about abnormal values: Residents or on-call physicians received alerts via smartphone or pager whenever a patient had any of six common abnormal laboratory values: low serum potassium (with two separate alerts depending on whether or not the patient had an active drug order for a diuretic that causes loss of potassium), high serum potassium, low or high serum sodium, high international normalized ratio, and low hemoglobin. Specific cutoff values existed for each alert. The process worked differently during normal working hours than at night and over the weekend, as outlined below:
    • Normal working hours: Senior residents received the alerts via smartphone and could forward them to junior residents if desired. Junior residents accessed the alerts via smartphone on one unit and via pager or computer workstation on the other.
    • Nights and weekends: On one unit, the on-call physician carried the senior resident's smartphone, allowing the alerts to come via this device. On the other unit, the on-call physician did not have the resident's phone and instead received word of the abnormal finding via a page from the laboratory. (The senior resident took call every fourth night on this unit and received alerts via smartphone on those nights.)
  • Decision support: Each alert contained a link to Web-based decision support related to the abnormal value. The link provided treatment recommendations that had been developed by physicians after reviewing relevant literature and existing hospital policies, in consultation with local specialists who reviewed and commented on early drafts.

Did It Work?

Back to Top

Results

The program had no impact on the clinical management of patients or the frequency of adverse events.
  • No change in clinical management: An analysis of 498 instances of abnormal laboratory values found that clinicians performed about half of clinical actions that could have been taken, regardless of whether the alert system was turned on or off. (The analysis compared actions taken during a 2-month period before the alerts had been turned on to a similar-length period after activation.)6
  • Statistically insignificant impact on adverse events: Before the alert system had been activated, roughly one-third abnormal values resulted in an adverse event. After activation, the frequency of adverse events increased modestly (to 42 percent), although this change did not meet the test of statistical significance.6

What They Learned

Back to Top
  • Integrate system into physician scheduling: This program will not be effective unless the right physician sees the alert in a timely manner. In pilot tests conducted at Sunnybrook (for this program and others), one in eight pages went to the wrong doctor (e.g., someone on vacation or not on duty).7,8 To correct this problem, the alert system will be integrated with the physician scheduling system, thus allowing it to know which doctor is responsible for the patient at the time of the alert (who may be someone other than the ordering physician). 
  • Integrate system into existing workflow: Physicians must be able to review, research, and act on any alert quickly and easily. During the initial trial, however, the physician often had to access multiple systems to accomplish relevant tasks, and in many cases skipped steps deemed nonessential. To address these issues, program leaders are working to integrate the alert system with the decision support and test ordering systems.
  • Make system and decision support easy to use: Usability experts emphasize the importance of making it as easy as possible to use products correctly. During the pilot test, physicians who received the alert on a pager had to scroll down to see the entire body of the text. Many did not realize that more information appeared at the bottom, including the link to decision support. Sunnybrook is currently revamping the system to promote usability, including making the link to decision support more visible and accessible.
  • Follow principles of user-centered design: To promote integration and usability, design the system with ongoing input from those who will be using it. Going forward, Sunnybrook designers will not turn any product over to programmers until functional prototypes have been tested with users and refined based on their feedback.
  • Be patient: Creating an effective system can take a long time, as unexpected issues may come up along the way. For example, program developers did not initially consider the potential for alerts to go to the wrong physician, and hence the need for integration with the scheduling system did not become apparent until the problem surfaced during testing.
  • Do not depend on audit and feedback: Monitoring performance and feeding back information to physicians requires significant effort that often yields lackluster results. The better approach is to create the integrated, easy-to-use system outlined earlier. Physicians will naturally gravitate toward using it, because doing so will save them time and effort, thus promoting better management of abnormalities without the need for intrusive, time-consuming audit and feedback.

More Information

Back to Top

Contact the Innovator

Edward Etchells, MD, MSc
Associate Director of the University of Toronto Centre for Patient Safety
Medical Director of Information Services and Staff Physician in the Division of General Internal Medicine at Sunnybrook Health Sciences Centre
Associate Professor of Medicine at the University of Toronto
2075 Bayview Avenue, Room H469
Toronto, Ontario M4N 3M5
Phone: (416) 480-5996
Fax: (416) 480-6777
E-mail: edward.etchells@sunnybrook.ca

References/Related Articles

Etchells E, Adhikari NKJ, Wu R, et al. Real-time automated paging and decision support for critical laboratory abnormalities. BMJ Qual Saf. 2011;20:924-30. [PubMed]

Wong BM, Quan S, Shadowitz S, et al. Implementation and evaluation of an alphanumeric paging system on a resident inpatient teaching service. Journal of Hospital Medicine. 2009 October;8(4):e34-40. Available at: http://utoronto.academia.edu/BrianWong/Papers/1083399

Wong BM, Quan S, Cheung CM, et al. Frequency and clinical importance of pages sent to the wrong physician. Arch Intern Med. 2009;169(11):1072-3. [PubMed] Available at: http://archinte.jamanetwork.com/article.aspx?articleid=415030

Footnotes

1 Paltiel O, Salakhov E, Ronen I, et al. Management of severe hypokalemia in hospitalized patients: a study of quality of care based on computerized databases. Arch Intern Med 2001;161:1089-95. [PubMed]
2 Acker CG, Johnson JP, Palevsky PM, et al. Hyperkalemia in hospitalized patients. Arch Intern Med. 1998;158:917-24. [PubMed]
3 Iorio A, Basileo M, Marchesini E, et al. Audit of the clinical use of fresh-frozen plasma in Umbria: study design and results of the pilot phase. Blood Transfus. 2008;6:211-19. [PubMed]
4 Rothschild JM, McGurk S, Honour M, et al. Assessment of education and computerized decision support interventions for improving transfusion practice. Transfusion. 2007;47:228-39. [PubMed]
5 Palevsky PM, Bhagrath R, Greenberg A. Hypernatremia in hospitalized patients. Ann Intern Med. 1996;124:197-202. [PubMed]
6 Etchells E, Adhikari NKJ, Wu R, et al. Real-time automated paging and decision support for critical laboratory abnormalities. BMJ Qual Saf. 2011;20:924-30. [PubMed]
7 Wong BM, Quan S, Shadowitz S, et al. Implementation and evaluation of an alphanumeric paging system on a resident inpatient teaching service. Journal of Hospital Medicine. 2009 October;8(4):e34-40. Available at: http://utoronto.academia.edu/BrianWong/Papers/1083399
8 Wong BM, Quan S, Cheung CM, et al. Frequency and clinical importance of pages sent to the wrong physician. Arch Intern Med. 2009;169(11):1072-3. [PubMed]
Comment on this Innovation

Disclaimer: The inclusion of an innovation in the Innovations Exchange does not constitute or imply an endorsement by the U.S. Department of Health and Human Services, the Agency for Healthcare Research and Quality, or Westat of the innovation or of the submitter or developer of the innovation. Read more.

Original publication: July 03, 2012.
Original publication indicates the date the profile was first posted to the Innovations Exchange.

Last updated: August 27, 2014.
Last updated indicates the date the most recent changes to the profile were posted to the Innovations Exchange.

Look for Similar Items by Subject
Patient Care Process:
Organizational Processes: