Triggers and Targeted Injury Detection Systems: Aiming for the Right Target With the Appropriate Tool
Stephanie Shimada, Ph.D.a
Peter E. Rivard, Ph.D.b
Hillary J. Mull, M.P.P.a
Jonathan R. Nebeker, M.D., M.S.c
Amy Rosen, Ph.D.a
Despite significant attention to detecting inpatient adverse events, very little research has focused on the ambulatory setting.C1-C3 As described in our task order (The Development and Use of Ambulatory Adverse Event Trigger Tools, AHRQ Task Order No. 3, HHSA 2902006000012), we aim to (1) generate a set of triggers to detect adverse events in the ambulatory setting through an iterative approach involving a literature review of existing triggers and input from a clinical advisory panel, focus groups, and a Delphi panel of experts in trigger tools and related methodologies; (2) conduct an indepth baseline assessment of ambulatory adverse events in the Veterans Health Administration, Intermountain Healthcare, and Boston Medical Center using these triggers; (3) revise and improve the triggers based on results obtained in (2); and (4) evaluate which triggers are most useful, actionable, and amenable to targeted injury detection system (TIDS) development through feedback to quality managers, patient safety officers, clinicians, and managers in each of the three systems.
The three sites assembled for this project combine the presence of strong local expertise and interest in trigger tool development and diversity of patient populations, ambulatory care settings, and information systems. The VA is the largest integrated health care system in the Nation, designed to address the health care needs of the Nation's veterans through a variety of ambulatory settings. Boston Medical Center is an urban academic medical center with a racially and socioeconomically diverse mix of ambulatory patients due to its presence in an underserved and working class neighborhood. Intermountain Healthcare is a not-for-profit integrated health care delivery system located in Utah and Idaho that provides more than 50 percent of all care delivered in the region through providers in both urban and rural settings.
Where Current Triggers/TIDS Work Can Contribute Most
Determining High-Priority Areas to Target
The decision was made a priori to focus our work on developing triggers to identify preventable adverse events related to diagnosis (loss to followup) and treatment (medication, surgery) in outpatient settings. The Delphi panel of clinical experts was therefore asked to rate all ambulatory adverse events they felt were most important to target in these areas. Table 1 lists the ambulatory adverse events that had a median priority rating of 3 or lower. The panelists rated missed or delayed diagnoses as the highest priority area for targeting with triggers, followed by surgical adverse events and adverse drug events.
Determining How to Focus Adverse Drug Event Triggers
In ambulatory care, adverse drug events are commonC4 and represent a large portion of total adverse events.C5 A systematic review of the literature revealed that cardiovascular drugs, analgesics, and hypoglycemic agents accounted for 86.5 percent of preventable adverse drug events in ambulatory care.C4 However, the most frequently involved agents appear to vary by patient populationC6–C8 and detection setting.C7,C9 Thus, to determine which drugs were most important to target for trigger development, the Delphi panel was asked to rate the relative importance of the drugs specific to the ambulatory setting. The panel determined the following drug classes to be highest priority for trigger development: analgesics, cardiovascular drugs, hematologic and oncologic agents, antibiotics, neuropsychological drugs, and glucose controllers.
Table 1. Highest priority adverse events for trigger development
(1=highest priority, 9=lowest priority)
|Missed/delayed diagnoses/loss to followup
||Missed/delayed cancer diagnosis
||Missed/delayed myocardial infarction diagnosis
||Missed/delayed infection diagnosis
|Surgical adverse events
||Pulmonary embolism/deep vein thrombosis
||Postoperative respiratory failure
||Postoperative wound infection
||Postoperative cardiac event
|Adverse drug events
||Falls (without injury)
||Electrolyte imbalance/decreased renal function
More research is needed to characterize the causes and incidence of adverse events in outpatient settings. The variation observed in studies to date suggests that there is significant variation across settings and patient populations in the prevalence and causes of adverse events. Although adverse drug events are the best characterized to date, there is still much to gain from developing and improving triggers to detect and prevent them. Adverse events resulting from loss to followup and ambulatory surgery are critical areas for trigger development, as very little has been done in these areas thus far.
Relative Advantages of Various Trigger Types
Choosing the Appropriate Trigger System
The type of trigger system that is most appropriate depends on its intended purpose. An interventionist system is one that provides actionable notifications that can be used at the patient level to prevent or intervene in an adverse event. Other trigger systems might provide nonactionable notifications useful at the system level to generate feedback to clinicians or to design system change. The clinical specificity and timing of trigger systems should also be compatible with their intended purpose. For example, general or global trigger systems are helpful for identifying the types of adverse events that occur, and they can be used for surveillance and for guiding system-level interventions. However, they would not be useful for targeting specific adverse events or for patient-level intervention. Compared to general systems, specific trigger systems can be readily translated into patient- and system-level interventions because they are clinically specific regarding the cause of the adverse event. However, specific systems cannot identify all iatrogenic events, just those for which triggers are in place. Concurrent trigger systems can identify patients at high risk during the clinical episode in which the adverse event occurred to guide clinical intervention. Retrospective trigger systems are useful only for raising awareness, event rate measurement, and evaluating system-level interventions. The ability to run a trigger concurrently or retrospectively depends on data availability and ability to incorporate trigger response into workflow.
Evaluating Trigger Characteristics
Physicians, nurses, pharmacists, quality managers, and informaticists participated in focus groups to evaluate a number of specific triggers for clinical relevance, utility, and ease of implementation. Some consistent general themes emerged.
Trigger systems should target adverse events that are both prevalent and preventable. "Prevalence" included the size of the population at risk, how frequently the cases occurred within that population, and the frequency of cases that would benefit from the trigger. "Preventability" included whether the trigger was likely to be able to prevent harm that was avoidable (e.g., not address side effects that were a conscious tradeoff as part of treatment).
Trigger systems should fill a need and add value. Triggers should not duplicate existing quality, safety, or performance measures or information already being captured to meet internal or external requirements.
Trigger notifications should be actionable. Concurrent triggers should be implemented so that the data are received by individuals who can act on the information within the appropriate timeframe to prevent or mitigate harm.
Trigger systems should have a good "signal-to-noise" ratio and cost-benefit ratios. The degree and cost of harm to the patient that the trigger is designed to prevent should be weighed against the cost of implementing the trigger. Trigger systems should have good sensitivity and positive predictive value.
Trigger systems should be easy to implement. The data needed to run the trigger should be readily available in the necessary format. Ideally, the trigger should be integrated with existing computer systems and/or processes of care.
The type of trigger system that is most appropriate depends on its intended use. The perceived importance of trigger types and characteristics may depend also on the individual doing the evaluating. Our focus groups, which were heavily composed of front-line clinicians, placed a greater value on specific, concurrent, interventionist trigger systems that would allow for patient-level intervention. Test characteristics such as positive predictive value or sensitivity were not the only important criteria for evaluating trigger systems for potential implementation.
C1. Elder NC, Dovey SM. Classification of medical errors and preventable adverse events in primary care: a synthesis of the literature. J Fam Pract 2002 Nov;51(11):927-932.
C2. Gandhi TK, Weingart SN, Borus J, et al. Adverse drug events in ambulatory care. N Engl J Med 2003 Apr 17;348(16):1556-1564.
C3. Gandhi TK, Burstin HR, Cook EF, et al. Drug complications in outpatients. J Gen Intern Med 2000 Mar;15(3):149-154.
C4. Thomsen LA, Winterstein AG, Sondergaard B, et al. Systematic review of the incidence and characteristics of preventable adverse drug events in ambulatory care. Ann Pharmacother 2007 Sep;41(9):1411-1426.
C5. Plews-Ogan ML, Nadkarni MM, Forren S, et al. Patient safety in the ambulatory setting. A clinician-based approach. J Gen Intern Med 2004 Jul;19(7):719-725.
C6. Ferranti J, Horvath MM, Cozart H, et al. Reevaluating the safety profile of pediatrics: a comparison of computerized adverse drug event surveillance and voluntary reporting in the pediatric environment. Pediatrics 2008 May;121(5):e1201-1207.
C7. Budnitz DS, Shehab N, Kegler SR, et al. Medication use leading to emergency department visits for adverse drug events in older adults. Ann Intern Med 2007 Dec 4;147(11):755-765.
C8. Kanjanarat P, Winterstein AG, Johns TE, et al. Nature of preventable adverse drug events in hospitals: a literature review. Am J Health Syst Pharm 2003 Sep 1;60(17):1750-1759.
C9. Budnitz DS, Pollock DA, Weidenbach KN, et al. National surveillance of emergency department visits for outpatient adverse drug events. JAMA 2006 Oct 18;296(15):1858-1866.
a Boston University School of Public Health, Health Policy and Management Department, and VA Center for Health Quality, Outcomes and Economic Research (CHQOER).
b Suffolk University, Sawyer School of Business, and VA Center for Health Quality, Outcomes and Economic Research (CHQOER).
c VA Salt Lake City Geriatrics Research, Education, and Clinical Center (GRECC); Department of Medicine, University of Utah; and Intermountain Institute for Healthcare Delivery.
Note: The views in this article are those of the authors and do not necessarily represent the views of the Department of Veterans Affairs.
Previous Section Contents Next Section