Reflections on the Institute for Healthcare Improvement (IHI) Global Trigger Tool
Roger Resar, M.D.a
Experience With the IHI Global Trigger Tool
The Institute for Healthcare Improvement (IHI) Global Trigger Tool (GTT) is based on 10 years of trigger tool component testing and experience in the acute inpatient setting,D1–D3 resulting in a clearly articulated and described methodology.D4 The current use of the GTT is to establish a baseline level of harm (adverse events) in an organization and then, using statistical process control rules, collect data points over time to determine improvement. In some organizations, an evaluation of actual adverse events allows targeting of improvement strategies at a fairly high level (postoperative infections, anticoagulant-related events, etc.). The GTT is the only method available to quantify harm at a system level in a reproducible fashion in any hospital system, whether paper based or electronically endowed. The combining of triggers to a robust, tested, and clearly articulated methodology is crucial to creating the reproducible findings. A two-step process, using primary reviewers whose consensus is validated by a physician reviewer, is required initially on prepared training records. Measured kappa scores for testing and training demonstrate high inter-rater agreement.D4
Without a believable acknowledgment of the level of harm created by health care, there is a lack of incentive to change and there is no way to determine whether the resources being devoted to safety are having any sustained effect. The harm surfaced (revealed) by the GTT, by definition, removes from consideration preventability (very subjective and would result in an ever-changing denominator), omissions (data almost always missing from the single chart review and very accusatory), error (harm looked at from the patient perspective, regardless of whether an error can be documented), and, for the most part, any considerations of cause. The harm identified is carefully tied to commission and represents a majority subset, but not all, of the totality of institutional harm. (Ninety percent or more of adverse events identified in retrospective traditional chart review are surfaced by the GTT.)D5
Evidence from hundreds of organizations using the GTT, as well as coordinated implementation directed by IHI faculty, have established baseline "ballpark" expectations of harm for organizations. As a caveat, these are treated as a general evaluation of the review process rather than as a benchmark. After some initial variability (we usually discount the first four to six data points as learning opportunities), most organizations report about 90 adverse events per 1,000 patient days, 40 adverse events per 100 admissions; about 30 percent of all admissions (using GTT criteria for definition of an admission) experience an adverse event.
The IHI GTT generates reproducible results due to a highly structured methodology of training, testing, and reviewing of charts and is utilized by organizations worldwide. The results have unfortunately provided evidence of even more harm than previously reported.
Triggers are built to surface harm. Harm can be identified in the construct of the GTT in multiple ways. These include, of course, the triggers themselves (some are, by definition, adverse events); triggers that require further chart investigation to determine if harm has occurred (glucose of 50 is harm only if symptoms of hypoglycemia are present); harm identified using the methodology of GTT chart review (reviewing the discharge codes and diagnoses without any use of triggers); and just plain luck while paging through the chart.
The harm surfaced with the GTT commonly is associated with a trigger but does not necessarily require a trigger. (Eighty percent of harm surfaced has an associated trigger; 20 percent does not.) A given harm may well have multiple triggers, and reviewers commonly will have identified different triggers to surface the same harm. Although it is tempting to be overly academic about the triggers chosen to review charts, at least in the retrospective view and with the methodology established by the IHI GTT, the specificity and sensitivity of the triggers are secondary.
The most important factor appears to be the ability to evaluate the trigger (is it easy to find in the record and easy to make a determination of harm?) and the total number of triggers (the greater the number, the more time it will take to review the chart). The IHI has by definition limited the time to retrospectively review a chart to 20 minutes, which forces a limitation on the number of triggers and the way in which the chart is reviewed. Starting from page one and continuing to page zed is not an option, particularly with very thick (long stay) charts. The primary concern with the IHI GTT triggers is to accomplish a quick review of key inpatient modules within an appropriate timeframe to allow the sampling methodology to be robust.
Triggers for the IHI GTT are not based on an evaluation grounded in specificity and sensitivity, but rather on an empiric model. Since there are multiple ways to surface harms in this retrospective methodology, the sensitivity and specificity of the triggers are actually secondary both to the sampling and to the robustly structured methodology.
D1. Resar RK, Rozich JD, Classen D. Methodology and rationale for the measurement of harm with trigger tools. Qual Saf Health Care 2003;12 Suppl 2:ii39-45.
D2. Rozich JD, Haraden CR, Resar RK. Adverse drug event trigger tool: a practical methodology for measuring medication related harm. Qual Saf Health Care 2003;12(3):194-200.
D3. Resar RK, Simmonds T, Haraden CR. A trigger tool to identify adverse events in the intensive care unit. Jt Comm J Qual Patient Saf 2006;32(10):585-590.
D4. Classen DC, Lloyd RC, Provost L, et al. Development and evaluation of the IHI Global Trigger Tool. Accepted for publication June 2008 in the Journal for Patient Safety.
D5. James B. Vice President—Medical Research and Continuing Medical Education, Intermountain Healthcare. Personal communication. Dec. 12, 2007.
aInstitute for Healthcare Improvement.
Previous Section Contents Next Section