Triggers and Targeted Injury Detection Systems (TIDS)
Expert Panel Meeting
Conference Summary Report
On June 30 and July 1, 2008, Agency for Healthcare Research and Quality (AHRQ) convened a meeting to develop and support the implementation of clinical triggers and targeted injury detection systems (TIDS) to identify patient safety risks and hazards.
Select for print version (PDF File, 933 KB).
Purpose of the Meeting
AHRQ is developing and supporting the implementation of clinical triggers and targeted injury detection systems (TIDS) to identify patient safety risks and hazards. Clinical triggers are data flags identifying patients who may be at risk of harm or clinical situations that have the potential for harm. Prospective trigger tools provide rapid, real-time identification of adverse events and enable timely intervention that can prevent or mitigate an adverse event. In contrast, retrospective trigger tools allow identification of risky situations after the clinical episodes have terminated and are most suitable for measurement of event rates.
With many of AHRQ's initial research projects relating to triggers and TIDS coming to a close, AHRQ decided to convene a meeting of trigger and TIDS researchers and Federal and private stakeholders to:
- Review progress on trigger/TIDS development.
- Identify (or explore) options for future work in this area.
Meeting Structure and Contents of This Publication
On June 30 and July 1, 2008, AHRQ convened a meeting to achieve the above-named goals. Forty-three researchers and key stakeholders participated in this meeting. In order to provide focus to the discussion, two panel presentations were held: one panel examined the methodological issues arising during triggers/TIDS development; the second panel explored issues surrounding the implementation of triggers/TIDS systems. These panel presentations were followed by extensive small-group discussion of issues raised during the presentations, as well as discussion of action options for assuring successful implementation and sustainable outcomes for future triggers/TIDS work.
This publication summarizes the key points raised at this meeting and contains:
- A literature review, prepared for this meeting, summarizing published work to date in the trigger/TIDS research domain.
- Seven brief articles submitted by the panelists, summarizing their research and implementation work to date.
- A glossary, also prepared specifically for this meeting, to ensure a common understanding of trigger-related terms.
Key aspects of each of these items are briefly summarized below.
The literature review (Section IIA) notes that the majority of work published to date on triggers is related to adverse drug events. These publications on triggers for adverse drug events display a wide variety in the amount of detail and type of data used to detect the adverse event. A less common type of trigger, with significant presence in the literature, relates to surgical adverse events. In this review, Mull and colleagues note that adverse events related to operations are both prevalent and costly. Another significant category of triggers identified in the review consists of triggers linked to causes of adverse events. The review notes a gap in the literature of work surrounding triggers in the ambulatory setting (including outpatient surgery), triggers for diagnostic errors, and circumstances of failure to follow up.
The methods panel noted that positive predictive value (PPV) is a commonly used measure for determining the value of a trigger system, but PPV is limited because it is dependent on prevalence. Other methods such as sensitivity can be used, but this requires large numbers of patients to be reviewed; therefore, the panelists felt that the field would benefit from the development of new methods.
The importance of developing triggers/TIDS that are appropriate for their intended purpose was also emphasized. It was specifically noted that trigger methods for diagnostic events and ambulatory adverse events are underdeveloped.
Panelists noted that it is critical to develop triggers/TIDS that are integrated into workflow; often this will require an integrated, advanced electronic health record. Paper-based trigger systems can be very informative in identifying adverse events, but generally, paper-based systems require significant staff training and labor-intensive, manual chart review.
Development of real-time, concurrent trigger/TIDS systems for both inpatient and ambulatory care would significantly advance the field.
Focus group work presented during the methods panel revealed clinicians are most interested in triggers that are associated with frequently occurring, critical, and preventable adverse events. The clinicians represented in the research focus group assigned the most value to actionable triggers.
Themes surrounding the implementation of triggers/TIDS were identified by the second group of panelists. In AHRQ-funded contracts, inpatient adverse drug event and pressure ulcer algorithms have proved useful for identifying adverse events and hold potential value for mitigating harm. The panelists noted that generally, clinicians perceived an incremental benefit from the triggers/TIDS systems when the systems filled a gap in clinical knowledge. Additionally, implementation of triggers/TIDS tended to be more successful when a multidisciplinary team approach was used. Like the methods panel, the implementation panel noted that implementation of paper-based systems is often cumbersome due to the large number of charts needing to be reviewed. Some participants in paper-based systems found it difficult to find significant clinical value when weighed against the efforts of implementation. Therefore, automated systems were identified as critical to long-term successful implementation across many health care organizations.
Although automated systems may be costly to develop, one panelist demonstrated methods to develop a clear business case for preventing adverse events based upon hospital discharge cost data and the institution's harm prevalence rates. The panelists noted that harm sustained because of adverse events during hospitalizations results in significant costs to the organization, with impact on reimbursement and opportunity costs. A second approach identified to aid with the business case development for triggers/TIDS was integration of trigger tools into existing quality processes within organizations.
Glossary of Terms
Both earlier researchers and attendees at the expert meeting identified a need for definitional alignment of trigger/TIDS elements. This publication contains a glossary of trigger/TIDS terms (Section III) that both clinicians and researchers will find valuable. The glossary was provided to the attendees for the purpose of advancing the discussions.
AHRQ would like to acknowledge the contributing experts who have provided this framework, who are listed in Section IV.
A number of future options for AHRQ to consider emerged during the small-group discussions. These included:
- Continued development of triggers/TIDS systems that match the intended purpose of use (in terms of logic, method of evaluation, etc.). Systems intended for ranking or rate estimation functions must be evaluated very stringently.
- Continued research to explore methods of integrating trigger/TIDS harm identification and mitigation into existing health information technology and into existing workflow. Implementation of the trigger tools is not easy and may require guidance and support from entities external to the implementing organization.
- Support of development and spread of the business case for triggers/TIDS and for systems that mitigate harm.
- Continued dialog among Federal agencies likely to be interested in triggers/TIDS and patient safety monitoring.
- New grant or contract funding for triggers/TIDS development on outpatient surgery, care transitions, long-term care, and ambulatory settings.
- Development of standards for definitions of triggers/TIDS variables to facilitate easier adoption and promotion of uniformity.
- Support for collaborative organizations seeking to apply triggers/TIDS.
- Support for demonstrations and research on trigger/TIDS implementation, with distribution of lessons learned to systems interested in implementation.
- Development of a standardized implementation tool for triggers/TIDS.
A final observation is perhaps in order regarding further work on trigger/TIDS development. No clear definition of triggers/TIDS emerged that clearly delineated a mechanism or mechanisms substantially different from current approaches found in either:
- Real-time alerts and reminders embedded in electronic health records.
- Retrospective chart review conducted using either explicit "triggers" (measures) or subjective analysis.
Further work may be needed to be define a unique and cost-effective role for new triggers/TIDS mechanisms.