The Health Quality & Safety Commission has facilitated a review of literature associated with the development and use of trigger tools to determine rates of harm in health care settings. Trigger tools can function either as a counting system that aims to estimate the rate of harm at an organisation or as an alerting system that aims to highlight the occurrence of a potential adverse event so that it can be mitigated.
This review of evidence focuses on the use of the global trigger tool (GTT) and related versions developed by the Institute of Healthcare Improvement (IHI) in the United States. The review was undertaken using a standard systematic review methodology. It includes a structured search for all published studies that have considered the IHI GTT and its related forms.
The key audience for the review is health professionals looking to use the IHI GTT to complement their other information sources about potential patient harm and to inform quality improvement projects. It is hoped that this is a useful document to assist with the implementation of the GTT methodology in New Zealand health care services.
The review was originally published in 2013 and updated in 2016. Both editions are available below.
Please email us if you would like a copy of this document in a different format.
Date of last publication
To review the literature associated with the development and use of trigger tools to determine rates of harm in health care settings with particular attention on the IHI GTT.
The review was undertaken using a standard systematic review methodology. It included a structured search for all published studies that have considered the IHI GTT and its related forms. All relevant studies are summarised and the main information is presented in tabular form. The key audience for the review is health professionals looking to use the IHI GTT to complement their other information sources about potential patient harm and to inform quality improvement projects. The main function of the review is to highlight the available literature. Only limited critical appraisal of the included studies is included. Instead, general comments are made about the limitations of the IHI GTT approach to measuring patient safety.
Detailed scope and methods for the review
The review describes all published studies including reviews of published studies addressing the IHI GTT and related trigger tools including versions designed for paediatric care, surgery, intensive care, ADEs and ambulatory care.
A systematic search was undertaken of the following electronic databases: MEDLINE and EMBASE. The databases were searched using a range of text keywords or Medical Subject Headings (MeSH) alone and in various combinations (trigger tool$, adverse event$, adverse drug event$, medication error$, adverse effect, detection system, surveillance, evaluation, review, screening, chart review, record review, incident report, voluntary report).
The search was undertaken in December 2012 and updated in February 2013. A further update was undertaken in July 2015. It was conducted without any limitations by language and it included all years from 1990 onwards. Studies in languages other than English were identified but not translated and were excluded from the review.
Further ‘snowball’ searching was undertaken of the reference lists of published studies.
A limited search of ‘grey’ literature was conducted. The search included important conference abstracts and key literature from relevant websites such as that belonging to the IHI.
After the abstracts were screened, all potentially relevant full-text publications were evaluated. Studies were included if they considered the use of a trigger tool system to identify patient harm and presented numeric data. The review focused on the use of the IHI GTT and any of its derivatives (specific tools for specialty areas such as paediatrics, mental health etc). It excluded studies that have not used all the stages of the GTT methodology (that is, sampling followed by screening for triggers and an assessment of whether an AE occurred). Therefore, for example, text mining studies that solely identified potential AEs but did not determine whether such an event had occurred were excluded as were studies that just assessed medical records in order to identify the presence of AEs without reference to the use of any triggers.
A structured template was used to extract relevant information from the included studies. This information included details about the study setting, sample, important methods, key results (such as ADE rate per 1000 inpatients) and authors’ conclusions.
Over 3200 potentially relevant studies were located by the searches. Some 148 studies were included in the review, after exclusions were applied for non-English language or out-of-scope reports.
A substantial number of studies have now been published that have used trigger tools including the IHI GTT to measure AE rates in health care organisations. From using these tools, it appears that AEs are common among inpatients including those in intensive care settings, occurring approximately 131 times per 1000 inpatients, 35 occasions per 100 admissions or among 29% of admissions. Most hospital events are relatively minor, and between 36% and 72% may be preventable. Adverse drug event (ADE) rates vary considerably when assessed by means of the tools but may be as high as 31% of admissions or 46 per 1000 inpatient days.
As there is no true gold standard, the accuracy of trigger tools cannot be reliably ascertained. The use of medical record review as the standard trigger tool appears to be an accurate method to detect iatrogenic harm with high sensitivity and specificity reported in some but not all studies. The tool also appears to be an efficient method to detect harm with high positive predictive values (PPVs) recorded in some studies. Assessments of the reliability of the tools suggest that there is moderate agreement amongst reviewers in their assessments of the occurrence of AEs. Limitations associated with this level of agreement may impact on the ability of the tool to reliably detect changes in patient outcomes at an organisation over time. Trigger tools are the best single method to detect harm and appear considerably more effective and cost-effective than voluntary reporting and pharmacist review to detect AEs. However, it seems likely that trigger tools also identify different types of harm compared with these methods and a comprehensive review of patient safety in an organisation should adopt multiple methods. Most experience with trigger tools has occurred in relation to ADEs, while experience is accumulating with intensive care and surgical patients. Recently, trigger tools have been applied in primary care and as part of quality improvement initiatives.
Conclusions: Trigger tools, particularly the IHI GTT, assist organisations to measure and monitor harm. They appear to be the most accurate and efficient method to identify AEs. Further work is needed to assess their reliability and validity. Trigger tools are most effective when combined with other measures and patient safety interventions in the reduction of iatrogenic harm.