Clinical question: Can a trigger tool identify harms for hospitalized children?
Background: An estimated 400,000 people die annually in the United States as a result of hospital-associated harm. The Centers for Medicare and Medicaid Services define harm as “unintended physical injury … by medical care that required additional monitoring, treatment, or hospitalization or that resulted in death.” Although harm is common, voluntary reporting of events has been shown to capture only 2%-8% of harm. Global Trigger Tools (GTT) are an alternative to voluntary reports. These tools use “triggers,” or clues, to help reviewers identify potential harms when reviewing the electronic heath record. The Institute for Healthcare Improvement (IHI) has created an adult-focused GTT; however, no pediatric-focused GTT exists.
Study design: Cross-sectional, retrospective chart review.
Setting: Children <22 years old discharged from six freestanding U.S. children’s hospitals in February 2012.
Synopsis: In a prior paper, the authors described how they used a modified Delphi technique to develop a pediatric GTT based upon the IHI GTT. Here they piloted this new pediatric-focused GTT through a retrospective chart review. One clinical nonphysician reviewer and one physician reviewer were selected from each site and received training on use of the pediatric GTT and the identification of harms. One hundred charts from each site were randomly selected for application of the GTT. The reviewers examined the charts for harms and then applied the GTT. When reviewers found a harm, they determined the likelihood that the harm was preventable.
Of the 600 records reviewed, 240 harms were found. The GTT identified 1,093 potential harms, leading to identification of 204 harms. The remaining 36 harms did not cause a trigger and were found by chart review. The positive predictive value of the aggregate GTT was 22%. There were 40 harms per 100 patients, and 24.3% of patients had one or more harm. Sixty-eight percent of harms were of the least severe type, and only one led to a patient death. The most common harms were intravenous catheter infiltration, respiratory distress, constipation, pain, and surgical complications.
Bottom line: The pediatric GTT appears to be a moderately sensitive indicator for inpatient pediatric harm. Inpatient pediatric harm occurs frequently, with about one in four pediatric inpatients suffering from harm. Serious harm appears uncommon.
Citation: Stockwell DC, Bisarya H, Classen DC, et al. A trigger tool to detect harm in pediatric inpatient settings. Pediatrics. 2015;135(6):1036-1042.
Dr. Stubblefield is a pediatric hospitalist at Nemours/Alfred I. Dupont Hospital for Children in Wilmington, Del., and assistant professor of pediatrics at Thomas Jefferson Medical College in Philadelphia.