Presented at the Global Health Institute Annual Symposium in Madison, WI, 2017
Kathleen K. Miller, MD; Amanda C. Becker, MD; Erin Chung; Peter Oluput-Oluput
Background: Appropriate triage and recognition of danger/ priority signs in critically ill children during the first 24 hours of admission is necessary to reduce child mortality, particularly in settings with limited resources. We hypothesized that a quality improvement project at Mbale Regional Referral Hospital (MRRH) to improve provider recognition of danger signs would result in increased documentation of danger signs and indirectly decrease mortality.
Methods: An educational refresher course was held for members of the medical team reviewing paediatric “danger signs” and effective triage. Handouts were given to all participants and posters were placed in the triage area and acute care ward listing defined danger signs. A retrospective chart review was performed to identify documentation of danger signs on admission to the pediatric acute care ward at MRRH for the 2 week period prior to the educational intervention. Data points included the number of danger signs, patient outcome, diagnosis, and length of stay. An additional retrospective chart review assessing the same data points was performed for the 2-week period after the intervention.
Results: 25 clinical staff attended the refresher training. 210 charts were reviewed prior to the intervention and 179 postintervention. The mortality rate in the pre-intervention group was 10%; of these patients, 76% died within the first 24 hours of admission. The average number of positive danger signs identified was 1.0 among patients who survived and 1.8 among patients who died, which was statistically significant (pvalue 0.0015). There was a statistically significant association between the presence of positive danger signs and death (pvalue 0.0048). In the post-intervention group, 179 charts were reviewed in the two weeks following the intervention. 8% of patients had zero danger signs identified. The average number of danger signs identified among patients who survived was 1.7, and 2.7 among those who died (p-value 0.0196). There was a statistically significant increase in the average number of danger signs documented in the pre-vs post intervention group, which averaged of 1.0 in the pre- group and 1.7 in the post-intervention group. Mortality decreased from 10% to 6%, but this was not statistically significant (p-value 0.058).
Conclusions: There was a statistically significant increase in the number of danger signs identified on admission after the educational intervention and placement of visual reminders in the work environment. There was a difference in mortality, although not statistically significant, which may have been related to the number of patients in the study. Refresher trainings for identification of pediatric danger signs lead to increased identification and documentation of danger signs, which could be associated with a decreased mortality over time.