Antibiotics in a Haitian neonatal intensive care unit: findings from a Plan-Do-Study-Act quality-improvement project

Abstract

BackgroundIn the critical care setting, delay in the administration of antibiotics has been associated with as much as a three-fold rise in mortality. Working in a neonatal intensive care unit (NICU) in rural Haiti, we attempted to reduce the time between patient admission and the administration of the first dose of antibiotics and to reduce the frequency of dosing errors by implementing a Plan-Do-Study-Act (PDSA) cycle-based quality improvement project. MethodsThe study took place at St Boniface Hospital—a 130 bed hospital serving 2·3 million people in rural Haiti. A retrospective chart review was done for every patient admitted to the 24-bed NICU over a 2-week period. Baseline data were recorded by noting the time of patient admission and the time of administration of the first doses of antibiotics. In addition, the dosage of each antibiotic was checked for accuracy. Two subsequent 2-week data collections were performed over 6 months, interspersed with PDSA cycles. These cycles included training and interventions to: improve staff awareness about the risks of delaying treatment and of incorrect dosing; identify barriers to care; introduce nurse-initiated standing orders; review data; and disseminate results to the staff. This study was approved by the Research Ethics Committee at St Boniface Hospital. FindingsThe baseline data collection between Jan 29, 2018, and Feb 11, 2018, included 26 patients; period 2 (April 25, 2018 to May 8, 2018), included 33 patients; and period 3 (July 5–18, 2018), 28 patients. At baseline, mean delay in antibiotic administration was 6·2 h (95% CI 4–8·4). In periods 2 and 3, mean delays were 0·6 h (0·35–0·85) and 0·4 h (0·16–0·64), respectively. At baseline, dosing errors were noted in 13 of 22 ampicillin orders (59%) and four of 22 gentamicin orders (18%). Four patients received antibiotics other than ampicillin or gentamicin. In periods 2 and 3 there were no errors reported. In the baseline period, two of 26 patients (7·6%) received the first dose of antibiotic within 1 h and in periods 2 and 3, 30/33 (91%) and 24/28 patients (86%), respectively, received antibiotics within 1 h. InterpretationIn a low-resource setting, errors in antibiotic dosing and delays in administration of antibiotics to newborns admitted to the NICU can be successfully reduced using a PDSA-based quality improvement project. This approach does not require expensive interventions nor new technology but it does have the potential to reduce neonatal morbidity and mortality. FundingNone.