Pediatric Intensive Care Nurse Staffing Measures and Patient Outcomes During the COVID-19 Pandemic

IMPORTANCE During the COVID-19 pandemic, pediatric intensive care units (PICUs) reportedly relied more on temporary contract nurses hired through staffing agencies (agency nurses) to fill gaps related to increased nurse turnover. It is unknown whether these changes in nurse staffing were associated with patient outcomes. OBJECTIVE To determine whether the proportion of PICU agency staff or nurse turnover was associated with patient outcomes during the pandemic. DESIGN, SETTING, AND PARTICIPANTS This retrospective, multicenter, linked database cross-sectional study was performed at PICUs in the US between January 1, 2019, and December 31, 2022. Participants included patients younger than 19 years admitted to a PICU with data in both the Children’s Hospital Association Pediatric Health Information Systems and PROSPECT Hospital Essentials databases. Data were analyzed from March 15, 2023, to August 1, 2024. EXPOSURE The yearly proportion of agency nurses and nurse turnover. MAIN OUTCOMES AND MEASURES Descriptive characteristics were examined for PICUs grouped by the median proportion of agency nurses or nursing turnover each year. A variable staffing metric was defined as greater than 10% change in the proportion of agency nurses from the intrahospital baseline in 2019. Multivariable negative binomial models controlling for illness severity and hospital effect with median yearly staffing metrics examined whether the proportion of agency nurses or nurse turnover was associated with mortality, PICU complications, central line–associated bloodstream infections, or ventilator-associated pneumonias. RESULTS There were 218 789 admissions across 20 PICUs, with a median patient age of 45 (IQR, 9-138) months (119 997 [54.8%] male). PICUs with a variable proportion of agency nurses had significantly more patients with PICU complications (9695 of 120 017 [8.1%] vs 7141 of 98 772 [7.2%]; P < .001). Multivariable analysis indicated an increase of 19 (95% CI, 7-39) ventilator-associated pneumonias per 10 000 ventilator-days in 2020. For a PICU with 10% higher nurse turnover than the median turnover, there were 10 (95% CI, 8-13) more PICU complications per 10 000 patient-days in 2022. CONCLUSIONS AND RELEVANCE In this cross-sectional study of nurse staffing measures and patient outcomes during COVID-19, complex associations existed throughout the study years. More data are needed to determine optimal staffing strategies during times of workforce strain.