Abstract
OBJECTIVES: This study aimed to identify the prevalence of and factors associated with diagnostic uncertainty when critically ill children are admitted to the PICU. Understanding diagnostic uncertainty is necessary to develop effective strategies to reduce diagnostic errors in the PICU. DESIGN: Multicenter retrospective cohort study with structured medical record review by trained clinicians using a standardized instrument to identify diagnostic uncertainty in narrative clinical notes. Diagnoses and diagnostic uncertainty were compared across time from PICU admission to hospital discharge. Generalized linear mixed models were used to determine patient, clinician, and encounter characteristics associated with diagnostic uncertainty at PICU admission. SETTING: Four academic tertiary-referral PICUs. PATIENTS: Eight hundred eighty-two randomly selected patients 0-18 years old who were nonelectively admitted to participating PICUs. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: PICU admission notes for 228 of 882 patients (25.9%) indicated diagnostic uncertainty. Patients with uncertainty decreased over time but 58 (6.6%) had remaining diagnostic uncertainty at hospital discharge. Multivariable analysis showed that diagnostic uncertainty was significantly associated with off hours admission (odds ratio [OR], 1.52; p = 0.037), greater severity of illness (OR, 1.04; p = 0.025), an atypical presentation (OR, 2.14; p = 0.046), diagnostic discordance at admission between attending intensivists and resident physicians/advanced practice providers (OR, 3.62; p < 0.001), and having a neurologic primary diagnosis (OR, 1.87; p = 0.03). Older patients (OR, 0.96; p = 0.014) and those with a respiratory (OR, 0.58; p = 0.009) or trauma primary diagnosis (OR, 0.08; p < 0.001) were less likely to have diagnostic uncertainty. There were no significant associations between diagnostic uncertainty and attending intensivists' characteristics. CONCLUSIONS: Diagnostic uncertainty at PICU admission was common and was associated with off hours admission, severe illness, atypical presentation, diagnostic discordance between clinicians, and a neurologic primary diagnosis. Further study on the recognition and management of diagnostic uncertainty is needed to inform interventions to improve diagnosis among critically ill children.