A multicenter observational study investigating care errors, staffing levels, and workload in small animal intensive care units.

Objective: To investigate associations among care errors, staffing, and workload in small animal ICUs.

Design: Multicenter observational cohort study conducted between January 2017 and September 2018.

Setting: Three small animal teaching hospital ICUs.

Animals: None.

Interventions: None.

Measurements And Main Results: Data on patient numbers, illness severity (assesed via the acute patient physiologic and laboratory evaluation [APPLE] score), care burden, staffing levels, technician experience/education level, and care errors were collected at each study site. Care errors were categorized as major (unanticipated arrest or death; patient endangerment through IV line, arterial catheter, chest tube or other invasive device mismanagement, or errors in drug calculation/administration) or minor. Median patient:technician ratio was 4.3 (range: 1-18). Median patient illness severity was 15.1 (4.7-27.1) APPLE score units. A total of 221 major and 3,317 minor errors were observed over the study period. The odds of a major error increased by an average of 11% (odds ratio [OR] = 1.11; 95% confidence interval [CI], 1.02-1.20; P = 0.012) for each 1 patient increase in the patient:technician ratio after averaging by ICU location. The major error incident rate ratio was 2.53 (95% CI, 1.84-3.54; P < 0.001) for patient:technician ratios of >4.0 compared with ≤4.0. The odds of a major error increased by 0.5% per total unit APPLE score increase (OR = 1.005; 95% CI, 1.002-1.007; P < 0.001). The major error incident rate ratio was 1.71 (95% CI, 1.30-2.25; P < 0.001) for APPLE :technician ratios of >73 compared with ≤73. The odds of a major error decreased by 2% (OR = 0.98; 95% CI, 0.97-0.99; P = 0.01) for each year increase in total technician years of ICU work experience.

Conclusions: Substantial reductions in major care errors may be achieved by maintaining ICU patient:technician ratios at ≤4. Technician experience and total unit burden of patient illness severity are also associated with error incidence, and should be taken into consideration when scheduling staff.