Unintentional use of mistaken insulin products: a retrospective review of poison center data.

BACKGROUND: There are no guidelines for the management of accidental insulin administration. We hypothesized that home monitoring of asymptomatic patients (pts) was safe following unintentionally insulin administration. METHODS: Retrospective review of poison center (PCC) charts from 1/1/2000-12/31/2010 looking for accidental insulin administrations. INCLUSION CRITERIA: pt must be prescribed insulin. Information recorded from charts: pt age/gender, "intended" and "mistaken" insulin formulations/doses, use of oral diabetic agents, management site, Emergency Department (ED) referral, symptoms, blood glucose values, and treatments. Defined outcomes: symptoms (e.g., altered sensorium); hypoglycemia (<60 mg/dL); management site; health care facility (HCF) admission; and death. Multiple logistic regression was used to determine outcome predictors. RESULTS: 652 charts met inclusion criteria. Mean age was 56.4 years; most (58.5%) were women. Most (89%) calls originated from home, 10.7% from a HCF, 0.3% from Emergency Medical Services (EMS). Overall, 397 (60.9%) pts were managed at home. Two pts managed at home were later evaluated by EMS; neither required admission. Symptoms developed in 56 (8.6%) pts. There were no deaths. Only 40 (6.1%) pts were admitted to a HCF; 18 (45%) pts were hypoglycemic. The development of hypoglycemia (odds ratio [OR] 5.94; p < 0.001) and amount of insulin accidentally administered (OR 1.04; p < 0.001) predicted HCF referral. The type and dose of insulin administered did not predict symptoms. CONCLUSIONS: Based on a retrospective analysis of a single PCC's cases, home observation of asymptomatic patients after unintentional administration of a wrong insulin formulation appears safe.