Background: Contaminated pharmaceutical products pose serious infection risks to patients and can lead to significant morbidity and mortality. Contamination at the point of manufacturing or compounding (intrinsic contamination) has the potential to affect large numbers of patients. Public health plays a critical role in detecting and investigating such events. We identified investigations involving intrinsically contaminated pharmaceuticals to characterize the burden and scope of harm associated with these events. Methods: We reviewed Centers for Disease Control and Prevention records to identify US investigations between January 1, 2009, and December 31, 2018, involving laboratory-confirmed contamination of manufactured medications and pharmacy-compounded preparations (P-CPs), using relevant search terms (eg, “medication contamination”). Laboratory confirmation was defined as identification of a pathogen from a manufactured medication or P-CP. We determined the number and type of patient infections associated with these investigations, the number of states involved, pathogens identified, type of medication (sterile or nonsterile), route of administration, and how the contamination event was first identified. We excluded investigations when the mode of production was unknown. Results: We identified 20 investigations in at least 20 states involving laboratory-confirmed contamination of manufactured medications (n = 12) and P-CPs (n = 8). Patient infections were identified in 16 (80%) investigations (9 involving manufactured medications and 7 involving P-CPs) resulting in at least 1,183 infections and at least 73 deaths. Bloodstream infections were the most common infection type (n = 7, 44%). Waterborne pathogens (eg, Serratia marcescens, Burkholderia cepacia) were cultured from medications in 83% (n = 10) of investigations involving manufactured medications and 75% (n = 6) of investigations involving P-CPs. Contamination of sterile pharmaceutical products occurred in 14 (70%) investigations; 11 (79%) of these involved injectables. Information regarding how contaminated pharmaceuticals were first identified was documented for 18 investigations; most cases (n = 14, 78%) started with investigation of patient infections by facilities, public health, or both, which led to laboratory testing of pharmaceuticals and confirmation of contamination. Conclusions: The events summarized here likely underestimate the frequency of intrinsic contamination of pharmaceutical products in the United States. These events can have devastating consequences that impact patients across the country. Waterborne pathogens appear to be the most frequently identified source of contamination in both manufactured medications and P-CPs.
Detection, investigation, control, and prevention of pharmaceutical contamination events benefit from collaboration between state and federal public health authorities; without public health intervention. Such contamination may have gone undetected and could have harmed additional patients.