Prospective identification versus administrative coding of adverse drug reaction-related hospitalizations in the elderly: A comparative analysis

Purpose To compare prospective identification of adverse drug reaction (ADR)-related hospital admissions in the elderly with administrative coding using the International Classification of Diseases 10(th) Revision Australian Modification (ICD-10-AM) coding system. Methods Results We linked the records of 768 enrolled patients from an earlier study, where clinical pharmacists identified ADRs using prospective data collection, to hospital administrative data. We identified patients in the study whose admissions were coded as ADRs using ICD-10-AM codes. We then compared the prevalence and characteristics of ADR-related hospital admissions identified by the two approaches. According to ICD-10-AM coding, 2.7% of patients were admitted due to ADRs, while 15.0% of patients were deemed to have been admitted due to ADRs based on prospective identification by clinical pharmacists. Most (85.7%) patients coded as having an ADR-related hospital admission were also identified as such prospectively. Hematological (23.1%) and metabolic reactions (23.1%) were frequent causes of ADRs identified by coding, whereas cardiovascular ADRs (27.8%) were more common causes of ADRs identified prospectively by pharmacists. Antidepressants (16.7%) and cardiac glycosides (16.7%) were the most commonly implicated drug groups in ADRs identified by coding, whereas diuretics (28.8%) and renin-angiotensin system inhibitors (17.0%) were frequently implicated as causes of ADRs identified prospectively by pharmacists. Conclusions Reliance on administrative coding potentially underestimates the extent of the problem of ADRs as a cause of hospitalization in the elderly, and more detailed prospective analysis of admissions provides additional targets for strategies to prevent ADRs. The types of ADRs identified also differ between the two approaches.