Session number: 1006
Abstract title: Adverse Drug Events Leading to Hospital Admissions
Author(s):
JM Hoffman - VA Salt Lake City GRECC
JF Jurdle - VA Salt Lake City GRECC; University of Utah
JN Nebeker - VA Salt Lake City GRECC; University of Utah
CR Weir - VA Salt Lake City GRECC; University of Utah
BS Roth - VA Salt Lake City GRECC
Objectives: Past research has been successful in demonstrating that adverse drug events (ADEs) cause a substantial number of hospital admissions. One meta-analysis reported an aggregate of 5% of hospital admissions due to ADEs. Establishing the number of admissions caused by preventable ADEs is of further importance. Our objectives are 1) to assess the rate of ADEs leading to hospitalizations, and 2) determine the characteristics of the set of preventable ADEs.
Methods: This prospective study is part of a larger investigation assessing a random sample of 937 admissions (40% of all admissions) over a period of 4.5 months in an acute-care hospital. Two clinical pharmacists conducted a comprehensive case review. Validation and characterization was conducted by a panel of two MDs, two PharmDs and one RN recording the drug(s) responsible, causality, medical severity, pharmacologic type, and error, among other variables.
Results: Twenty-five percent (233/937) of admissions were due in whole or in part to an ADE. When using a conservative preventability definition, 122 of the 233 were judged to be preventable (13% of all admissions studied). Causes of preventable ADEs included non-compliance (62%), monitoring error (45%), and improper dose (33%) [not mutually exclusive]. Warfarin, insulin, and olanzapine were the leading medications involved in non-compliance. There were also significant differences in syndromes and drug classes when categorized by error. Counterintuitively, age < 75 years old was significantly associated with non-compliance leading to admission (9.0% vs. 6.0%, p<0.05). Resource utilization ascribed solely to preventable ADEs (excluding the standard resource utilization associated with admission itself) included 133 ADEs in which additional labs were ordered, 89 additional diagnostic procedures, and 94 follow-up appointments.
Conclusions: In this VA population, ADEs are responsible in whole or in part for 25% of admissions, which is much higher than reported in the literature. Thirteen percent of the ADEs in our study were preventable and the majority of the preventable ADEs were secondary to non-compliance. Preventable ADEs lead to extensive resource utilization during hospitalization.
Impact statement: Substantial reduction in patient morbidity and system cost could be realized if admissions due to preventable ADEs were decreased with analysis and intervention in the outpatient setting.