Abstract and Introduction
Abstract
Purpose: The likelihood of hospitalization caused by adverse drug reactions (ADRs) from commonly implicated therapeutic groups is discussed.
Methods: A retrospective analysis of the computerized records of exposure cases involving pharmaceutical substances reported to the New Jersey Poison Information and Education System (NJPIES) was conducted from 2000 through 2007. The cases in the National Poisoning Data System that were categorized as an ADR were included in the study set. Only reports involving a single drug were selected for inclusion in the analyses. Characteristics of the ADRs, such as the sex and age of the patient, the therapeutic group involved, and the medical outcome of the exposure, were examined. Reports of ADRs with the most frequently implicated therapeutic groups were analyzed based on whether the patients were managed onsite, referred to a health care facility, or managed at a health care facility. The Adverse Drug Reaction Hospitalization (ADRH) index was calculated for all therapeutic groups, but the focus of the analyses was on the groups that were implicated in 5% or more of all ADRs.
Results: A total of 454,520 cases of human poisoning exposure were reported to NJPIES from 2000 through 2007. Of these cases, 162,105 were exposures implicating a single drug, of which 5,461 (3.4%) were classified as an ADR. Of the 5,461 cases, 385 patients were admitted into a health care facility. Antidepressants had the highest ADRH index (20.4%) among the therapeutic groups implicated, and antimicrobials had the lowest (2.2%).
Conclusion: The analyses revealed a substantial variation in the likelihood of hospitalization associated with ADRs within different therapeutic groups. Among the groups that were most frequently implicated in ADRs, antidepressants showed the highest probability for an ADR-related hospitalization, followed by dietary supplements, herbals, and homeopathics and then by sedatives, hypnotics, and antipsychotics.
Introduction
Adverse drug reactions (ADRs) present a serious public health problem that can affect patients, caregivers, pharmaceutical companies, and the health care system as a whole. The systematic collection and analyses of postmarketing data concerning the frequency, severity, and types of ADRs are of particular importance. Premarketing clinical trials do not have the power to demonstrate an association between a drug and rare adverse events, which may be observed postmarketing when the drug is used by a larger number of individuals. Furthermore, the pre-approval clinical efficacy and safety studies cannot account for ADRs associated with unlabeled use of drugs after they become available on the market. Accordingly, considerable efforts are put into postmarketing surveillance activities, which in essence are driven by safety concerns and designed to identify rare events or unexpected occurrences of known ADRs.
Postmarketing safety studies usually collect drug-specific information on ADR occurrences that permits better estimation of the benefit–risk profile of individual drugs. To get a broader perspective about the overall burden associated with drug-specific ADRs, however, consideration should be given not only to the frequency and types of such events but also to their potential to further affect both patients and the health care system, particularly if they involve hospitalization. Published data regarding these issues are scarce. In particular, even as ADRs often result in hospital admissions, knowledge about the likelihood of drug-specific ADRs to prompt hospitalization is limited.
While many studies have assessed the morbidity and mortality associated with ADRs, very few have focused on evaluating the potential for additional burden associated with ADR-related hospital admissions. In a prospective observational study conducted with 18,820 hospital patients in the United Kingdom, Pirmohamed et al. reported that 6.5% of the hospital admissions were related to ADRs. The authors concluded that although many of the implicated drugs had proven clinical benefits, efforts should be made to improve their overall benefit–burden balance.
Results of a prospective 19-week observational study conducted in the United States revealed that 7.5% of all admissions to a medical intensive care unit were related to ADRs. Lazarou et al. conducted an earlier meta-analysis of prospective studies in U.S. hospitals and estimated that the frequency of serious ADRs resulting in hospital admission was 4.7%. Results of another, more recent study in hospital emergency departments conducted by Budnitz et al. indicated that 16.7% of all ADRs were severe enough to require hospitalization. While there is variability in these estimates, a rate of 5% suggests that more than 1.5 million hospital admissions are related to ADRs. Given that a substantial number of ADRs lead to a hospital admission, it is important to know the probability of additional burden that different drug groups could pose by causing ADRs that result in hospitalization.
The primary objective of this study was to characterize the likelihood of therapeutic group-specific ADRs that result in hospital admissions. To accomplish this objective, the Adverse Drug Reaction Hospitalization (ADRH) index, a composite measure proposed by the authors, was calculated for ADRs with the most frequently implicated groups of drugs reported to the New Jersey Poison Information and Education System (NJPIES).