Adverse Drug Reactions in Hospitalized Patients: A Critique of a Meta-analysis

Marion Kvasz, MD, MPH; I. Elaine Allen, PhD; Matthew J. Gordon, BA; Eric Y. Ro, BA; Rhonda Estok, RN; Ingram Olkin, PhD; and Susan D. Ross, MD, FRCPC

In This Article

Literature Search

In the meta-analysis, four electronic databases were searched. We also searched these same databases: MEDLINE, Excerpta Medica, International Pharmaceutical Abstracts, and Science Citation Index. The meta-analysis authors' search went to 1996, but ours was extended to 1998, to also see how many new citations had appeared in the interval since the original search. The MEDLINE search strategy was as follows: (1) explode hospitalization, (2) explode pharmaceutical preparations/adverse event, (3) explode drug therapy/adverse event, (4) 2 or 3, and (5) 1 and 4. This strategy was developed according to the cited medical subject headings in the meta-analysis. For the other databases, these terms may not have been relevant and the key words used were adverse drug, adverse reaction, drug-related, drug-induced, and hospital. MEDLINE was searched from 1966 to present, Excerpta Medica from 1980 to present, International Pharmaceutical Abstracts from 1970 to present, and Science Citation Index from 1989 to present. A manual search of reference sections of all accepted papers supplemented the electronic search in the original meta-analysis, and we did likewise. The meta-analysis authors also sent letters to researchers requesting unpublished data. We omitted this step, because it was not stated by the meta-analysis authors that this step yielded useful information. In addition, we obtained the list of studies rejected by the meta-analysis authors to compare our rejected study list with theirs.