Skip navigation

Adverse Drug Reactions

In Bandolier 28 we highlighted a report on adverse drug reactions. That report has now been included with 38 others in a meticulous meta-analysis of adverse drug reactions both in US hospital inpatients and those patients admitted with an adverse drug reaction [1].


The searching strategy was heroic, and included examination of a number of electronic databases together with postal questionnaires to researchers. Studies were restricted to US hospitals, since the bulk of information was from that country. The methods and definitions are meticulously laid out, allowing those who are interested in the figures to follow all the manipulations.


The outcome the researchers used was the WHO definition: any noxious, unintended, and undesired effect of a drug, which occurs at doses used in humans for prophylaxis, diagnosis, or therapy. This excludes therapeutic failures, intentional or accidental poisoning or drug abuse, and adverse effects due to errors in administration or compliance. Not everyone likes this definition (see Bandolier 30 ), but it was the most common in the reports found.
A serious ADR was defined as one that requires hospital admission, prolongs hospital stay, is permanently disabling, or results in death. Serious ADRs therefore include fatal ADRs, which were also analysed separately.


The main results (based on 33 million US hospital admissions in 1994) are shown in the Table. They show that 2.1% of in-patients experienced a serious ADR, and that 4.7% of hospital admissions were due to a serious ADR. Fatal ADRs occurred in 0.19% of in-patients and 0.13% of admissions.
Incidence of adverse drug reactions in the USA
  Number of studies Patients studied Incidence of ADR (95% CI) Estimated number (thousands, with 95% CI)
ADRs in patients while in hospital
Serious 12 22,500 2.1 (1.9 to 2.3) 702 (635 to 770)
Fatal 10 28,900 0.19 (0.13 to 0.26) 63 (41 to 85)
ADRs in patients admitted to hospital
Serious 21 28,000 4.7 (3.1 to 6.2) 1547 (1033 to 2060)
Fatal 6 17,800 0.13 (0.04 to 0.21) 43 (15 to 71)
Overall ADR incidence
Serious 33 50,500 6.7 (5.2 to 8.2) 2216 (1721 to 2711)
Fatal 16 46,600 0.32 (0.23 to 0.41) 106 (76 to 137)
Overall it was estimated that some 2.2 million serious ADRs would have occurred in 1994, with 106,000 of them fatal.
Although the reports covered some four decades, there was no evidence of any systematic change in the rate of serious ADRs, and the population studied was similar to that in the US hospitals. Medical wards were over-represented, though the authors did considerable work to investigate how this might have affected results using the most conservative estimates (not much). Issues like hospital setting (teaching versus non-teaching), and possible errors in compiling data in the original reports are dealt with thoroughly.


This report suggested that fatal ADRs rank from the fourth to the sixth leading cause of death in the USA after heart disease, cancer, and stroke, and similar to pulmonary disease and accidents. Costs associated with ADRs were estimated at up to $4 billion a year.
This is an important clinical issue. Those who want to think about it, and how it might be tackled, should read, digest and understand this careful and thoughtful piece of work.


  1. J Lazarou, BM Pomeranz, PN Corey. Incidence of adverse drug reactions in hospitalized patients: A meta-analysis of prospective studies. JAMA 1998 279: 1200-5.

previous or next story in this issue