Risk of clinical blood dyscrasia in a cohort of antibiotic users.

Blood dyscrasias, although rare, can be fatal. Many drugs, including antibiotics, are associated with these dyscrasias. We conducted a cohort study with a nested case-control analysis using data from the General Practice Research Database to estimate incidence rates of clinical blood dyscrasias in the general population and to examine their association with use of antibiotic drugs. The study population consisted of patients aged 5-69 years receiving at least one antibiotic prescription from January 1994-September 1998. The final cohort consisted of 822,048 persons who received 1,507,307 antibiotic prescriptions during the study period. The main outcome measure was a diagnosis of neutropenia, agranulocytosis, hemolytic anemia, thrombocytopenia, bicytopenia, pancytopenia, or aplastic anemia. We confirmed 122 patients who developed clinical blood dyscrasias. The incidence was 3.3/100,000 person-years in the general population. Patients older than 60 years (relative risk [RR] 2.8, 95% confidence interval [CI] 1.6-5.0) and those who took phenothiazines (RR 49.0, 95% CI 4.9-488.2) had an increased risk of blood dyscrasia. Users of antibiotics had an RR of 4.4 (95% CI 2.6-7.5), and patients taking more than one class of antibiotics had an RR of 29.1 (95% CI 9.1-92.8). Among individual antibiotic classes, the greatest risk was with cephalosporins (RR 13.8, 95% CI 3.6-52.6). Although uncommon, our study supports an association between blood dyscrasias and antibiotics.