Temporal trends in drug use in one UK region, revealed by chemical group matching.

(1) The pharmaceutical pricing data for Northern Ireland were amended to include defined daily dosages (DDD) for all single chemical entities. Eight therapeutic groups were studied: antiasthmatics, antidepressants, antimicrobials, benzodiazepines, hormone replacement therapy (HRT), hypoglycaemics, lipid-lowering agents and ulcer-healing drugs. Each group was then subdivided into its main chemical groups. The regional use of each chemical group was defined as the combined DDDs of its individual chemical entities per quarter year, from January 1989 until December 1994. (2) During this period, drug use increased in all eight therapeutic groups and in most of their constituent chemical groups. Increased use of newer drugs did not cause the expected decrease in use of established drugs. Use of all broad-spectrum antimicrobials increased by 314%. Use of sedative benzodiazepines decreased slowly and steadily (16%) throughout the study period but use of all hypnotics increased inexplicably by 21% in 1992 reaching a plateau in 1993 and 1994. SSRI antidepressant use increased sharply (5333%) following their introduction in 1989, accompanied by a 24% increase in use of tricyclic antidepressants. There was a 23,626% increase in the use of proton pump inhibitors and a smaller but steady increase of 38% in use of histamine H(2) antagonists; it is unlikely that much of the prescribing of anti-ulcer and antimicrobials was accurately targeted and rationally defensible. (3) More positively, use of beta(2)-agonist inhalers increased by 45% despite a 254% increase in the use of inhaled steroids. Use of HRT increased by 389% though evidence of under-use is given. There was a steady increase in the use of both insulins (28%) and oral hypoglycaemics (34%). The use of 'statins' (690%) and fibrates (123%) increased. (4) The possible interpretations and implications of these patterns of drug use is discussed, together with their potential as proxies for morbidity incidence in the community.