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 entitie s. Eight therapeutic groups were studied: antiasthmatics, antidepressa nts, 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 in dividual chemical entities per quarter year, from January 1989 until D ecember 1994. (2) During this period, drug use increased in all eight therapeutic groups and in most of their constituent chemical groups: I ncreased use of newer drugs did not cause the expected decrease in use of established drugs. Use of all broad-spectrum antimicrobials increa sed by 314%. Use of sedative benzodiazepines decreased slowly and stea dily (16%) throughout the study period but use of all hypnotics increa sed inexplicably by 21% in 1992 reaching a plateau in 1993 and 1994. S SRI antidepressant use increased sharply (5333%) following their intro duction in 1989, accompanied by a 24% increase in use of tricyclic ant idepressants. There was a 23626% increase in the use of proton pump in hibitors 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-u lcer and antimicrobials was accurately targeted and rationally defensi ble. (3) More positively, use of beta(2)-agonist inhalers increased by 45 % despite a 254% increase in the use of inhaled steroids. Use of H RT increased by 389% though evidence of under-use is given. There was a steady increase in the use of both insulins (28%) and oral hypoglyca emics (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. (C) 1997 by John Wiley & Sons, Ltd.