HYPERTENSION IN THE ELDERLY - IMPLICATIONS AND GENERALIZABILITY OF RANDOMIZED TRIALS

Objective.-To estimate morbidity and mortality benefits of drug therap y for hypertensive elderly subjects, compare these benefits with effec ts in younger subjects, and provide a framework for generalizing resul ts derived from trials to actual patients. Data Sources.-A literature search using MEDLINE from 1966 to 1993, references from reviews and tr ial articles, and experts. Study Selection.-Randomized trials lasting at least 1 year that evaluated effects of drug treatment on morbidity and mortality outcomes in hypertensive persons. Data Extraction.-Four independent reviewers appraised protocol characteristics and quality o f selected trials. Data Synthesis.-There were 13 trials involving 16 5 64 elderly persons (age 60 years and older). The prevalence of cardiov ascular risk factors, cardiovascular disease, and competing comorbid d iseases was lower among trial participants than the general population of hypertensive elderly persons. When the six large high-quality tria ls were combined, trial results showed 43 subjects (95% confidence int erval [CI], 31 to 69) and 61 subjects (95% CI, 39 to 141) needed to be treated for 5 years to prevent one cerebrovascular event and one coro nary heart disease event, respectively. Including the other seven tria ls did not change the results significantly. Only 18 subjects (95% CI, 14 to 25) needed to be treated to prevent one cardiovascular event (c erebrovascular or cardiac). Twelve trials in primarily younger and mid dle-aged adults involved approximately 33 000 persons. For all outcome s except cardiac mortality, two to four times as many of the younger s ubjects as the older subjects needed to be treated for 5 years to prev ent morbid and mortal events. No significant effect on cardiac mortali ty was seen among younger subjects, while 78 older subjects (95% CI, 5 0 to 180) needed to be treated to prevent a fatal cardiac event. Concl usions.-Randomized trials demonstrate that treating healthy older pers ons with hypertension is highly efficacious. Five-year morbidity and m ortality benefits derived from trials are greater for older than young er subjects. Extrapolating benefits from trials to individual patients is difficult, but should take into account multiple issues including the patient's risk factors, preexisting cardiovascular disease, and co mpeting comorbid illnesses.