Abstract:

Presented is a cohort analysis of the use of health services by U.S. elderly. Standard cohort tables are constructed, using data on approximately 50,000 individuals from the 1972, 1976, & 1980 U.S. Health Interview Surveys. Six 4-year age cohorts are tracked: the youngest, aged 56-59 in 1972; the oldest, aged 84-87 in 1980. Significant age & period effects on MD & hospital contact are shown; eg, the aging effect reflects an increase in the % of individuals who either saw a MD on an outpatient basis or were hospitalized in the previous year. Analysis of the volume of MD visits reveals an inverse J-curve relationship with age, with the pivotal point occurring at or about age 80. Consideration of 9 alternative explanations of the J-curve phenomenon shows that 3 are plausible on either conceptual or empirical grounds. Two of these explanations suggest the substitution of either (1) hospital-based services for ambulatory-based services, or (2) social supports (especially in the form of older children as caregivers) for MDs' services; the third explanation focuses on the involuntary severance of patient-practitioner relationships resulting from the retirement or death of long-standing family MDs. Competing policy implications of these explanations are discussed.