Strategies for the analysis of imputed data from a sample survey. The National Medical Care Utilization and Expenditure Survey.

Missing data in sample surveys is virtually unavoidable, whether it is an entire unit that is missing or only an item for a responding unit. Compensation for unit nonresponse is usually made through the assignments of weights to responding units; for item nonresponse, the compensation often is by an imputation procedure. This paper reviews the extent of missing data in a large federal survey, the National Medical Care Utilization and Expenditure Survey, and the imputation procedures used to compensate for item missing data. The effects of imputation on several types of estimates from the survey are examined. In addition, several methods for analyzing survey data with imputed values are reviewed, and recommendations about preferred strategies are made for selected circumstances.

Costs of illness: United States, 1980.

The total costs of illness and injury in the U.S. civilian noninstitutionalized population in 1980 amounted to $381.7 billion. The direct costs of illness and injury--resource expenditures for the diagnosis, treatment, and management of medical and dental conditions--were $153.9 billion, or 40.3 percent of total costs. Indirect costs--economic losses from morbidity and mortality--were $227.9 billion, or 59.7 percent of total costs. Of indirect costs, $104.9 billion resulted from productivity losses because of morbidity, and $123.0 billion represent the present value of lost productivity from premature mortality based on a net effective discount rate of 4 percent. These estimates, based on data from the 1980. National Medical Care Utilization and Expenditure Survey (NMCUES), differ from other estimates of the costs of illness and injury in 1980 (Gibson and Waldo, 1982; Rice, Hodgson, and Kopstein, 1985). The differences, which can be resolved, are attributable to two major factors: (1) NMCUES includes only the civilian noninstitutionalized population, but the other estimates include the institutionalized population and the military; and (2) NMCUES indirect cost estimates for the population unable to work include persons who were retired for health reasons in 1979 and 1980, disabled homemakers, and other persons who were disabled for the entire year 1980 but were not retired for health reasons in 1979, but the Rice et al. estimates do not include the last two categories in the population unable to work. The principal NMCUES findings on the total costs of illness in the civilian noninstitutionalized population reinforce the importance of considering distributional effects. Persons 65 years of age and over represent one-tenth of this population yet account for more than one-fourth of direct costs and more than their share of total costs, even though the institutionalized elderly are excluded. More than two-thirds of total costs for this age category are accounted for by direct costs. Direct costs also account for more than two-thirds of total costs for people under 17 years of age. However, this youngest age category, which constitutes over one-fourth of the civilian noninstitutionalized population, generates only 12.3 percent of direct costs. In contrast, indirect costs account for well more than 60 percent of total costs for the working-age population (17-64 years of age). Within the working-age population, per capita direct costs are highest among persons who are not full participants in the work force, many of whom are not working full time or at all because of injury or ill health.(ABSTRACT TRUNCATED AT 400 WORDS)