This study included all the office-based physician visits included in the National Ambulatory Medical Care Survey (NAMCS) between 2000 and 2002 that were made by persons over the age of 60. A dementia/AD visit was defined based upon ICD-9 diagnosis codes for senile dementia, vascular dementia, AD and senility (290.xx, 294.9x, 331.xx and 797.xx). A cholinesterase inhibitor visit was defined as either one or more prescription indications for Aricept 10 mg, galantamine and tacrine.
NAMCS is an annual survey conducted by the National Center for Health Statistics (NCHS) that provides a probabilistic sample of office-based physician visits. NAMCS is a population-based instrument, and it incorporates a multistage probability design with sampling procedures that have been discussed extensively. In brief, NAMCS is collected annually from selected nonfederally employed physicians, excluding those in the specialties of pathology, anesthesiology and radiology. The unit of analysis in NAMCS is a patient visit record made to an office-based physician, and each has a provided patient visit weight value to allow extrapolation to national estimates. The primary goal of conducting NAMCS is to measure the national healthcare utilization of ambulatory care services among the noninstitutionalized population in the United States.
A cross-sectional analysis using NAMCS data from 2000-2002 was conducted with goals to estimate prevalence of dementia/AD diagnosis and cholinesterase inhibitor prescription and to investigate predictive factors of cholinesterase inhibitor prescription in the ambulatory care setting. Pooled annual data from three years were used in order to increase power to obtain reliable estimates, as the computation of weighted estimates using NAMCS renders visit rates based on 30 or fewer actual visits unreliable for calculating national estimates. Therefore, some of the findings from subgroup analyses reported unweighted data as those visit rates were lower than 30 records. propecia 1 mg
Univariate and bivariate analyses were done in order to describe the overall study population. For the predictive model, the dependent variable was a dichotomous measure of cholinesterase inhibitor prescription. The independent variables based on visit characteristics included age, race, gender, dementia or AD diagnosis, insurance type, region, physician’s specialty and metropolitan status. Adjusted and unadjusted odds ratios with 95% confidence intervals (CIs) were derived from the logistic regression model and were used to determine the statistically significant predictors of cholinesterase inhibitor prescription.