Determinants of specialist physician ambulatory visits: a neurology example

• Published in: Journal of medical economics Vol. 22; no. 8; p. 830
• Main Authors: van der Goes, David N, Ney, John P, Garrison, Louis P
• Format: Journal Article
• Language: English
• Published: England 03-08-2019
• Subjects: Adult; Age Factors; Aged; Ambulatory Care - statistics & numerical data; Comorbidity; Female; Health Services Accessibility - statistics & numerical data; Humans; Male; Middle Aged; Neurologists - statistics & numerical data; Residence Characteristics - statistics & numerical data; Sex Factors; Socioeconomic Factors; United States; United States; determinants of medical care utilization; Specialist physicians; I1; ambulatory care; I12; I14; neurology
• ISSN: 1941-837X
• DOI: 10.1080/13696998.2019.1618861

Economic theory argues that specialization in medicine improves efficiency. Current literature suggests that access to and utilization of specialist care vary widely based on many determinants. Thus, understanding the determinants of specialist physician ambulatory care utilization is integral to healthcare policy.
The objective is to investigate the individual and community determinants of specialist ambulatory care utilization-specifically neurologists. The aim was to find predictors of specialist utilization and to identify the particular determinants that can be modified by regulatory or legislative action.
A large claims database, Truven Health Analytics™ Marketscan data, was used from 2007-2010 as the sample. These data are supplemented with data from the American Academy of Neurology (for geographic distribution of neurologists) and the US Census American FactFinder (for community demographic factors). Multivariate regression analysis was run to test the hypotheses. Several robustness tests of our models were included.
Most importantly, neurologists per capita has a meaningful impact on utilization. Additionally, the difference in neurologist usage by neurological condition is an important factor. It was also found that union status, age, comorbidities, and diagnosis are significant individual level determinants, and that the percentage of Hispanic residents and median income are significant community level determinants.
There are two predictors believed to be the most important. The first is the unique neurologists per 1,000 capita variable, which shows a small increase in the number of neurologists would be correlated with a small increase in the probability of seeing a neurologist. We suggest that this is within policymakers' control, and policymakers should consider this action in the face of the predicted shortage. The second is what appears to be possible sorting by neurologists of patients based on diagnosis - the large difference in the fraction of patients seeing a neurologist by disease.