Dec 31, 1999
Published in: Medical Care Research and Review, v. 57, no. 2, Jun. 2000, p. 182-195
Posted on RAND.org on December 31, 1999
This study examines how preauthorization affects outpatient behavioral health utilization under managed care by comparing plans with similar benefits, but differing in the number of visits authorized. The authors compare plans primarily authorizing in increments of 5 visits to plans authorizing in increments of 10 visits. They analyze the likelihood of terminating outpatient service between the two groups using conditional logistic regression. Results suggest that patients whose treatment is authorized in increments of 5 sessions are nearly 3 times more likely to terminate treatment at exactly the fifth visit than if their treatment is authorized in increments of 10 sessions conditional on being in treatment until the 5th visit. The likelihood of termination peaks in both the 5- and 10- session authorization at the 10th visit, but the difference is not statistically significant. The authorization effect differs by provider type and is weaker among psychiatrists than among nonphysician providers.