Article Abstract Online ISSN: 1099-176X    Print ISSN: 1091-4358 The Journal of Mental Health Policy and Economics Volume 5, Issue 4, 2002. Pages: 141-152 Published Online: 5 May 2003 Copyright © 2002 ICMPE.

*Correspondence to: Prof. Mingshan Lu, Department of Economics, University of Calgary, 2500 University Drive, NW, Calgary, AB, Canada, T2N 1N4.
Tel.:  +1-403-220 5488
Fax:  +1-403-282 5262
E-mail: lu@ucalgary.ca      

Source of Funding: Research was funded by grant R21-AA12886 from the National Institute on Alcohol Abuse and Alcoholism; Lu also gratefully acknowledges financial support from the Alberta Heritage Foundation for Medical Research and the Institute of Health Economics.

Abstract

We assess the consistency of clinicians' reports on clients across two data sets. The first one, the Maine Addiction Treatment System (MATS), was an administrative data set; the state government used it for program performance monitoring and evaluation. The second was a set of medical records abstracted directly from the clinical records. We look for evidence of inconsistencies in five categories: admission alcohol use frequency, discharge alcohol use frequency, termination status, admission employment status, and discharge employment status. Multiple imputation methods are employed to address the problem of missing values in the record abstract data set. For admission and discharge alcohol use frequency measures, we find, respectively, strong and supporting evidence for inconsistencies. We find equally strong evidence for consistency in reports of admission and discharge employment status. Finally, we find mixed evidence on termination status. Patterns of inconsistency may be due to both altruistic and self-interest motives.

Background: In the health care market managed care has become the latest innovation for the delivery of services. For efficient implementation, the managed care organization relies on accurate information. So clinicians are often asked to report on patients before referrals are approved, treatments authorized, or insurance claims processed. What are clinicians’ responses to solicitation for information by managed care organizations? The existing health literature has already pointed out the importance of provider “gaming,” “sincere reporting,” “nudging,” and “dodging the rules.” Aims of the Study: We assess the consistency of clinicians’ reports on clients across administrative data and clinical records. Methods: For about 1,000 alcohol abuse treatment episodes, we  compare clinicians’ reports across two data sets. The first one, the Maine Addiction Treatment System (MATS), was an administrative data set; the state government used it for program performance monitoring and evaluation. The second was a set of medical record abstracts, taken directly from the clinical records of treatment episodes. A clinician’s reporting practice exhibits an inconsistency if the information reported in MATS differs from the information reported in the medical record in a statistically significant way. We look for evidence of inconsistencies in five categories: admission alcohol use frequency, discharge alcohol use frequency, termination status, admission employment status, and discharge employment status. Chi-square tests, Kappa statistics, and sensitivity and specificity tests are used for hypothesis testing. Multiple imputation methods are employed to address the problem of missing values in the record abstract data set. Results: For admission and discharge alcohol use frequency measures, we find, respectively, strong and supporting evidence for inconsistencies. We find equally strong evidence for consistency in reports of admission and discharge employment status, and mixed evidence on report consistency on termination status. Patterns of inconsistency may be due to both altruistic and self-interest motives. Discussion and Limitations: Payment contracts based on  performance may be subject to provider mis-reporting, which could seriously undermine its purpose. However, further analysis is needed to determine how much of the inconsistencies observed are results of clinician gaming in reporting. Implications for Health Policy: Increasing system accountability is becoming more and more important for health care policy makers.  Results of this study will lead to a better understanding of physician reporting behavior.  Implications for Future Research: Our work in this paper on the data sets confirms the statistical significance of strategic reporting in alcohol addiction treatment. It will be of interest to confirm  our finding in other data sets. Our on-going research will model the  motives behind strategic reporting. We will hypothesize that both altruistic and financial incentives are present. Our empirical  identification strategy will use Maine’s Performance-Based Contracting system and client insurance sources to test how these incentives affect the direction of clinician’s strategic reporting.

Received 27 November 2002; accepted 9 April 2003

Copyright © 2002 ICMPE