Multistage samples and the minimum sum method for Medicare fraud investigations.

摘要

Random sampling of paid medicare claims has been legally acceptable for investigating suspicious billing practices by health care providers since 1986. A population of payments made to a given provider during a given time frame is isolated and a probability sample selected for investigation. A lower confidence bound for the total overpaid amount in the population is then used as a recoupment demand. Edwards et al. (Health Services and Outcomes Research Methodology, 2005) show that methods based on the Central Limit Theorem can fail badly and propose an alternative method, called the minimum sum method, for fixed sample sizes. In this dissertation we consider two different sampling approaches: two-stage sampling and sequential sampling. In the two-stage plan, in case of little abuse in the first stage "probe" sample the investigation is stopped; otherwise a second sample is examined. Based on this strategy an exact conditional lower confidence bound for the total number of universe payments in error and a corresponding lower bound for the total overpayment amount are defined. Criteria for choosing the sampling design parameters are considered. The sequential plan examines claims one at a time and if there is enough evidence of abuse (high error rate) it continues the sampling. Several measures for "enough evidence" are considered, and exact conditional lower confidence bounds for number of payment in error and total overpayment amount are defined. Relative efficiencies of the fixed sample, two-stage and sequential sampling plans are studied.; Keywords: medicare fraud, two-stage sampling, sequential sampling, acceptance sampling, hypergeometric distribution, lower confidence bounds