Publication Details
Abstract
The home and community-based services offered by Medicaid have been important in facilitating the elderly and disabled groups although the programs have been experiencing massive problems in regard to fraud, waste, and abuse. Fragmented data environments, intricate service delivery models, and lack of integration between administrative and clinical systems broaden the scope of collusive fraud by Licensed Home Care Services Agencies (LHCSAs) and Social Adult Day Care services. Medicaid improper payments are still enormous and the lapses in documentation, discrepancies in eligibility, and misrepresentation of services often conceal the organized fraudsters among providers, caregivers, and beneficiaries. This study examines the role of a well-organized data governance system to strengthen the integrity of the Medicaid program by facilitating fraud that has been practiced in a collusive manner within the LHCSA and the Social Adult Day Care industries. The analysis aims at integrating and governing heterogeneous data, such as Medicaid claims, patient records, caregiver timesheets, billing submissions and Electronic Visit Verification (EVV). logs, in a systematic way. The proposed framework will overcome inherent weaknesses in effective fraud detection by defining common data format, quality control, lineage management, role access control, and the potential of using AI-based forecasting tools. Based on this governance background, the paper puts the analytical methods of rule-based validation, anomaly detector, and machine learning-based risk score into action to discover abnormal service patterns that are indicative of collusion. These trends involve the overbilling of services not yet provided, exaggerated service times, joint reimbursement of services between the caregivers and the beneficiaries, and manipulations of structural loopholes between the consumer-directed and the agency-directed care systems. The analytical results are incorporated into a human-with-the-loop review procedure, which makes them interpretable, regulation ally defensible and operationally relevant. Results show that analytics based on governance can be used to a significant level of success in identifying and prioritizing high-risk claims and provider behavior and minimizing false positives and the workload of investigation. The research concludes that data governance is neither a support functionality nor an enabler of sophisticated fraud detection but a basic enabler of advanced fraud detection functions. When the governance practices are aligned to the analytical intelligence, the Medicaid agencies are able to tighten their control, minimize improper payments and secure the social resources without hindering the provision of needed home and adult day care services.