A Fraud Detection System for Health Insurance in Nigeria

Résumé

Background and Purpose: This research developed a Fraud Detection System for National Health Insurance Scheme (NHIS) in Nigeria. This was with a view to addressing the fraudulent activities of some stakeholders in NHIS; as many researches have proven that the lack of appropriate tools to do this has negatively affected service providers as well as the beneficiaries of this Scheme. 

Methods: In order to achieve the aim of this research, an inspection of organizational documents, direct observation and collection of existing data from NHIS accredited health facilities and Health Maintenance Organizations in Nigeria were carried out. The system was designed using Unified Modelling Language (UML) tools. The implementation of the system was done using MongoDB as the big data storage mechanism for the input, Comma Separated Values (CSV) files as a storage facility for the intermediate results generated during processing and MySQL as the storage mechanism for the final output, Apache MapReduce as the big data processing platform, Association Rule Mining as the data analytics algorithm, and Java programming language as the implementation technology.

Results: The system modules of comprised of four modules: user management, enrollment processing, referral processing and claims processing. With this, it identified different types of frauds in NHIS such as double billing, billing for services not provided, ghost patients, identity theft, self-referral, collusion with providers and kickback schemes.

Conclusions: This paper developed a system for the detection of the fraudulent activities of the actors of NHIS. This system employed data from the Nigerian NHIS which was categorised into: enrollment, referral and claim data with different file formats: pdf, jpg, png, csv and excel.

Keywords: Fraud, Detection, Health, Insurance, Prevention, NHIS.