Payroll and Agency-Worker Fraud in Fragmented Health and Social Care Data: A Critical Narrative Review and Forensic Analytics Framework
Adedire Temitope Odelabu *
Neurological Service, Headway Suffolk & Norfolk Neurological Service, Ipswich, United Kingdom.
Ridwan Abiodun Alimi
Department of Statistics, University of Ilorin, Ilorin, Nigeria.
Akeem Babatunde Lawal
Department of Accounting & Finance, McPherson University, Ogun State, Nigeria.
*Author to whom correspondence should be addressed.
Abstract
Payroll and agency-worker fraud presents a difficult assurance problem for health and social care organisations because payments, staffing decisions and evidence of work performed are often distributed across systems and institutions. A payment may appear legitimate in a payroll or invoicing ledger even when its relationship to an authorised worker, approved shift, agreed rate or actual attendance cannot be established. This critical narrative review integrates scholarship on healthcare fraud analytics, continuous auditing, process mining, record linkage, workforce research and information governance with authoritative public-sector fraud guidance. It distinguishes deliberate misrepresentation from payroll error, disputed employment arrangements and legitimate variations in temporary staffing. The evidence indicates that record completeness, worker identity resolution, reliable event chronology and organisational controls are prerequisites for defensible analytics. Studies of healthcare insurance claims provide useful methods but do not directly validate detection of payroll or agency timesheet fraud. Research on temporary staffing and patient outcomes establishes the operational importance of workforce data, without providing evidence that temporary employment itself is fraudulent. The review develops a proposed forensic analytics framework linking source provenance, privacy-aware entity resolution, contract and rostering reconciliation, risk indicators, human adjudication and continuous learning. It examines strengths and limitations of rule-based tests, anomaly detection, graph analytics and process mining, and proposes evaluation against investigated cases rather than conveniently labelled administrative exceptions. Particular attention is given to false accusations, data-access constraints, bias against mobile or atypically scheduled workers, and the potential disruption of essential care. Prospective, multi-organisation evaluations and independently adjudicated reference sets are identified as priority research needs. The framework is a testable synthesis rather than a validated fraud-detection system. Its central proposition is that credible assurance requires evidence reconciliation and proportionate investigation, not an algorithmic presumption that unusual payments or working patterns establish dishonesty.
Keywords: Occupational fraud, temporary staffing, workforce assurance, entity resolution, process mining, continuous auditing, administrative data linkage, public-sector accountability