An integrated, field-level synthesis of digitally enabled performance measurement and management systems (PM/PMS) in healthcare is provided, addressing fragmentation across tool-centered streams and limited cross-stage comparison of control configurations. The thematic structure of the field is mapped, themes are positioned along the balance between technical and social controls, and their placement across PM/PMS stages (design, implementation and use/revision) is made explicit.
A bibliometric science-mapping design was applied to Web of Science Core Collection records from 2000 to 2025. Records were filtered by indices, document type (articles, reviews), language (English) and subject-category refinements aligned with health management and information-systems relevance. Keyword variants were harmonized using a domain thesaurus, and low-specificity tokens were filtered. VOSviewer generated keyword co-occurrence networks (full counting, a minimum occurrence of 10 and association-strength normalization) and an overlay by average publication year. Clusters were interpreted through control-balance and stage lenses, and descriptive profiles were produced by outlets, journals, periods and country groups.
A coherent thematic structure is revealed, organized around infrastructures that stabilize measurement, implementation arrangements that make indicators governable through review routines and accountability and service contexts in which measures are enacted and revised. A temporal overlay indicates a recency gradient from established infrastructure and measurement foundations toward more recent emphases in distributed care and analytics-augmented practice. Artificial intelligence (AI), machine learning (ML) and natural language processing terms are positioned as extensions at the edge of improvement-oriented work rather than as a detached methods stream. Descriptive profiles indicate uneven anchoring across outlet types and country contexts.
Coverage is limited by reliance on a single database and on author and indexer keywords, which can under- or overrepresent niche topics. Co-occurrence indicates conceptual proximity rather than causal effects, and indexing updates can change record counts over time. The geographic concentration of contributing country contexts also limits the transferability of findings across diverse health-system settings. Even with these constraints, field-level configurations are made visible in a transparent, reproducible way. Robustness could be tested by extending the corpus to Scopus and PubMed, using alternative vocabularies where available and examining longitudinal slices of the corpus.
Implications are offered as navigation cues rather than prescriptive best practices. At design, attention can be directed to data quality, traceability and interoperability so that later routines receive reliable signals. At implementation, dashboards and indicator sets become consequential when coupled to explicit review cycles, escalation paths and role responsibilities that enable cross-boundary sensemaking. At use and revision, measures can be aligned with pathway and modality demands, including access, continuity and patient experience, with equity monitored explicitly. Where AI or ML is used, post-deployment monitoring and controlled updating can be integrated into existing review routines.
System-level value can be supported when measurement is used for learning and coordination rather than compliance alone. Distributed care settings raise equity and access concerns and intensify demands for patient-facing coordination. Social-integrative control, including participation, leadership-enabled review routines and cross-boundary coordination, can help mitigate gaming and tunnel vision while sustaining improvement. Attention to equity stratification and patient-reported experience and outcome measures can help ensure that technical tools translate into fairer and safer care.
A reproducible, map-based synthesis is offered that moves beyond tool-specific silos and interprets the field through combined lenses of technical versus social control and PM/PMS stages. An implementation-oriented bridge is surfaced between infrastructure and use, and AI-related vocabulary is positioned as embedded within improvement work. A recency gradient is identified that distinguishes established foundations from newer emphases in distributed care and analytics-augmented practice. Descriptive profiles across outlets, journals, periods and country groups add boundary conditions for transferability claims.
