PERFORMANCE MEASUREMENT SYSTEM MODEL ORIENTED TO HEALTH SERVICES MANAGEMENT
Performance Measurement System; Health Services Management; Primary Health Care; Machine Learning; Analytical Digital Twin; Decision Support.
This thesis develops and empirically validates a Performance Measurement System
model oriented toward the management of complex health services, using Primary Health Care
as a critical empirical field. The study recognizes that public health services operate under
conditions of organizational complexity, fragmented data, institutional constraints, territorial
inequalities, limited technological maturity, and permanent demand for evidence-informed
decision-making. Methodologically, the thesis articulates three complementary studies aligned
with its specific objectives. The first study identifies, through a systematic literature review,
constructs, methods, indicators, dimensions, and research gaps associated with Performance
Measurement Systems in Primary Health Care. The second study develops and empirically
validates a PMS in a vulnerable semi-arid municipality, based on formative diagnosis, co-
production with local actors, indicator construction, pilot application, and longitudinal
monitoring. This process resulted in an 18-indicator PMS structured for managerial use in
Primary Health Care. The third study extends the system using Machine Learning techniques
and a PMS-based Analytical Digital Twin, leveraging seven months of monitoring data from
seven PHC units to assess forecasting, scenario analysis, systemic interpretation, and decision-
support capacity. The findings indicate that the central contribution of the thesis lies in
proposing a situated methodological architecture that articulates scientific evidence, contextual
adaptation, stakeholder participation, longitudinal monitoring, and analytical-predictive
expansion. The thesis contributes to the field of Performance Measurement Systems by showing
how a PMS model can be developed, validated, and analytically expanded in complex health
services under real public management conditions, without reducing performance governance
to isolated indicators, retrospective dashboards, or purely technological solutions.