Changes in mental health system performance: the case of Gipuzkoa (Basque Country, Spain)

• Published in: International journal of integrated care 2019-08, Vol.19 (4), p.204
• Main Authors: Almeda, Nerea, García-Alonso, Carlos R., Salinas-Pérez, José A., Iruin Sanz, Álvaro, Salvador-Carulla, Luis
• Format: Article
• Language: English
• Subjects: management; mental health system; planning; policy; relative technical efficiency
• ISSN: 1568-4156; 1568-4156
• DOI: 10.5334/ijic.s3204

Use of decision support systems may improve policy-making for management of mental health services and systems. System performance can be analyzed by using Relative Technical Efficiency (RTE), stability and entropy indicators. These indicators summarize resource availability, utilization and results as a balance between inputs (resources) and outputs (outcomes). Nevertheless, performance and stability assessment of mental health systems is complex because of difficulty related to data collection, results interpretation and translation of information into practice. The present mental health context requires better ways of planning for allocating resources and improving outcomes. The objective of this study is to assess the RTE, stability and entropy 2012-2015 variations as a consequence of a policy developed by the Mental Health Network of Gipuzkoa (Basque Country, Spain). The Mental Health Network of Gipuzkoa is structured by thirteen small health areas. In these catchment areas, mental health services were standardized by using the DESDE-LTC codification tool. Mental health services were classified according to the main type of care provided (outpatient, day and residential). In the analysis, 57 variables were included, which were classified in resources –inputs- (availability, placement and workforce capacity) and results –outputs- (service utilization, readmissions, discharges and length of stay). A hybrid decision support system, that integrates statistical, operational and artificial intelligence techniques, has been used to analyze the indicators. The main statistical procedure was a Monte-Carlo simulation engine to include the uncertainty of real contexts. The data envelopment analysis, an operational technique, was utilized to assess the RTE. In addition, a prototype of fuzzy inference engine was included for interpreting expert knowledge according to the basic community mental health care model. The stability was calculated by analyzing the frequency distributions of the RTE and, finally, the Shannon’s entropy to estimate system disorder. The main structure of the real policy was identified by developing structured interviews to senior managers and planners of Mental Health System of Gipuzkoa. Results provided information about the changes of the selected indicators throughout three years (2012-2015). The impact of the policy developed can be considered positive but the stability remains poor as new real interventions have to take into account tha