Estudio comparativo de los servicios de mantenimiento utilizando técnicas de minería de datos

Translated title of the contribution: Comparative study of maintenance services using data mining techniques

Antonio Miguel Cruz, Wilmer A. Aguilera-Huertas, Darío A. Días-Mora

Research output: Contribution to journalArticlepeer-review

6 Scopus citations


Objectives: To compare the quality of the maintenance service of two health entities. One has outsourced maintenance services and the other has its own. The quality indicator under study is the time of change of state. Study the behavior of health service productivity versus availability.
Materials and Methods. An inventory of medical and non-medical equipment was taken at two dialysis facilities located in two health facilities. Both third level. These have similar characteristics in terms of dialysis units and medical equipment. Each of them has 16 patient care units and 92 medical teams. The difference is the maintenance service; one is outsourced while the other is outsourced. Maintenance work orders were collected for a period of 6 months and the indicators: Response Time, Actual Duration, Lost Times and Change of Status Time (TAT) were calculated for each work order. A predictor was built for the TAT variable (dependent variable) as a function of the other variables.
Results. The quality of maintenance service by the entity that has its own staff is better than the entity with outsourced services. The TAT indicator is an average of 2.95 hours for the entity with its own maintenance, while for the other entity it is 3.4 hours.
Conclusions: The behaviour of service productivity versus availability was found to be of a linear positive type.
Translated title of the contributionComparative study of maintenance services using data mining techniques
Original languageSpanish
Pages (from-to)641 - 653
Number of pages12
JournalRevista de Salud Publica
Issue number4
StatePublished - 2009


Dive into the research topics of 'Comparative study of maintenance services using data mining techniques'. Together they form a unique fingerprint.

Cite this