Análisis y predicción de cargas de trabajo en un sistema de dispensación de medicamentos por dosis unitarias en un hospital general

PÉREZ RUIXO, J J; CLIMENTE MARTÍ, M; ORDOVÁS BAINES, J P; BORRÁS ALMENAR, C; PÉREZ PEIRÓ, C; JUAN COLOMER, J; JIMÉNEZ TORRE, N V

Farmacia Hospitalaria

ISSN: 1130-6343

Farmacia Hospitalaria (Farm Hosp) is the official publication of the Spanish Society of Hospital Pharmacists (SEFH), a professional and scientific organization that currently has about 4.351 members.

The Mission of Farm Hosp is to be the support for health professionals for expanding the influence of projects, publications, and training programs, and to promote the exchange of experiences.

Vision Farm Hosp is:
- to increase the state of knowledge about the hospital pharmacy and whose actions should increase the adequate and safe use of medicines and health products.
- to be an instrument of exchange and dissemination of scientific knowledge of pharmacy practice through the best use and safe medication and health outcomes.

It is a peer-reviewed double-blind scientific journal aimed at publishing in the broad field of Hospital Pharmacy advances in Pharmacy Services. This journal included the following topics: Drug evaluation. Quality of life. Patient Reported Outcome Measures (PROMs), and Patient Reported Experience Measures (PREMs) in the use of drugs. Monitoring medication. Pharmaceutical compounding. Medication administration. Management of Pharmacy Services (logistics, quality control). Technologies and information systems in the process of using medication. Medical devices. Patient Education. Pharmacist´s role in the society and /or in the Health System in Spain, and other topics related to clinical, scientific, and institutional role of hospital pharmacists.

Farmacia Hospitalaria publishes bimonthly (6 issues yearly) originals, short originals, reviews, clinical cases, special articles, protocols, consensus documents, clinical guidelines, letters to the editor, editorials.

All the articles are published in Spanish and English language regardless of the language in which they are received.

Farmacia Hospitalaria is an Open Access journal, whose content is available free of charge to users or their institutions. Users are authorised to read, download, copy, distribute, print, search, or link to entire texts of the articles in this journal without prior permission from the editor or author, according to the BOAI definition of Open Access. The papers may be re-used under the terms of the Creative Commons 4.0 license. (CC BY-NC-ND). Such use shall be subject to authorisation from SEFH, the journal owner.

Farmacia Hospitalaria journal charges no fees for submission of papers or publication of articles.

Adherence to international ethics recommendations

Manuscripts must follow the most recent recommendations of the International Committee of Medical Journal Editors (available at http://www.icmje.org).

The articles published in Farmacia Hospitalaria are indexed in PubMed and other 17 date bases:
- Evaluation databases: MEDLINE, DOAJ, Scopus, Latindex y Scimago Journal and Country Rank (SJR). Índice H: 18, quartile 3 (Medicine, miscellaneous / Pharmacology)
- Bibliographic databases: CINHAL, EMBASE Excerpta Medica, Índice Bibliográfico Español en Ciencias de la Salud (IBECS), Índice Español de Ciencia Y Tecnología (ICyT), Índice Médico Español (IME), International Pharmaceutical Abstracts (IPA).
- Journal collections: Dialnet, Medes, Scientific Electronic Library Online (SciELO), REDIB, ScienceDirect, ClinicalKey.

Indexed in:

Medline, Emerging Sources Citation Index (ESCI), Directory of Open Access Journals (DOAJ)

Objetivo: Caracterizar la temporalidad de la carga de trabajo mensual en un sistema de dispensación individualizada de medicamentos en dosis unitarias y desarrollar un modelo matemático para su predicción. Material y métodos: Tipo de hospital: hospital general universitario de 573 camas. Extracción de datos: estadísticas mensuales de dispensación de medicamentos por el sistema de dosis unitarias. Periodo de estudio: enero 1994 ¿ diciembre 1999. Grupos: los datos disponibles hasta diciembre de 1998 se utilizan para identificar el modelo estadístico (grupo de identificación) y los datos de 1999 se utilizan para validar el mismo (grupo de validación). Modelos estadísticos: la variable dependiente es el número de dosis unitarias dispensadas/mes. El modelo de regresión lineal múltiple utiliza como variables independientes, el tiempo transcurrido desde el inicio del estudio (años), el mes y el año en que se efectúan las dispensaciones. El desarrollo del modelo ARIMA se realiza mediante la metodología de Box-Jenkins. Evaluación de la capacidad predictiva: media del error relativo de predicción (exactitud) y raíz cuadrada de la media del error cuadrático de predicción (precisión) de las predicciones realizadas durante 1999 mediante el modelo ARIMA desarrollado. R e s u l t a d o s : Durante el periodo de tiempo estudiado, se dispensaron mensualmente 98.760 ± 9.074 dosis unitarias. En el modelo de regresión lineal desarrollado (r2 = 0,78) se observa una tendencia lineal anual que incrementa en 3.392 (IC 95%: 2.609 a 4.175) el número de dosis unitarias dispensadas/mes. Asimismo, respecto al mes de enero, las dispensaciones del mes de marzo se incrementan en 6.783 (IC95%: 138 a 13.429) unidades y, las dispensaciones de los meses de agosto y septiembre disminuyen en 17.762 (IC 95%: 11.386 a 24.137) y 11.115 (IC 95%: 4.740 a 17.490) unidades, respectivamente. El modelo desarrollado puede ser empleado con fines predictivos, ya que posee la exactitud, -3,06% (IC 95%: 0,91 a ¿5,20), y precisión, 3,37% (inferior al 15%) adecuadas. Conclusión: La disponibilidad de un modelo matemático para predecir la carga de trabajo en el sistema de dispensación en dosis unitarias es una herramienta eficaz para planificar la adecuación de recursos humanos en el Área de Dispensación del Servicio de Farmacia.

Palabras clave:

Sistema de dispensación de medicamentos en dosis unitarias

Carga de trabajo

Series temporales

Optimización de recursos

Objective: To characterize the time course of monthly workload in a unit-dose drug dispensing system and to develop a mathematic model in order to predict it. M e t h o d s : Hospital: general university hospital with 573 beds. Data collecting: monthly drug dispensed statistics of a unit-dose system. Study period: January 1994 ¿ December 1999. Groups: data available till decembre of 1998 were used to identify the mathematical model (training group) and data collected during 1999 were used to validated the model (validation group). Statistical models: the dependent variable was the unit-dose dispensed monthly. The multiple linear regression used the time since the begining of the study (years), the month and the year of the monthly unit-dose dispensed, as an independient variables. The ARIMA model development was made with Box-Jenkins methodology. Predictive performance analyses: mean of relative prediction error (bias) and root square of the mean of the squared relative prediction error (precission) in 1999 were used in validation step. R e s u l t s : During the time studied, the mean (standard deviation) of unit-dose dispensed was 98.760 (9.074) unit-dose per month. The multiple lineal regression model (r2= 0.78) identified a annual lineal tendency to increase by 3,392 (IC 95%: 2,609 a 4,175) unitdose dispensed per month. If we compare with january, the unitdose dispensed in march increases by 6,783 (IC 95%: 138 to 13,429) units, but in august and september, decreased by 17,762 (IC 95%: 11,386 to 24,137) and 11,115 (IC 95%: 4,740 to 17,490) units, respectively. The model developed could be used in predictive setting because it has an optimal bias, -3,06% (IC 95%: 0.91 to ¿5.20), and precision, 3.37% (lower than 15%). Conclusión: The availability of a mathematical model in order to predict the workload in a unit-dose dispensing system is a usefull tool for human resources planification in the Pharmacy Department.

Keywords:

Unit-dose drug dispensing system

Workload

Time series analysis

Resource optimization

Full text is only available in PDF