Uso de modelos lineales mixtos para la estimación del tiempo de vida útil de un medicamento
DOI:
https://doi.org/10.17533/udea.vitae.v24n2(2)a09Keywords:
Vida útil, Estabilidad, modelos lineales mixtos, medicamentosAbstract
Antecedentes: el campo de la confiabilidad estadística ha realizado importantes avances en el modelamiento de los datos obtenidos en estudios de estabilidad para diferentes equipos y productos, por ejemplo electrodomésticos, componentes de vehículos, circuitos eléctricos entre otros, sin embargo, su permeación en el área farmacéutica no se ha visto promovida aun considerando la complejidad de los ensayos de estabilidad y el impacto económico y social de la definición de la vida útil de un medicamento; adicionalmente, se ha identificado que debido a los procesos de fabricación por lotes de los medicamentos, el efecto de la variación asociada a ellos debe ser considerada a la hora de realizar el análisis estadístico de los datos de estabilidad. Objetivos: en este trabajo se buscó mostrar la conveniencia de utilizar un nuevo método de análisis de la información derivada de un estudio de estabilidad acelerado. Métodos: la simulación estadística se utilizó para comparar dos estimadores de vida útil propuestos basados en el modelo lineal mixto vs uno existente fundamentado en el modelo de regresión lineal simple según la metodología colombiana; se contempló el efecto que podría tener la variabilidad inherente al proceso de fabricación de un medicamento, considerando la variabilidad entre lote e intra lote, además, se utilizaron diferentes niveles de los factores que el investigador puede controlar en el estudio de estabilidad, como lo son el número de lotes, el número de tiempos de observación y el número de muestras por lote y por tiempo. Resultados: se determinó que los dos estimadores propuestos permiten obtener estimaciones mucho menos sesgadas que las encontradas mediante el estimador clásico, así mismo, se presentó el comportamiento de las estimaciones en función de los componentes de varianza intrínsecos del estudio y la afectación de los factores experimentales del mismo. Conclusiones: las estimaciones realizadas mediante modelos lineales mixtos presentan una exactitud mayor frente a las realizadas de manera tradicional, esta disminución en el sesgo podría mejorar la productividad de las empresas farmacéuticas sin comprometer la salud pública.Downloads
References
International Conference on Harmonization. Stability testing of new drug substances and products Q1A(R2), 2003.
República de Colombia – INVIMA. Guía para el desarrollo de estudios de estabilidad. 2006.
Komka K., Kemeny S. A modified error model for the assessment of stability of pharmaceutical products. Chemometrics and Intelligent Laboratory Systems. 2004, 72: 161-165.
Loftsson T. Drug stability for pharmaceutical scientists. Oxford: Academic Press, Elsevier; 2014. 35p.
Fitzmaurice G., Laird N., Ware J. Applied longitudinal analysis. New Jersey: Wiley & Sons; 2004. 187p.
Escobar L., Meeker W. A review of accelerated test models. Statistical Science. 2006, 21(4): 552-577.
Pinheiro J., Bates D. Mixed -Effects Models in S and S-PLUS. New York: Springer-Verlag; 2000. 528p.
Meeker W., Escobar L. Statistical methods for reliability data. New York: Wiley; 1998. 472p
R Core Team (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.
Chow S-C., Shao J. Estimating drug shelf-life with random batches. Biometrics. 1991, 47(3): 1071-1079.
Ruberg S., Hsu J. Multiple comparison procedures for pooling batches in stability. Technometrics. 1992, 34(4): 465-472.
Meeker W., Escobar L., Lu J. Accelerated degradation tests: modeling and analysis. Technometrics. 1998, 40(2): 89-99.
Escobar L., Meeker W. A review of accelerated test models. Statistics Preprints. 2006, Paper 13.
Downloads
Additional Files
Published
How to Cite
Issue
Section
License
Copyright (c) 2017 Vitae
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.
Copyright Notice and Open Access Statement
The Journal Vitae works under the Open Access license, and the published manuscripts remain available for the public, both on the Journal's website and in databases, under the Creative Commons license, "Noncommercial Attribution" and "Share alike" systems, adopted in Colombia. Hence, when the authors agree to publish in the Journal Vitae, they will not have the right to economic retributions on publications and reproductions through different diffusion media. The documents are freely available to the internet public, permitting users to read, download, copy, distribute, print, search, or link to the full texts and pass them as data to software. The only constraint on reproduction and distribution, should be to give authors control over the integrity of their work and the right to be appropriately acknowledged and cited.
Authors declare that:
-
They are the intellectual property owners and are responsible for all the information stated in the article.
-
This manuscript has not been submitted or published in other printed or digital media. They accept the responsibility for the judgments, opinions, and points of view expressed in the published article and, therefore, they exonerate Universidad de Antioquia and Journal Vitae from any process.
-
They exempt Universidad de Antioquia and Journal Vitae from settling conflicts or disputes related to the authorship of the referred article.
-
They accept the revision of the original manuscript by suitable personnel, and they bind themselves to perform the corrections appointed or suggested by the assessors.
-
Therefore, they know the editorial process and will not bind the Editorial Board of the Journal to assume any obligations regarding the volume and issue in which the article is published.
-
They transfer the rights of publication, reprinting, and distribution of the article from the moment of its approval, in print and digital format, without the right to economic rewards, and under the licensing conditions considered relevant by Journal Vitae.
-
They fully authorize Universidad de Antioquia and Journal Vitae to submit the published material to the diverse databases and indexing systems where the Journal can be found to comply with the requirements of the regulatory authorities to maintain the national classification of journals.
-
They will assume the article publication costs established for the current issue, and they will make the payment as soon as they are informed about the volume and the issue in which the final version of the article is published.
-
After the article is published, you can share digital or printed copies in a noncommercial manner. You will be able to use the paper in your institution or company for educational or research purposes, including the use in course programs.
Conflict of interest: Authors are responsible for recognizing and disclosing any financial or other benefits that could be perceived to bias their work, acknowledging all financial support and any personal connections with potential sponsors. Examples of such conflicts include receiving research funds or honoraria, serving on advisory boards, stock ownership, or employment and consulting arrangements. Authors without such connections should clearly state that they have no financial support or personal relationships that could be perceived to bias their work. All conflicts of interest should be disclosed on the author's identification page of the manuscript.
