Logotipo del repositorio

Identificación de patrones anómalos en facturación de medicamentos antibióticos: Un enfoque basado en clusterización

Vista sencilla de ítem
dc.contributor.advisorGarcia Diaz, Misael
dc.contributor.authorBurgos Garces, Angela María
dc.contributor.authorEscobar Espinosa, Sandra Beatriz
dc.date.accessioned2025-07-14T15:41:45Z
dc.date.available2025-07-14T15:41:45Z
dc.date.issued2025-06
dc.description.abstractLa facturación de medicamentos en los sistemas de salud es un área crítica propensa a actividades fraudulentas como la sobrefacturación, la prescripción inapropiada y la facturación de medicamentos no entregados, lo cual pone en riesgo tanto la sostenibilidad financiera como la calidad de la atención. Este estudio propone un modelo de detección de anomalías basado en técnicas de aprendizaje automático no supervisado, aplicado al análisis de registros de facturación de antibióticos en Colombia. Se analizó un conjunto de datos de 666.182 transacciones mediante diversos algoritmos de agrupamiento, seleccionando HDBSCAN por su capacidad para detectar clústeres de densidad variable sin requerir el número de clústeres como entrada. El proceso de detección se mejoró con Random Forest y Análisis de Valores Extremos basado en percentiles para mejorar la robustez del modelo. Los resultados indican que el 0,85% de las transacciones presentan patrones anómalos, concentrados en un conjunto limitado de proveedores y medicamentos, lo que destaca áreas de riesgo específicas. Estos hallazgos demuestran el potencial de los modelos adaptativos de detección de anomalías como herramientas estratégicas para apoyar los sistemas de auditoría de la atención médica.
dc.description.abstractenglishMedication billing in healthcare systems is a critical area prone to fraudulent activities such as overbilling, inappropriate prescribing, and invoicing for undelivered drugs, which jeopardize both financial sustainability and care quality. This study proposes an anomaly detection model based on unsupervised machine learning techniques, applied to the analysis of antibiotic billing records in Colombia. A dataset comprising 666,182 transactions was analyzed using various clustering algorithms, with HDBSCAN selected for its ability to detect variable-density clusters without requiring the number of clusters as input. The detection process was enhanced with Isolation Forest and percentile-based extreme value analysis to improve model robustness. Results indicate that 0.85% of transactions exhibit anomalous patterns, concentrated among a limited set of providers and drugs, highlighting specific risk areas. These findings demonstrate the potential of adaptive anomaly detection models as strategic tools to support healthcare auditing systems.
dc.identifier.urihttps://hdl.handle.net/20.500.12495/14942
dc.language.isoen
dc.relation.referencesD. Gallo, “Sobrecostos de más de $60.000 millones en la compra de medicamentos de la Policía denunció la Contraloría,” Infobae, Feb. 13, 2024. [Online]. Available: https://www.infobae.com/colombia/2024/02/13/contraloria-denuncio-sobrecostos-de-mas-de-60-mil-millones-de-pesos-en-la-compra-de-medicamentos-de-la-policia/. [Accessed: Apr. 12, 2025].
dc.relation.referencesJ. W. Drabiak, “What Should Health Care Organizations Do to Reduce Billing Fraud and Abuse?,” AMA J. Ethics, vol. 22, no. 3, pp. E221–E231, 2020. doi: 10.1001/amajethics.2020.221.
dc.relation.referencesY. Timofeyev et al., “Predictors of loss due to pharmaceutical fraud: Evidence from the U.S.,” Cost Eff. Resour. Alloc., vol. 20, no. 1, pp. 1–10, 2022. doi: 10.1186/s12962-022-00342-0.
dc.relation.referencesA. Coustasse et al., “Upcoding Medicare: Is healthcare fraud and abuse increasing?,” Perspect. Health Inf. Manag., vol. 18, pp. 1–10, 2021.
dc.relation.referencesN. F. Stowell, M. Schmidt, and N. Wadlinger, “Healthcare fraud under the microscope: improving its prevention,” J. Financial Crime, vol. 25, no. 4, pp. 1039–1061, 2018. doi: 10.1108/JFC-05-2017-0041.
dc.relation.referencesK. J. B. Osorno Pareja, “El rol de la auditoría en cuentas médicas dentro de los procesos de facturación del área administrativa del sector salud en Colombia,” Bachelor's thesis, Univ. de La Salle, Bogotá, Colombia, 2021.
dc.relation.referencesMinisterio de Salud y Protección Social, Colombia, “Resolución 2284 de 2023,” 2023. [Online]. Available: https://www.minsalud.gov.co/Normatividad_Nuevo/Resoluci%C3%B3n%20No%202284%20de%202023.pdf. [Accessed: Apr. 12, 2025].
dc.relation.referencesZ. A. Mansour Zoubeirou and M. Riveill, “Multiple Inputs Neural Networks for Medicare fraud Detection,” arXiv preprint arXiv:2203.05842, 2023. [Online]. Available: https://arxiv.org/abs/2203.05842. [Accessed: Apr. 12, 2025].
dc.relation.referencesV. Snorovikhina and A. Zaytsev, “Unsupervised anomaly detection for discrete sequence healthcare data,” arXiv preprint arXiv:2007.10098, 2020. [Online]. Available: https://arxiv.org/abs/2007.10098. [Accessed: Apr. 12, 2025].
dc.relation.referencesA. Kumari, N. S. Punk, S. K. Sonbhadra, and S. Agarwal, “Impact of the composition of feature extraction and class sampling in medicare fraud detection,” arXiv preprint arXiv:2206.01413, 2022. [Online]. Available: https://arxiv.org/abs/2206.01413. [Accessed: Apr. 12, 2025].
dc.relation.referencesM. H. Soleymani et al., “Detecting medical prescriptions suspected of fraud using an unsupervised data mining algorithm,” DARU J. Pharm. Sci., vol. 26, no. 1, pp. 1–9, 2018. doi: 10.1186/s40199-018-0202-1.
dc.relation.referencesA. Mehbodniya et al., “Financial fraud detection in healthcare using machine learning and deep learning techniques,” Security Commun. Netw., vol. 2021, pp. 1–12, 2021. doi: 10.1155/2021/9986655.
dc.relation.referencesM. Lokanan, “The determinants of investment fraud: A machine learning and artificial intelligence approach,” Front. Big Data, vol. 5, pp. 1–12, 2022. doi: 10.3389/fdata.2022.855792.
dc.relation.referencesO. Zapata-Cortes et al., “Machine learning models and applications for early detection,” Sensors, vol. 24, no. 1, pp. 1–15, 2024. doi: 10.3390/s24010001.
dc.relation.referencesV. Nalluri et al., “Building prediction models and discovering important factors of health insurance fraud using machine learning methods,” J. Ambient Intell. Humaniz. Comput., vol. 14, pp. 1–12, 2023. doi: 10.1007/s12652-023-04213-9.
dc.relation.referencesJ. Vajiram, N. Senthil, and N. Adhith, “Correlating Medi-Claim Service by Deep Learning Neural Networks,” arXiv preprint arXiv:2308.04469, 2020. [Online]. Available: https://arxiv.org/abs/2308.04469. [Accessed: Apr. 12, 2025].
dc.relation.referencesMinisterio de Salud y Protección Social, Colombia, “Resolución 2275 de diciembre 2023,” 2023.
dc.relation.referencesInstituto Nacional de Vigilancia de Medicamentos y Alimentos (INVIMA), Colombia, “Circular 420 de noviembre de 2006,” 2006.
dc.relation.referencesE. Montagud Penadés et al., “Implantación y análisis de Atención Farmacéutica basada en la comunicación entre Farmacéuticos Comunitarios y Médicos de Familia...,” Pharmaceutical Care España, vol. 27, e872, 2025. doi: 10.60103/phc.v27.e872.
dc.relation.referencesV. Sheth, U. Tripathi, and A. Sharma, “A Comparative Analysis of Machine Learning Algorithms for Classification Purpose,” Procedia Comput. Sci., vol. 215, pp. 422–431, 2022. doi: 10.1016/j.procs.2022.12.044.
dc.relation.referencesL. Marrec and A.-F. Bitbol, “Resist or perish: Fate of a microbial population subjected to a periodic presence of antimicrobial,” PLoS Comput. Biol., vol. 16, no. 4, e1007798, 2020. doi: 10.1371/journal.pcbi.1007798.
dc.relation.referencesWHO, “Anatomical Therapeutic Chemical (ATC) Classification,” [Online]. Available: https://www.who.int/tools/atc-ddd-toolkit/atc-classification. [Accessed: Apr. 12, 2025].
dc.relation.referencesL. Kaufman and P. J. Rousseeuw, Finding Groups in Data: An Introduction to Cluster Analysis. Hoboken, NJ, USA: Wiley, 1990. doi: 10.1002/9780470316801.
dc.relation.referencesM. Nowakowska et al., “Antibiotic choice in UK general practice: rates and drivers of potentially inappropriate antibiotic prescribing,” J. Antimicrob. Chemother., 2019. doi: 10.1093/jac/dkz345.
dc.relation.referencesF. Gómez and Á. Ruiz-Guillermo, “La influencia de la obra de Vilfredo Pareto en el análisis económico moderno...,” Iber. J. Hist. Econ. Thought, vol. 10, pp. 1–10, 2023. doi: 10.5209/ijhe.84463.
dc.relation.referencesZ. Zhu et al., “Improving access to medicines and beyond: the national volume-based procurement policy in China,” BMJ Global Health, vol. 8, no. 7, e011535, Jul. 2023. doi: 10.1136/bmjgh-2022-011535.
dc.relation.referencesD. A. Ovalle and Y. Riaño Bejarano, Pago por desempeño: El prestador de servicios de salud de cara a la negociación con las aseguradoras..., Pontificia Univ. Javeriana, Bogotá, 2019.
dc.relation.referencesH. Yu et al., “Self-paced learning for K-means clustering algorithm,” Pattern Recogn. Lett., vol. 132, pp. 69–75, 2020. doi: 10.1016/j.patrec.2018.08.028.
dc.relation.referencesK. Anbarasan, “K-Medoid Clustering Algorithm – A Review,” Academia.edu. [Online]. Available: https://www.academia.edu/8446443/K_Medoid_Clustering_Algorithm_A_Review.
dc.relation.referencesD. Deng, “DBSCAN Clustering Algorithm Based on Density,” in Proc. 7th Int. Forum on Electr. Eng. Autom. (IFEEA), Hefei, China, 2020, pp. 949–953. doi: 10.1109/IFEEA51475.2020.00199.
dc.relation.referencesM. Ankerst et al., “OPTICS: Ordering points to identify the clustering structure,” in Proc. ACM SIGMOD Int. Conf. Manage. Data, 1999, pp. 49–60. doi: 10.1145/304181.304187.
dc.relation.referencesD. Reynolds, “Gaussian Mixture Models,” [Online]. Available: http://leap.ee.iisc.ac.in/sriram/teaching/MLSP_16/refs/GMM_Tutorial_Reynolds.pdf.
dc.relation.referencesD. Dueck, “Affinity Propagation: Clustering Data by Passing Messages,” 2009. [Online]. Available: https://utoronto.scholaris.ca/items/db515c5e-ff5d-448e-8c43-1015d968a6cd/full.
dc.relation.referencesF. Ramadhani et al., “Improve BIRCH algorithm for big data clustering,” IOP Conf. Ser.: Mater. Sci. Eng., vol. 725, no. 1, p. 012090, 2020.
dc.relation.referencesC. Malzer and M. Baum, “A Hybrid Approach to Hierarchical Density-based Cluster Selection,” arXiv preprint arXiv:1911.02282, 2021.
dc.relation.referencesR. J. G. B. Campello et al., “Density-Based Clustering Based on Hierarchical Density Estimates,” in Adv. Knowl. Discov. Data Mining (PAKDD), Springer, 2013, pp. 160–172. doi: 10.1007/978-3-642-37456-2_14.
dc.relation.referencesL. McInnes, J. Healy, and S. Astels, “hdbscan: Hierarchical density based clustering,” J. Open Source Softw., vol. 2, no. 11, p. 205, 2017. doi: 10.21105/joss.00205.
dc.relation.referencesM. Das et al., “Fast rule mining over multi-dimensional windows,” in Proc. SIAM Int. Conf. Data Mining, 2011, pp. 582–593. doi: 10.1137/1.9781611972818.50.
dc.relation.referencesF. T. Liu, K. M. Ting, and Z.-H. Zhou, “Isolation Forest,” in Proc. 2008 Eighth IEEE Int. Conf. Data Mining, pp. 413–422, 2008. doi: 10.1109/ICDM.2008.17.
dc.relation.referencesK. J. Ruskin and J. P. Bliss, “Alarm Fatigue and Patient Safety,” Anesth. Patient Saf. Found. J., vol. 34, no. 1, pp. 1–28, Jun. 2019.
dc.rightsAttribution-NonCommercial-ShareAlike 4.0 Internationalen
dc.rights.urihttp://creativecommons.org/licenses/by-nc-sa/4.0/
dc.subjectAprendizaje automático
dc.subjectAprendizaje no supervisado
dc.subjectCluster
dc.subjectFraude farmacéutico
dc.subject.keywordsMachine Learning
dc.subject.keywordsUnsupervised Learning
dc.subject.keywordsClustering
dc.subject.keywordsPharmaceutical Fraud
dc.titleIdentificación de patrones anómalos en facturación de medicamentos antibióticos: Un enfoque basado en clusterización
dc.title.translatedIdentification of Anomalous Patterns in Antibiotic Medication Billing: A Clustering-Based Approach (2025)

Archivos

Bloque original

Mostrando 1 - 1 de 1
Miniatura
Nombre:
Trabajo de grado.pdf
Tamaño:
426.47 KB
Formato:
Adobe Portable Document Format
Descargar

Bloque de licencias

Mostrando 1 - 3 de 3
Miniatura
Nombre:
license.txt
Tamaño:
1.95 KB
Formato:
Item-specific license agreed upon to submission
Descripción:
Descargar
Miniatura
Nombre:
Carta de autorizacion.pdf
Tamaño:
221.98 KB
Formato:
Adobe Portable Document Format
Descripción:
Descargar
Miniatura
Nombre:
Anexo 1 Acta de aprobacion.pdf
Tamaño:
381.8 KB
Formato:
Adobe Portable Document Format
Descripción:
Descargar

Colecciones

Maestría Estadística Aplicada y Ciencia de Datos

Software DSpace copyright © 2002-2025 LYRASIS