Risk adjustment revisited using machine learning techniques

Risk adjustment is vital in health policy design. Risk adjustment defines the annual capitation payments to health insurers and is a key determinant of insolvency risk for health insurers. In this study we compare the current risk adjustment formula used by Colombia's Ministry of Health and Soc...

Full description

Autores:
Riascos Villegas, Alvaro José
Romero, Mauricio
Serna, Natalia
Tipo de recurso:
Work document
Fecha de publicación:
2017
Institución:
Universidad de los Andes
Repositorio:
Séneca: repositorio Uniandes
Idioma:
spa
OAI Identifier:
oai:repositorio.uniandes.edu.co:1992/8676
Acceso en línea:
http://hdl.handle.net/1992/8676
Palabra clave:
Risk adjustment
Diagnostic Related Groups
Risk selection
Machine learning
Seguros de salud - Colombia
Evaluación de riesgos contra la salud - Colombia
Política de salud - Colombia
I11, I13, I18, C45, C55
Rights
openAccess
License
http://creativecommons.org/licenses/by-nc-nd/4.0/
id UNIANDES2_e93b3437780c788a008980bcc28a857b
oai_identifier_str oai:repositorio.uniandes.edu.co:1992/8676
network_acronym_str UNIANDES2
network_name_str Séneca: repositorio Uniandes
repository_id_str
dc.title.none.fl_str_mv Risk adjustment revisited using machine learning techniques
dc.title.alternative.none.fl_str_mv Revisión del ajuste de riesgo con técnicas de aprendizaje de máquinas
title Risk adjustment revisited using machine learning techniques
spellingShingle Risk adjustment revisited using machine learning techniques
Risk adjustment
Diagnostic Related Groups
Risk selection
Machine learning
Seguros de salud - Colombia
Evaluación de riesgos contra la salud - Colombia
Política de salud - Colombia
I11, I13, I18, C45, C55
title_short Risk adjustment revisited using machine learning techniques
title_full Risk adjustment revisited using machine learning techniques
title_fullStr Risk adjustment revisited using machine learning techniques
title_full_unstemmed Risk adjustment revisited using machine learning techniques
title_sort Risk adjustment revisited using machine learning techniques
dc.creator.fl_str_mv Riascos Villegas, Alvaro José
Romero, Mauricio
Serna, Natalia
dc.contributor.author.none.fl_str_mv Riascos Villegas, Alvaro José
Romero, Mauricio
Serna, Natalia
dc.subject.keyword.none.fl_str_mv Risk adjustment
Diagnostic Related Groups
Risk selection
Machine learning
topic Risk adjustment
Diagnostic Related Groups
Risk selection
Machine learning
Seguros de salud - Colombia
Evaluación de riesgos contra la salud - Colombia
Política de salud - Colombia
I11, I13, I18, C45, C55
dc.subject.armarc.none.fl_str_mv Seguros de salud - Colombia
Evaluación de riesgos contra la salud - Colombia
Política de salud - Colombia
dc.subject.jel.none.fl_str_mv I11, I13, I18, C45, C55
description Risk adjustment is vital in health policy design. Risk adjustment defines the annual capitation payments to health insurers and is a key determinant of insolvency risk for health insurers. In this study we compare the current risk adjustment formula used by Colombia's Ministry of Health and Social Protection against alternative specifications that adjust for additional factors. We show that the current risk adjustment formula, which conditions on demographic factors and their interactions, can only predict 30% of total health expenditures in the upper quintile of the expenditure distribution. We also show the government's formula can improve significantly by conditioning ex ante on measures indicators of 29 long-term diseases. We contribute to the risk adjustment literature by estimating machine learning based models and showing non-parametric methodologies (e.g., boosted trees models) outperform linear regressions even when fitted in a smaller set of regressors.
publishDate 2017
dc.date.issued.none.fl_str_mv 2017
dc.date.accessioned.none.fl_str_mv 2018-09-27T16:55:45Z
dc.date.available.none.fl_str_mv 2018-09-27T16:55:45Z
dc.type.spa.fl_str_mv Documento de trabajo
dc.type.coarversion.fl_str_mv http://purl.org/coar/version/c_970fb48d4fbd8a85
dc.type.driver.spa.fl_str_mv info:eu-repo/semantics/workingPaper
dc.type.coar.spa.fl_str_mv http://purl.org/coar/resource_type/c_8042
dc.type.content.spa.fl_str_mv Text
dc.type.redcol.spa.fl_str_mv https://purl.org/redcol/resource_type/WP
format http://purl.org/coar/resource_type/c_8042
dc.identifier.issn.none.fl_str_mv 1657-5334
dc.identifier.uri.none.fl_str_mv http://hdl.handle.net/1992/8676
dc.identifier.eissn.none.fl_str_mv 1657-7191
dc.identifier.doi.none.fl_str_mv 10.57784/1992/8676
dc.identifier.instname.spa.fl_str_mv instname:Universidad de los Andes
dc.identifier.reponame.spa.fl_str_mv reponame:Repositorio Institucional Séneca
dc.identifier.repourl.spa.fl_str_mv repourl:https://repositorio.uniandes.edu.co/
identifier_str_mv 1657-5334
1657-7191
10.57784/1992/8676
instname:Universidad de los Andes
reponame:Repositorio Institucional Séneca
repourl:https://repositorio.uniandes.edu.co/
url http://hdl.handle.net/1992/8676
dc.language.iso.none.fl_str_mv spa
language spa
dc.relation.ispartofseries.none.fl_str_mv Documentos CEDE No. 27 Marzo de 2017
dc.relation.repec.spa.fl_str_mv https://ideas.repec.org/p/col/000089/015601.html
dc.rights.uri.*.fl_str_mv http://creativecommons.org/licenses/by-nc-nd/4.0/
dc.rights.accessrights.spa.fl_str_mv info:eu-repo/semantics/openAccess
dc.rights.coar.spa.fl_str_mv http://purl.org/coar/access_right/c_abf2
rights_invalid_str_mv http://creativecommons.org/licenses/by-nc-nd/4.0/
http://purl.org/coar/access_right/c_abf2
eu_rights_str_mv openAccess
dc.format.extent.none.fl_str_mv 19 páginas
dc.format.mimetype.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Universidad de los Andes, Facultad de Economía, CEDE
publisher.none.fl_str_mv Universidad de los Andes, Facultad de Economía, CEDE
institution Universidad de los Andes
bitstream.url.fl_str_mv https://repositorio.uniandes.edu.co/bitstreams/bfcd0faf-23de-4f58-b915-8568643c3233/download
https://repositorio.uniandes.edu.co/bitstreams/0a9080cf-c5b9-408c-99d8-5de8024bd2bd/download
https://repositorio.uniandes.edu.co/bitstreams/82d474e0-da62-4e4d-8401-a0aeebfdbd95/download
bitstream.checksum.fl_str_mv 6b5ced765e3d82f292a225affbc18ddb
1c67554b832d579f299db94b105e805b
0741de651b4bdfbb9779fa4ea490be14
bitstream.checksumAlgorithm.fl_str_mv MD5
MD5
MD5
repository.name.fl_str_mv Repositorio institucional Séneca
repository.mail.fl_str_mv adminrepositorio@uniandes.edu.co
_version_ 1812133820824027136
spelling Al consultar y hacer uso de este recurso, está aceptando las condiciones de uso establecidas por los autores.http://creativecommons.org/licenses/by-nc-nd/4.0/info:eu-repo/semantics/openAccesshttp://purl.org/coar/access_right/c_abf2Riascos Villegas, Alvaro José9d5edd1a-2800-42af-b5b7-6cd5baeac192500Romero, Mauriciod63643dd-3340-4e4c-a638-03fc8e702a58500Serna, Nataliadb5c992a-c714-4729-bbc2-587809d00acf5002018-09-27T16:55:45Z2018-09-27T16:55:45Z20171657-5334http://hdl.handle.net/1992/86761657-719110.57784/1992/8676instname:Universidad de los Andesreponame:Repositorio Institucional Sénecarepourl:https://repositorio.uniandes.edu.co/Risk adjustment is vital in health policy design. Risk adjustment defines the annual capitation payments to health insurers and is a key determinant of insolvency risk for health insurers. In this study we compare the current risk adjustment formula used by Colombia's Ministry of Health and Social Protection against alternative specifications that adjust for additional factors. We show that the current risk adjustment formula, which conditions on demographic factors and their interactions, can only predict 30% of total health expenditures in the upper quintile of the expenditure distribution. We also show the government's formula can improve significantly by conditioning ex ante on measures indicators of 29 long-term diseases. We contribute to the risk adjustment literature by estimating machine learning based models and showing non-parametric methodologies (e.g., boosted trees models) outperform linear regressions even when fitted in a smaller set of regressors.El ajuste de riesgo es un componente esencial en el diseño de la política del sector de la salud. El ajuste de riesgo define los pagos de capitación que se le hacen a las aseguradoras en salud y determina el riesgo de insolvencia de las mismas. En este estudio comparamos la formula usada actualmente por el Ministerio de Salud y Protección Social para determinar el ajuste de riesgo, contra especificaciones alternativas que controlan por factores de riesgo adicionales. Mostramos que la formula actual, la cual ajusta los pagos solamente a factores de riesgo demográficos, predice tan solo el 30 % del gasto en el quintil superior de la distribución del gasto. Nuestros resultados muestran que incorporar indicadores de 29 enfermedades de larga duración en un modelo lineal, como el que usa el gobierno, mejora su capacidad predictiva considerablemente. Finalmente, contribuimos a la literatura de ajuste de riesgo a través de la estimación de modelos de aprendizaje de máquinas y mostramos que modelos no paramétricos (e.g., boosting de arboles) tienen un mejor desempeño que los modelos lineales, incluso cuando se ajustan sobre un conjunto de regresores más pequeño.19 páginasapplication/pdfspaUniversidad de los Andes, Facultad de Economía, CEDEDocumentos CEDE No. 27 Marzo de 2017https://ideas.repec.org/p/col/000089/015601.htmlRisk adjustment revisited using machine learning techniquesRevisión del ajuste de riesgo con técnicas de aprendizaje de máquinasDocumento de trabajoinfo:eu-repo/semantics/workingPaperhttp://purl.org/coar/resource_type/c_8042http://purl.org/coar/version/c_970fb48d4fbd8a85Texthttps://purl.org/redcol/resource_type/WPRisk adjustmentDiagnostic Related GroupsRisk selectionMachine learningSeguros de salud - ColombiaEvaluación de riesgos contra la salud - ColombiaPolítica de salud - ColombiaI11, I13, I18, C45, C55Facultad de EconomíaPublicationORIGINALdcede2017-27.pdfdcede2017-27.pdfapplication/pdf554715https://repositorio.uniandes.edu.co/bitstreams/bfcd0faf-23de-4f58-b915-8568643c3233/download6b5ced765e3d82f292a225affbc18ddbMD51THUMBNAILdcede2017-27.pdf.jpgdcede2017-27.pdf.jpgIM Thumbnailimage/jpeg10326https://repositorio.uniandes.edu.co/bitstreams/0a9080cf-c5b9-408c-99d8-5de8024bd2bd/download1c67554b832d579f299db94b105e805bMD55TEXTdcede2017-27.pdf.txtdcede2017-27.pdf.txtExtracted texttext/plain53617https://repositorio.uniandes.edu.co/bitstreams/82d474e0-da62-4e4d-8401-a0aeebfdbd95/download0741de651b4bdfbb9779fa4ea490be14MD541992/8676oai:repositorio.uniandes.edu.co:1992/86762024-06-04 15:16:43.71http://creativecommons.org/licenses/by-nc-nd/4.0/open.accesshttps://repositorio.uniandes.edu.coRepositorio institucional Sénecaadminrepositorio@uniandes.edu.co