Models solely using claims-based administrative data are poor predictors of rheumatoid arthritis disease activity. Sauer, B. C., Teng, C., Accortt, N. A., Burningham, Z., Collier, D., Trivedi, M., & Cannon, G. W. Arthritis research & therapy, 19(1):86, May, 2017. Place: Englanddoi abstract bibtex BACKGROUND: This study developed and validated a claims-based statistical model to predict rheumatoid arthritis (RA) disease activity, measured by the 28-joint count Disease Activity Score (DAS28). METHOD: Veterans enrolled in the Veterans Affairs Rheumatoid Arthritis (VARA) registry with one year of data available for review before being assessed by the DAS28, were studied. Three models were developed based on initial selection of variables for analyses. The first model was based on clinically defined variables, the second leveraged grouping systems for high dimensional data and the third approach prescreened all possible predictors based on a significant bivariate association with the DAS28. The least absolute shrinkage and selection operator (LASSO) with fivefold cross-validation was used for variable selection and model development. Models were also compared for patients with \textless5 years to those ≥5 years of RA disease. Classification accuracy was examined for remission (DAS28 \textless 2.6) and for low (2.6-3.1), moderate (3.2-5.1) and high (\textgreater5.1) activity. RESULTS: There were 1582 Veterans who fulfilled inclusion criteria. The adjusted r-square for the three models tested ranged from 0.221 to 0.223. The models performed slightly better for patients with \textless5 years of RA disease than for patients with ≥5 years of RA disease. Correct classification of DAS28 categories ranged from 39.9% to 40.5% for the three models. CONCLUSION: The multiple models tested showed weak overall predictive accuracy in measuring DAS28. The models performed poorly at predicting patients with remission and high disease activity. Future research should investigate components of disease activity measures directly from medical records and incorporate additional laboratory and other clinical data.
@article{sauer_models_2017,
title = {Models solely using claims-based administrative data are poor predictors of rheumatoid arthritis disease activity.},
volume = {19},
issn = {1478-6362 1478-6354},
doi = {10.1186/s13075-017-1294-0},
abstract = {BACKGROUND: This study developed and validated a claims-based statistical model to predict rheumatoid arthritis (RA) disease activity, measured by the 28-joint count Disease Activity Score (DAS28). METHOD: Veterans enrolled in the Veterans Affairs Rheumatoid Arthritis (VARA) registry with one year of data available for review before being assessed by the DAS28, were studied. Three models were developed based on initial selection of variables for analyses. The first model was based on clinically defined variables, the second leveraged grouping systems for high dimensional data and the third approach prescreened all possible predictors based on a significant bivariate association with the DAS28. The least absolute shrinkage and selection operator (LASSO) with fivefold cross-validation was used for variable selection and model development. Models were also compared for patients with {\textless}5 years to those ≥5 years of RA disease. Classification accuracy was examined for remission (DAS28 {\textless} 2.6) and for low (2.6-3.1), moderate (3.2-5.1) and high ({\textgreater}5.1) activity. RESULTS: There were 1582 Veterans who fulfilled inclusion criteria. The adjusted r-square for the three models tested ranged from 0.221 to 0.223. The models performed slightly better for patients with {\textless}5 years of RA disease than for patients with ≥5 years of RA disease. Correct classification of DAS28 categories ranged from 39.9\% to 40.5\% for the three models. CONCLUSION: The multiple models tested showed weak overall predictive accuracy in measuring DAS28. The models performed poorly at predicting patients with remission and high disease activity. Future research should investigate components of disease activity measures directly from medical records and incorporate additional laboratory and other clinical data.},
language = {eng},
number = {1},
journal = {Arthritis research \& therapy},
author = {Sauer, Brian C. and Teng, Chia-Chen and Accortt, Neil A. and Burningham, Zachary and Collier, David and Trivedi, Mona and Cannon, Grant W.},
month = may,
year = {2017},
pmid = {28482933},
pmcid = {PMC5422885},
note = {Place: England},
keywords = {Aged, Humans, Male, Middle Aged, United States/epidemiology, Female, Veterans/*statistics \& numerical data, *Disease Progression, Arthritis, Rheumatoid/*diagnosis/*epidemiology, Disease activity, Insurance Claim Review/*statistics \& numerical data/trends, Models, Statistical, Predictive Value of Tests, Prospective Studies, Registries/statistics \& numerical data, Rheumatoid arthritis, Statistical methods, United States Department of Veterans Affairs/*statistics \& numerical data/trends},
pages = {86},
}
Downloads: 0
{"_id":"nLxe85zMhwXDavX3c","bibbaseid":"sauer-teng-accortt-burningham-collier-trivedi-cannon-modelssolelyusingclaimsbasedadministrativedataarepoorpredictorsofrheumatoidarthritisdiseaseactivity-2017","author_short":["Sauer, B. C.","Teng, C.","Accortt, N. A.","Burningham, Z.","Collier, D.","Trivedi, M.","Cannon, G. W."],"bibdata":{"bibtype":"article","type":"article","title":"Models solely using claims-based administrative data are poor predictors of rheumatoid arthritis disease activity.","volume":"19","issn":"1478-6362 1478-6354","doi":"10.1186/s13075-017-1294-0","abstract":"BACKGROUND: This study developed and validated a claims-based statistical model to predict rheumatoid arthritis (RA) disease activity, measured by the 28-joint count Disease Activity Score (DAS28). METHOD: Veterans enrolled in the Veterans Affairs Rheumatoid Arthritis (VARA) registry with one year of data available for review before being assessed by the DAS28, were studied. Three models were developed based on initial selection of variables for analyses. The first model was based on clinically defined variables, the second leveraged grouping systems for high dimensional data and the third approach prescreened all possible predictors based on a significant bivariate association with the DAS28. The least absolute shrinkage and selection operator (LASSO) with fivefold cross-validation was used for variable selection and model development. Models were also compared for patients with \\textless5 years to those ≥5 years of RA disease. Classification accuracy was examined for remission (DAS28 \\textless 2.6) and for low (2.6-3.1), moderate (3.2-5.1) and high (\\textgreater5.1) activity. RESULTS: There were 1582 Veterans who fulfilled inclusion criteria. The adjusted r-square for the three models tested ranged from 0.221 to 0.223. The models performed slightly better for patients with \\textless5 years of RA disease than for patients with ≥5 years of RA disease. Correct classification of DAS28 categories ranged from 39.9% to 40.5% for the three models. CONCLUSION: The multiple models tested showed weak overall predictive accuracy in measuring DAS28. The models performed poorly at predicting patients with remission and high disease activity. Future research should investigate components of disease activity measures directly from medical records and incorporate additional laboratory and other clinical data.","language":"eng","number":"1","journal":"Arthritis research & therapy","author":[{"propositions":[],"lastnames":["Sauer"],"firstnames":["Brian","C."],"suffixes":[]},{"propositions":[],"lastnames":["Teng"],"firstnames":["Chia-Chen"],"suffixes":[]},{"propositions":[],"lastnames":["Accortt"],"firstnames":["Neil","A."],"suffixes":[]},{"propositions":[],"lastnames":["Burningham"],"firstnames":["Zachary"],"suffixes":[]},{"propositions":[],"lastnames":["Collier"],"firstnames":["David"],"suffixes":[]},{"propositions":[],"lastnames":["Trivedi"],"firstnames":["Mona"],"suffixes":[]},{"propositions":[],"lastnames":["Cannon"],"firstnames":["Grant","W."],"suffixes":[]}],"month":"May","year":"2017","pmid":"28482933","pmcid":"PMC5422885","note":"Place: England","keywords":"Aged, Humans, Male, Middle Aged, United States/epidemiology, Female, Veterans/*statistics & numerical data, *Disease Progression, Arthritis, Rheumatoid/*diagnosis/*epidemiology, Disease activity, Insurance Claim Review/*statistics & numerical data/trends, Models, Statistical, Predictive Value of Tests, Prospective Studies, Registries/statistics & numerical data, Rheumatoid arthritis, Statistical methods, United States Department of Veterans Affairs/*statistics & numerical data/trends","pages":"86","bibtex":"@article{sauer_models_2017,\n\ttitle = {Models solely using claims-based administrative data are poor predictors of rheumatoid arthritis disease activity.},\n\tvolume = {19},\n\tissn = {1478-6362 1478-6354},\n\tdoi = {10.1186/s13075-017-1294-0},\n\tabstract = {BACKGROUND: This study developed and validated a claims-based statistical model to predict rheumatoid arthritis (RA) disease activity, measured by the 28-joint count Disease Activity Score (DAS28). METHOD: Veterans enrolled in the Veterans Affairs Rheumatoid Arthritis (VARA) registry with one year of data available for review before being assessed by the DAS28, were studied. Three models were developed based on initial selection of variables for analyses. The first model was based on clinically defined variables, the second leveraged grouping systems for high dimensional data and the third approach prescreened all possible predictors based on a significant bivariate association with the DAS28. The least absolute shrinkage and selection operator (LASSO) with fivefold cross-validation was used for variable selection and model development. Models were also compared for patients with {\\textless}5 years to those ≥5 years of RA disease. Classification accuracy was examined for remission (DAS28 {\\textless} 2.6) and for low (2.6-3.1), moderate (3.2-5.1) and high ({\\textgreater}5.1) activity. RESULTS: There were 1582 Veterans who fulfilled inclusion criteria. The adjusted r-square for the three models tested ranged from 0.221 to 0.223. The models performed slightly better for patients with {\\textless}5 years of RA disease than for patients with ≥5 years of RA disease. Correct classification of DAS28 categories ranged from 39.9\\% to 40.5\\% for the three models. CONCLUSION: The multiple models tested showed weak overall predictive accuracy in measuring DAS28. The models performed poorly at predicting patients with remission and high disease activity. Future research should investigate components of disease activity measures directly from medical records and incorporate additional laboratory and other clinical data.},\n\tlanguage = {eng},\n\tnumber = {1},\n\tjournal = {Arthritis research \\& therapy},\n\tauthor = {Sauer, Brian C. and Teng, Chia-Chen and Accortt, Neil A. and Burningham, Zachary and Collier, David and Trivedi, Mona and Cannon, Grant W.},\n\tmonth = may,\n\tyear = {2017},\n\tpmid = {28482933},\n\tpmcid = {PMC5422885},\n\tnote = {Place: England},\n\tkeywords = {Aged, Humans, Male, Middle Aged, United States/epidemiology, Female, Veterans/*statistics \\& numerical data, *Disease Progression, Arthritis, Rheumatoid/*diagnosis/*epidemiology, Disease activity, Insurance Claim Review/*statistics \\& numerical data/trends, Models, Statistical, Predictive Value of Tests, Prospective Studies, Registries/statistics \\& numerical data, Rheumatoid arthritis, Statistical methods, United States Department of Veterans Affairs/*statistics \\& numerical data/trends},\n\tpages = {86},\n}\n\n","author_short":["Sauer, B. C.","Teng, C.","Accortt, N. A.","Burningham, Z.","Collier, D.","Trivedi, M.","Cannon, G. W."],"key":"sauer_models_2017","id":"sauer_models_2017","bibbaseid":"sauer-teng-accortt-burningham-collier-trivedi-cannon-modelssolelyusingclaimsbasedadministrativedataarepoorpredictorsofrheumatoidarthritisdiseaseactivity-2017","role":"author","urls":{},"keyword":["Aged","Humans","Male","Middle Aged","United States/epidemiology","Female","Veterans/*statistics & numerical data","*Disease Progression","Arthritis","Rheumatoid/*diagnosis/*epidemiology","Disease activity","Insurance Claim Review/*statistics & numerical data/trends","Models","Statistical","Predictive Value of Tests","Prospective Studies","Registries/statistics & numerical data","Rheumatoid arthritis","Statistical methods","United States Department of Veterans Affairs/*statistics & numerical data/trends"],"metadata":{"authorlinks":{}},"html":""},"bibtype":"article","biburl":"https://bibbase.org/network/files/RyxHRMzXC8JQ3ifpb","dataSources":["YrmuDu8kXyyMRQSAd","Paz5SQAsutNo54mvh"],"keywords":["aged","humans","male","middle aged","united states/epidemiology","female","veterans/*statistics & numerical data","*disease progression","arthritis","rheumatoid/*diagnosis/*epidemiology","disease activity","insurance claim review/*statistics & numerical data/trends","models","statistical","predictive value of tests","prospective studies","registries/statistics & numerical data","rheumatoid arthritis","statistical methods","united states department of veterans affairs/*statistics & numerical data/trends"],"search_terms":["models","solely","using","claims","based","administrative","data","poor","predictors","rheumatoid","arthritis","disease","activity","sauer","teng","accortt","burningham","collier","trivedi","cannon"],"title":"Models solely using claims-based administrative data are poor predictors of rheumatoid arthritis disease activity.","year":2017}