Standard

Accuracy of approximations to recover incompletely reported logistic regression models depended on other available information. / Takada, Toshihiko; Hoogland, Jeroen; van Lieshout, Chris; Schuit, Ewoud; Collins, Gary S.; Moons, Karel G. M.; Reitsma, Johannes B.

In: Journal of clinical epidemiology, Vol. 143, 01.03.2022, p. 81-90.

Research output: Contribution to journalArticleAcademicpeer-review

Harvard

APA

Vancouver

Takada T, Hoogland J, van Lieshout C, Schuit E, Collins GS, Moons KGM et al. Accuracy of approximations to recover incompletely reported logistic regression models depended on other available information. Journal of clinical epidemiology. 2022 Mar 1;143:81-90. https://doi.org/10.1016/j.jclinepi.2021.11.033

Author

Takada, Toshihiko ; Hoogland, Jeroen ; van Lieshout, Chris ; Schuit, Ewoud ; Collins, Gary S. ; Moons, Karel G. M. ; Reitsma, Johannes B. / Accuracy of approximations to recover incompletely reported logistic regression models depended on other available information. In: Journal of clinical epidemiology. 2022 ; Vol. 143. pp. 81-90.

BibTeX

@article{3cd7cd9372634a42bf31efb64eb6c9b9,
title = "Accuracy of approximations to recover incompletely reported logistic regression models depended on other available information",
abstract = "Objective: To provide approximations to recover the full regression equation across different scenarios of incompletely reported prediction models that were developed from binary logistic regression. Study design and setting: In a case study, we considered four common scenarios and illustrated their corresponding approximations: (A) Missing: the intercept, Available: the regression coefficients of predictors, overall frequency of the outcome and descriptive statistics of the predictors; (B) Missing: regression coefficients and the intercept, Available: a simplified score; (C) Missing: regression coefficients and the intercept, Available: a nomogram; (D) Missing: regression coefficients and the intercept, Available: a web calculator. Results: In the scenario A, a simplified approach based on the predicted probability corresponding to the average linear predictor was inaccurate. An approximation based on the overall outcome frequency and an approximation of the linear predictor distribution was more accurate, however, the appropriateness of the underlying assumptions cannot be verified in practice. In the scenario B, the recovered equation was inaccurate due to rounding and categorization of risk scores. In the scenarios C and D, the full regression equation could be recovered with minimal error. Conclusion: The accuracy of the approximations in recovering the regression equation varied depending on the available information.",
keywords = "Equation, Intercept, Logistic regression, Prediction model, Reporting, Reverse engineering",
author = "Toshihiko Takada and Jeroen Hoogland and {van Lieshout}, Chris and Ewoud Schuit and Collins, {Gary S.} and Moons, {Karel G. M.} and Reitsma, {Johannes B.}",
note = "Funding Information: JH and JBR were supported by a TOP grant of the Netherlands Organisation for Health Research and Development (grant: 91215058). Funding Information: Conflict of interests: GSC was supported by the NIHR Biomedical Research Centre, Oxford, and Cancer Research UK (programme grant: C49297/A27294 ). All other authors declared no competing interests. Publisher Copyright: {\textcopyright} 2021",
year = "2022",
month = mar,
day = "1",
doi = "10.1016/j.jclinepi.2021.11.033",
language = "English",
volume = "143",
pages = "81--90",
journal = "Journal of clinical epidemiology",
issn = "0895-4356",
publisher = "Elsevier USA",

}

RIS

TY - JOUR

T1 - Accuracy of approximations to recover incompletely reported logistic regression models depended on other available information

AU - Takada, Toshihiko

AU - Hoogland, Jeroen

AU - van Lieshout, Chris

AU - Schuit, Ewoud

AU - Collins, Gary S.

AU - Moons, Karel G. M.

AU - Reitsma, Johannes B.

N1 - Funding Information: JH and JBR were supported by a TOP grant of the Netherlands Organisation for Health Research and Development (grant: 91215058). Funding Information: Conflict of interests: GSC was supported by the NIHR Biomedical Research Centre, Oxford, and Cancer Research UK (programme grant: C49297/A27294 ). All other authors declared no competing interests. Publisher Copyright: © 2021

PY - 2022/3/1

Y1 - 2022/3/1

N2 - Objective: To provide approximations to recover the full regression equation across different scenarios of incompletely reported prediction models that were developed from binary logistic regression. Study design and setting: In a case study, we considered four common scenarios and illustrated their corresponding approximations: (A) Missing: the intercept, Available: the regression coefficients of predictors, overall frequency of the outcome and descriptive statistics of the predictors; (B) Missing: regression coefficients and the intercept, Available: a simplified score; (C) Missing: regression coefficients and the intercept, Available: a nomogram; (D) Missing: regression coefficients and the intercept, Available: a web calculator. Results: In the scenario A, a simplified approach based on the predicted probability corresponding to the average linear predictor was inaccurate. An approximation based on the overall outcome frequency and an approximation of the linear predictor distribution was more accurate, however, the appropriateness of the underlying assumptions cannot be verified in practice. In the scenario B, the recovered equation was inaccurate due to rounding and categorization of risk scores. In the scenarios C and D, the full regression equation could be recovered with minimal error. Conclusion: The accuracy of the approximations in recovering the regression equation varied depending on the available information.

AB - Objective: To provide approximations to recover the full regression equation across different scenarios of incompletely reported prediction models that were developed from binary logistic regression. Study design and setting: In a case study, we considered four common scenarios and illustrated their corresponding approximations: (A) Missing: the intercept, Available: the regression coefficients of predictors, overall frequency of the outcome and descriptive statistics of the predictors; (B) Missing: regression coefficients and the intercept, Available: a simplified score; (C) Missing: regression coefficients and the intercept, Available: a nomogram; (D) Missing: regression coefficients and the intercept, Available: a web calculator. Results: In the scenario A, a simplified approach based on the predicted probability corresponding to the average linear predictor was inaccurate. An approximation based on the overall outcome frequency and an approximation of the linear predictor distribution was more accurate, however, the appropriateness of the underlying assumptions cannot be verified in practice. In the scenario B, the recovered equation was inaccurate due to rounding and categorization of risk scores. In the scenarios C and D, the full regression equation could be recovered with minimal error. Conclusion: The accuracy of the approximations in recovering the regression equation varied depending on the available information.

KW - Equation

KW - Intercept

KW - Logistic regression

KW - Prediction model

KW - Reporting

KW - Reverse engineering

UR - http://www.scopus.com/inward/record.url?scp=85122042612&partnerID=8YFLogxK

U2 - 10.1016/j.jclinepi.2021.11.033

DO - 10.1016/j.jclinepi.2021.11.033

M3 - Article

C2 - 34863904

VL - 143

SP - 81

EP - 90

JO - Journal of clinical epidemiology

JF - Journal of clinical epidemiology

SN - 0895-4356

ER -

ID: 21075559