Prognostic modelling with logistic regression analysis: a comparison of selection and estimation methods in small data sets

Steyerberg, Ewout W.; Eijkemans, Marinus J.C.; Harrell, Frank E.; Habbema, J. Dik F.

doi:10.1002/(sici)1097-0258(20000430)19:8<1059::aid-sim412>3.0.co;2-0

Cited by 654 publications

(369 citation statements)

References 43 publications

Supporting

Mentioning

342

Contrasting

Unclassified

Order By: Relevance

“…A direct automatic selection method was not reasonable in our case, since the number of cases of dyslexia compared to the number of variables is low (35 dyslexics/29 variables) (Steyerberg, Eijkemans, & Habbema 1999;Steyerberg, Eijkemans, Harrel, & Habbema 2000). A pre-selection among the initial variables was done using PCA in order to reduce the number of variables to be analyzed in the second step.…”

Section: Discussionmentioning

confidence: 99%

Multivariate predictive model for dyslexia diagnosis

et al. 2010

View full text Add to dashboard Cite

Dyslexia is a specific disorder of language development that mainly affects reading. Etiological researches have led to multiple hypotheses which induced various diagnosis methods and rehabilitation treatments so that many different tests are used by practitioners to identify dyslexia symptoms. Our purpose is to determine a subset of the most efficient ones by integrating them into a multivariate predictive model. A set of screening tasks that are the most commonly used and representative of the different cognitive aspects of dyslexia was proposed to 78 children from elementary school (mean age = 9 years ± 7 months) exempt from identified reading difficulties and to 35 dyslexic children attending a specialized consultation for dyslexia. We proposed a multi-step procedure: within each category, we first selected the most representative tasks using principal component analysis and then we implemented logistic regression models on the preselected variables. Spelling and reading tasks were considered separately. The model with the best predictive performance includes eight variables from four categories of tasks and classifies correctly 94% of the children. The sensitivity (91%) and the specificity (95%) are both high. Forty minutes are necessary to complete the test.

show abstract

Section: Discussionmentioning

confidence: 99%

Multivariate predictive model for dyslexia diagnosis

et al. 2010

View full text Add to dashboard Cite

show abstract

“…In this sense, the purpose was similar to recent studies [15,16] where in-depth methodological analysis has been the primary focus, rather than the elucidation of a pooled estimate of effect across an exhaustive catalogue. Original references of the studies cited in the meta-analyses were accessed and where indicated, data were transcribed for analysis.…”

Section: Methodsmentioning

confidence: 89%

Methodology in meta–analysis: a study from Critical Care meta–analytic practice

Moran

Solomon

Warn

2004

Health Serv Outcomes Res Method

View full text Add to dashboard Cite

Methodological aspects of meta-analytic practice, heterogeneity, publication bias, metaregression and effect metric, were investigated in 14 meta-analyses reflecting major therapeutic concern in Critical Care practice.Compared with the standard Q test, the exact Zelen test was more sensitive in identifying heterogeneity. Assessment of heterogeneity impact by the I 2 statistic was consistent with inferences afforded by both the Q and Zelen test. Publication bias was subject to test and metric determination: funnel plots exhibited variable asymmetry across studies and between metrics; the regression asymmetry test appeared more sensitive than the rank correlation test; the "trim and fill" method was the most sensitive, but suggested, on the basis of quantification of the effects of potentially missing studies, that meta-analyses may be resistant to such missingness. Metaregression of treatment effect against control risk using Bayesian hierarchical regression in all metrics (log odds ratio, log risk ratio and RD) suggested that naïve linear regression approaches over-diagnosed significant relationships and exhibited regression dilution.Heterogeneity, publication bias and risk related treatment effects all demonstrate estimator and metric dependence; the RD metric would appear the most capricious in this regard.

show abstract

“…Because the logistic regression analysis utilized a relatively small data set (n=79), there was a possibility that the predictive ability of the models was overestimated or overfit [19]. In order to accommodate for this possibility, further analysis which utilized a full model with all independent variables was performed to provide adjusted formulas known as shrinkage and Akaike's information criterion (AIC), both of which helped attain a more unbiased assessment [19].…”

Section: Methodsmentioning

confidence: 99%

Determining Predictors of Success on the American Board of Anesthesiology Written Certification Exam

Schmitz

Bailey

2014

Med.Sci.Educ.

View full text Add to dashboard Cite

Introduction The purpose of this research was to determine what, if any, relationship existed between in-training exams (ITE) given during the categorical years of anesthesiology residency training and first attempt performance on Part One of the American Board of Anesthesiology (ABA) certification examination (CE). Additionally, several variables available to residency program directors, to include gender, medical degree, medical school, and United States Medical Licensing Examination (USMLE) Step 1 and Step 2 CK scores, were analyzed to determine if a predictive relationship existed between any variable(s) and the Part One CE. Current research in anesthesiology only looked at PGY-II ITE scores, while other specialties have found relationships in later years of training not yet studied in anesthesiology. Methods The authors studied data from one US anesthesiology residency program. In total, there was data from the residency files of 79 residents. First attempt ABA CE results were coded as a binary dependent variable, either pass or fail. ITE scores at all three categorical years of training, as well as other available variables, were used to see if any predictive relationship exists. Results The PGY-IV ITE was found to be the only year of ITE that was statistically significant in predicting eventual first attempt performance on the ABA Part One CE. Logistic regression, utilizing the backward elimination technique, returned a model utilizing three variables: PGY-IV ITE, gender, and USMLE Step 2 CK scores. This model correctly predicted 89.9 % of eventual performance. Discussion The major finding in this study is that higher ABA/ASA ITE scores taken at the end of the final year of anesthesiology residency served as a statistically significant predictor of first attempt performance on the ABA Part One CE. Future studies should be performed at a national level to identify whether this finding is applicable to a wider population.

show abstract

Prognostic modelling with logistic regression analysis: a comparison of selection and estimation methods in small data sets

Cited by 654 publications

References 43 publications

Multivariate predictive model for dyslexia diagnosis

Multivariate predictive model for dyslexia diagnosis

Methodology in meta–analysis: a study from Critical Care meta–analytic practice

Determining Predictors of Success on the American Board of Anesthesiology Written Certification Exam

Contact Info

Product

Resources

About