Risk prediction models: I. Development, internal validation, and assessing the incremental value of a new (bio)marker

Moons, Karel G.M.; Kengne, André Pascal; Woodward, Mark; Royston, Patrick; Vergouwe, Yvonne; Altman, Douglas G.; Grobbee, Diederick E.

doi:10.1136/heartjnl-2011-301246

Cited by 722 publications

(785 citation statements)

References 94 publications

Supporting

Mentioning

773

Contrasting

Unclassified

Order By: Relevance

“…In conclusion, we have developed a predictive model for ESRD in children with CKD (27,29,36). Our model uses routinely available clinical, laboratory, and imaging data and can predict the long-term risk for renal impairment with accuracy.…”

Section: Discussionmentioning

confidence: 99%

“…The prognostic score derived was then grouped into three categories: low-risk, medium-risk, and high-risk groups. We assessed the predictive accuracy of the derived model by looking at the components of accuracy (i.e., discrimination and calibration) (27)(28)(29)(30). Discrimination was evaluated using the c statistic, which represents the area under the receiveroperating characteristic curve (for which larger values indicate better discrimination) (31).…”

Section: Statistical Analyses and Development Of Risk Prediction Modelmentioning

confidence: 99%

“…To adjust for overfitting and overoptimistic performance of the model, we performed an internal validation of our model with a bootstrapping technique (36). In each bootstrap sample, the entire modeling process was repeated, resulting in shrinkage of the regression coefficients when applicable (27,29,37).…”

Section: Statistical Analyses and Development Of Risk Prediction Modelmentioning

confidence: 99%

See 2 more Smart Citations

A Predictive Model of Progression of CKD to ESRD in a Predialysis Pediatric Interdisciplinary Program

Cerqueira¹,

Soares²,

Silva

et al. 2014

Clinical Journal of the American Society of Nephrology

View full text Add to dashboard Cite

Background and objectives The incidence of ESRD in children has increased over the last two decades. Nevertheless, there are still limited data on risk factors related to the emergence of ESRD among patients with CKD. The aim of this study was to develop a model of prediction of ESRD in children and adolescents with CKD (stages 2-4) enrolled in a predialysis interdisciplinary management program.Design, setting, participants, & measurements In this retrospective cohort study, 147 patients with CKD admitted from 1990 to 2008 were systematically followed up at a tertiary pediatric nephrology unit for a median of about 4.5 years. The primary outcome was the progression to CKD stage 5. A predictive model was developed using Cox proportional hazards model and evaluated by c statistics.Results The median renal survival was estimated at 98.7 months (95% confidence interval [95% CI], 68.7 to 129.6 months). The probability of reaching CKD stage 5 was estimated as 52% in 10 years. The most accurate model included eGFR, proteinuria at admission, and primary renal disease. Risk score ranged from 0 to 13 points (median, 4 points). The accuracy of the score applied to the sample was high, with c statistics of 0.865 (95% CI, 0.80 to 0.93) and 0.837 (95% CI, 0.76 to 0.91) at follow-up of 2 and 5 years, respectively. By survival analysis, it was estimated that at 10 years after admission, the probability of renal survival was about 63% for patients in the lowrisk group and 43% for the medium-risk group; all patients assigned to the high-risk group had CKD stage 5 (P,0.001). ConclusionThe predictive model of progression of CKD might contribute to early identification of a subgroup of patients at high risk for accelerated renal failure.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Statistical Analyses and Development Of Risk Prediction Modelmentioning

confidence: 99%

Section: Statistical Analyses and Development Of Risk Prediction Modelmentioning

confidence: 99%

See 1 more Smart Citation

A Predictive Model of Progression of CKD to ESRD in a Predialysis Pediatric Interdisciplinary Program

Cerqueira¹,

Soares²,

Silva

et al. 2014

Clinical Journal of the American Society of Nephrology

View full text Add to dashboard Cite

show abstract

“…Second, backward selection procedure with (P < 0.05) was used to choose the covariates in the final multivariable model (22). Covariates eliminated were reentered in the final multivariable model, with adjustment for the remaining significant covariates to ensure that no omitted covariate significantly reduced the log likelihood c 2 of the model (23). The performance of the multivariable model was quantified by Harrell's concordance (C) statistics, which is analogous to the receiver-operating characteristics (ROC) curve for binary data with confidence interval estimates using the jackknife resampling method (24).…”

Section: Statistical Analysesmentioning

confidence: 99%

LLPi: Liverpool Lung Project Risk Prediction Model for Lung Cancer Incidence

Marcus

Chen

Raji

et al. 2015

Cancer Prevention Research

View full text Add to dashboard Cite

Identification of high-risk individuals will facilitate early diagnosis, reduce overall costs, and also improve the current poor survival from lung cancer. The Liverpool Lung Project prospective cohort of 8,760 participants ages 45 to 79 years, recruited between 1998 and 2008, was followed annually through the hospital episode statistics until January 31, 2013. Cox proportional hazards models were used to identify risk predictors of lung cancer incidence. C-statistic was used to assess the discriminatory accuracy of the models. Models were internally validated using the bootstrap method. During mean follow-up of 8.7 years, 237 participants developed lung cancer. Age [hazard ratio (HR), 1.04; 95% confidence interval (CI), 1.02-1.06], male gender (HR, 1.48; 95% CI, 1.10-1.98), smoking duration (HR, 1.04; 95% CI, 1.03-1.05), chronic obstructive pulmonary disease (HR, 2.43; 95% CI, 1.79-3.30), prior diagnosis of malignant tumor (HR, 2.84; 95% CI, 2.08-3.89), and early onset of family history of lung cancer (HR, 1.68; 95% CI, 1.04-2.72) were associated with the incidence of lung cancer. The LLPi risk model had a good calibration (goodness-of-fit c 2 7.58, P ¼ 0.371). The apparent C-statistic was 0.852 (95% CI, 0.831-0.873) and the optimism-corrected bootstrap resampling C-statistic was 0.849 (95% CI, 0.829-0.873). The LLPi risk model may assist in identifying individuals at high risk of developing lung cancer in population-based screening programs. Cancer Prev Res; 8(6); 570-5. Ó2015 AACR.

show abstract

“…Methodological issues involved in the assessment and comparison of the performance of risk models using different biomarkers As reviewed recently [4][5][6], it is important to consider methodological issues in the development of risk prediction models. Before the incremental value of novel biomarkers for diabetes prediction can be evaluated, several specific issues related to model performance deserve attention and will be briefly summarised.…”

Section: Introductionmentioning

confidence: 99%

The potential of novel biomarkers to improve risk prediction of type 2 diabetes

et al. 2013

View full text Add to dashboard Cite

The incidence of type 2 diabetes can be reduced substantially by implementing preventive measures in highrisk individuals, but this requires prior knowledge of disease risk in the individual. Various diabetes risk models have been designed, and these have all included a similar combination of factors, such as age, sex, obesity, hypertension, lifestyle factors, family history of diabetes and metabolic traits. The accuracy of prediction models is often assessed by the area under the receiver operating characteristic curve (AROC) as a measure of discrimination, but AROCs should be complemented by measures of calibration and reclassification to estimate the incremental value of novel biomarkers. This review discusses the potential of novel biomarkers to improve model accuracy. The range of molecules that serve as potential predictors of type 2 diabetes includes genetic variants, RNA transcripts, peptides and proteins, lipids and small metabolites. Some of these biomarkers lead to a statistically significant increase of model accuracy, but their incremental value currently seems too small for routine clinical use. However, only a fraction of potentially relevant biomarkers have been assessed with regard to their predictive value. Moreover, serial measurements of biomarkers may help determine individual risk. In conclusion, current risk models provide valuable tools of risk estimation, but perform suboptimally in the prediction of individual diabetes risk. Novel biomarkers still fail to have a clinically applicable impact. However, more efficient use of biomarker data and technological advances in their measurement in clinical settings may allow the development of more accurate predictive models in the future.

show abstract

Risk prediction models: I. Development, internal validation, and assessing the incremental value of a new (bio)marker

Cited by 722 publications

References 94 publications

A Predictive Model of Progression of CKD to ESRD in a Predialysis Pediatric Interdisciplinary Program

A Predictive Model of Progression of CKD to ESRD in a Predialysis Pediatric Interdisciplinary Program

LLPi: Liverpool Lung Project Risk Prediction Model for Lung Cancer Incidence

The potential of novel biomarkers to improve risk prediction of type 2 diabetes

Contact Info

Product

Resources

About