SlideShare ist ein Scribd-Unternehmen logo
1 von 43
Downloaden Sie, um offline zu lesen
Clinical prediction models
Regression analyses, calibration, internal and external validation
Maarten van Smeden, PhD
Leiden University Medical Center
The Netherlands
Department of Clinical Epidemiology
15 Feb 2019
Clinical Trials and Research Methodology in Cardiology Meeting
Turkish Society of Cardiology
Istanbul
Sides available: https://www.slideshare.net/MaartenvanSmeden
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Prediction algorithms
SPAM
• $20 billion annual expenses
• Spam filter algorithms since 1990s
• Prediction (!)
• Noise filter function
• Dominated by classification
Rao, Journal of Economic Perspectives, 2012. doi: 10.1257/jep.26.3.87
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Prediction algorithms
SPAM
• $20 billion annual expenses
• Spam filter algorithms since 1990s
• Prediction (!)
• Noise filter function
• Dominated by classification
Healthcare
• $7, 200 billion expenses (WHO, 2015)
• Algorithms since at least 1950s
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Apgar score
Apgar, JAMA, 1958. doi: 10.1001/jama.1958.03000150027007
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Prediction algorithms
SPAM
• $20 billion annual expenses globally
• Spam filter algorithms since 1990s
• Prediction (!)
• ”Noise” filtering function
• Dominated by classification
Healthcare
• $7, 200 billion expenses (WHO, 2015)
• Algorithms popularized since 1950s
• Prediction (!)
• Informing medical decision making
• Dominated by explicit risk prediction
It is important to distinguish [risk] prediction and classification.
In many decision making contexts, classification represents a
premature decision, because classification combines prediction
and decision making..."
Frank Harrell
source: http://www.fharrell.com/post/classification/ (last accessed: Feb 12, 2019); bold-facing and [risk] added for clarity
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Risk estimation example: SCORE
10 year fatal cardiovascular disease risk
Conroy, European Heart Journal, 2003. doi: 10.1016/S0195-668X(03)00114-3
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Development of a risk prediction model
Probability of outcome = f (predictor variables)
Pr(Y = 1) = f (X)
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Development of a risk prediction model
Pr(10 year coronary heart disease risk) = f (age, cholesterol, SBP, diabetes, smoking)
Pr(Y = 1) = f (X)
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Development of a risk prediction model
Pr(10 year coronary heart disease risk) = f (age, cholesterol, SBP, diabetes, smoking)
Pr(Y = 1) = f(β1age + β2cholesterol+ β3SBP + β4diabetes + β5smoking)
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Development of a risk prediction model
Pr(10 year coronary heart disease risk) = f (age, cholesterol, SBP, diabetes, smoking)
Pr(Y = 1) = f (β1age + β2cholesterol+ β3SBP + β4diabetes + β5smoking)
These are the building blocks (simplified) of the Framingham risk score.
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Framingham risk score
10 year CVD risk
To online calculator
D’Agostino, Circulation, 2008. doi: 10.1161/CIRCULATIONAHA.107.699579
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Model specification
f (X) → linear predictor (lp)
Simplest case: lp = β0 + β1x1 + . . . + βP xP (only ”main effects”)
Note: in practice this simplest case is often too simple
linear regression
Y = lp +
logistic regression
ln{Pr(Y = 1)/(1-Pr(Y = 1))} = lp
Pr(Y = 1) = 1/(1+exp{-lp})
Cox regression
h(t)=h0(t)exp(lp)
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Logistic function
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Discrimination
• Sensitivity/specificity trade-off
• Arbitrary choice threshold → many
possible sensitivity/specificity pairs
• All pairs in 1 graph: ROC curve
• Area under the ROC-curve:
probability that a random individual
with event has a higher predicted
probability than a random individual
without event
• Area under the ROC-curve: the c-
statistic (for logistic regression) takes
on values between 0.5 (no better
than a coin-flip) and 1.0 (perfect
discrimination)
Read more: Sedgwick, BMJ, 2015, doi: 10.1136/bmj.h2464
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Calibration plot
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
The curse of statistical modeling: overfitting
What you see is not what you get1
Idiosyncrasies in the data are fitted rather than
generalizable patterns. A model may hence not be
applicable to new patients, even when the setting of
application is very similar to the development setting2
Note: prediction models are developed for new patients
1Babyak, Psychosomatic Medicine, 2004, PMID: 15184705; 2 Steyerberg, 2009, Springer, ISBN 978-0-387-77244-8.
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Overfitting artist impression
https://twitter.com/LesGuessing/status/997146590442799105
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Overfitting causes and consequences
Steyerberg, 2009, Springer, ISBN 978-0-387-77244-8.
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Overfitting: typical calibration plot
• Low probabilities are predicted too low
• high probabilities are predicted too high
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
How to avoid overfitting?
Be conservative selecting/removing variable predictor variables
• Avoid univariable, stepwise and forward selection
• When using backward elimination use conservative p-values (e.g. p = 0.10 or 0.20)
Figure: Steyerberg, JCE, 2018, doi: 10.1016/j.jclinepi.2017.11.013; Read more: Heinze, Biometrical J, 2018, doi: 10.1002/bimj.201700067
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
How to avoid overfitting?
Apply penalized regression
• Ridge regression (penalizes high regression coefficients)
• Lasso regression (penalizes high regression coefficients + automatic variable selection)
See: https://www.slideshare.net/MaartenvanSmeden/improving-predictions-lasso-ridge-and-steins-paradox-91544782
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
How to avoid overfitting?
Apply penalized regression
• Ridge regression (penalizes high regression coefficients)
• Lasso regression (penalizes high regression coefficients + automatic variable selection)
See: https://www.slideshare.net/MaartenvanSmeden/improving-predictions-lasso-ridge-and-steins-paradox-91544782
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
How to avoid overfitting?
• Adequate sample size: sufficient number of ”events” relative to number of variables
(considered) in the prediction model
• Traditional rule of thumb (10 events per variable) has been shown to have no theoretical
basis and perform poorly in simulation studies1; in many cases too lenient for development
of prediction models
• Alternative and more formal sample size calculations have recently been proposed
van Smeden, BMC med res meth, 2016, doi: 10.1186/s12874-016-0267-3
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Optimism
Optimsm
Predictive performance evaluations are too optimistic when estimated
on the same data where the risk prediction model was developed. This
is therefore called apparent performance of the model
• Optimism can be large, especially in small datasets and with a large number of predictors
• To get a better estimate of the predictive performance:
- Internal validation (same data sample)
- External validation (other data sample)
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Internal validation
• Evaluate performance of risk prediction model on data from the same population from
which model was developed
• Say that we start with one dataset with all data available: the original data
• Option 1: Splitting original data
- One portion to develop (’training set’); one portion to evaluate (’test set’)
- Non-random vs random split
- Generates 1 test of performance
• Option 2: Resampling from original data
- Cross-validation
- Bootstrapping
- Generates a distribution of performances
• General advice: avoid splitting (option 1) because
- Inefficient → especially when original data is small
- Usually leads to a too small test set
See: Steyerberg, JCE, 2001, doi: 10.1016/S0895-4356(01)00341-9
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
External validation
• Study of the predictive performance of the risk prediction model in data of new subjects
that were not used to develop it
• The larger the difference between development and validation data, the more likely the
model will be useful in (as yet) untested populations
- Case-mix (distributions of predictors and outcome)
• External validation is the strongest test of a prediction model
- Different time period (’temporal’)
- Different areas/centres (’geographical’)
- Ideally by independent investigators
See: Collins, BMJ, 2012, doi: 10.1136/bmj.e3186
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
External validation is not
It is not repeating model development steps
• Whether the same predictors, regression coefficients and predictive performance would be
found in new data is not in question
It is not re-estimating a previously developed model
• Updating regression coefficients is sometimes done when the performance at external
validation is unsatisfactory. This can be viewed as model (model revision) and calls for new
external validation
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
What to expect at external validation
• Decreased predictive performance compared to development is expected
• Many possible causes:
- Overfitting of the model at development
- Different type of patients (case mix)
- Different outcome occurrence
- Differences in care over time
- Differences in treatments
- Improvement in measurements over time (e.g.previous CTs less accurate than spiral
CT for PE detection)
- . . .
• When predictive performance is judged too low → consider model updating
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Model updating
• Recalibration in the large: re-estimate the intercept
• Recalibration: re-estimate the intercept + additional factor that multiplies all coefficients
with same factor (calibration slope)
Table from Vergouwe, Stat Med, 2017, doi: 10.1002/sim.7179
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Discrimination vs calibration
• Discrimination: extent to which risks differentiate between cases on non-cases
• Calibration: extent to which estimated risks are valid
• Discrimination is usually the no. 1 performance measure
- Risk models are typically compared on discriminative performance; not calibration
- A risk prediction model with no discriminative performance is uninformative
- A risk prediction model that is poorly calibrated is misleading
Read more: Van Calster, JCE, 2016, doi: 10.1016/j.jclinepi.2015.12.005
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Books
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Books
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
TRIPOD statement
TRIPOD, Ann Int Med, 2016, doi: 10.7326/M14-0697 and 10.7326/M14-0698
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Final remarks
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Toward machine learning and artificial intelligence?
Source: Topol, Nature Medicine, 2019. doi: 10.1038/s41591-018-0300-7
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Value of ML and AI for clinical prediction models?
A systematic review of 282 direct comparisons between machine learning and logistic regression:
”We found no evidence of superior performance of ML over LR for clinical prediction modeling...”
Christodoulou, J Clin Epi, 2019, doi: 10.1016/j.jclinepi.2019.02.004
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Do we even need new clinical prediction models at all?
• > 110 models for prostate cancer (Shariat 2008)
• > 100 models for traumatic brain injury (Perel 2006)
• 83 models for stroke (Counsell 2001)
• 54 models for breast cancer (Altman 2009)
• 43 models for type 2 diabetes (Collins 2011; Dieren 2012)
• 31 models for osteoporotic fracture (Steurer 2011)
• 29 models in reproductive medicine (Leushuis 2009)
• 26 models for hospital readmission (Kansagara 2011)
• > 25 models for length of stay in cardiac surgery (Ettema 2010)
• > 350 models for cardiovascular disease outcomes (Damen 2016)
• What if your model becomes number 300-something?
• What about the clinical benefit/utility of number 300-something?
Courtesy of KGM Moons and GS Collins for this overview
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Flow diagram
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
Flow diagram in Turkish
Courtesy of Prof Ibrahim Halil Tanboga
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
This presentation is available at https://www.slideshare.net/MaartenvanSmeden
Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019

Weitere ähnliche Inhalte

Was ist angesagt?

Survival Data Analysis for Sekolah Tinggi Ilmu Statistik Jakarta
Survival Data Analysis for Sekolah Tinggi Ilmu Statistik JakartaSurvival Data Analysis for Sekolah Tinggi Ilmu Statistik Jakarta
Survival Data Analysis for Sekolah Tinggi Ilmu Statistik JakartaSetia Pramana
 
4.5. logistic regression
4.5. logistic regression4.5. logistic regression
4.5. logistic regressionA M
 
Choosing Regression Models
Choosing Regression ModelsChoosing Regression Models
Choosing Regression ModelsStephen Senn
 
7. logistics regression using spss
7. logistics regression using spss7. logistics regression using spss
7. logistics regression using spssDr Nisha Arora
 
Correcting for missing data, measurement error and confounding
Correcting for missing data, measurement error and confoundingCorrecting for missing data, measurement error and confounding
Correcting for missing data, measurement error and confoundingMaarten van Smeden
 
Logistic regression (blyth 2006) (simplified)
Logistic regression (blyth 2006) (simplified)Logistic regression (blyth 2006) (simplified)
Logistic regression (blyth 2006) (simplified)MikeBlyth
 
Basic survival analysis
Basic survival analysisBasic survival analysis
Basic survival analysisMike LaValley
 
Multinomial Logistic Regression Analysis
Multinomial Logistic Regression AnalysisMultinomial Logistic Regression Analysis
Multinomial Logistic Regression AnalysisHARISH Kumar H R
 
Machine learning versus traditional statistical modeling and medical doctors
Machine learning versus traditional statistical modeling and medical doctorsMachine learning versus traditional statistical modeling and medical doctors
Machine learning versus traditional statistical modeling and medical doctorsMaarten van Smeden
 
Logistic regression
Logistic regressionLogistic regression
Logistic regressionDrZahid Khan
 
R workshop xiv--Survival Analysis with R
R workshop xiv--Survival Analysis with RR workshop xiv--Survival Analysis with R
R workshop xiv--Survival Analysis with RVivian S. Zhang
 
Improving epidemiological research: avoiding the statistical paradoxes and fa...
Improving epidemiological research: avoiding the statistical paradoxes and fa...Improving epidemiological research: avoiding the statistical paradoxes and fa...
Improving epidemiological research: avoiding the statistical paradoxes and fa...Maarten van Smeden
 
Five questions about artificial intelligence
Five questions about artificial intelligenceFive questions about artificial intelligence
Five questions about artificial intelligenceMaarten van Smeden
 
Machine learning in medicine: calm down
Machine learning in medicine: calm downMachine learning in medicine: calm down
Machine learning in medicine: calm downBenVanCalster
 
Rage against the machine learning 2023
Rage against the machine learning 2023Rage against the machine learning 2023
Rage against the machine learning 2023Maarten van Smeden
 
Multivariate analysis
Multivariate analysisMultivariate analysis
Multivariate analysisSubodh Khanal
 
Logistic Ordinal Regression
Logistic Ordinal RegressionLogistic Ordinal Regression
Logistic Ordinal RegressionSri Ambati
 

Was ist angesagt? (20)

Survival Data Analysis for Sekolah Tinggi Ilmu Statistik Jakarta
Survival Data Analysis for Sekolah Tinggi Ilmu Statistik JakartaSurvival Data Analysis for Sekolah Tinggi Ilmu Statistik Jakarta
Survival Data Analysis for Sekolah Tinggi Ilmu Statistik Jakarta
 
4.5. logistic regression
4.5. logistic regression4.5. logistic regression
4.5. logistic regression
 
Choosing Regression Models
Choosing Regression ModelsChoosing Regression Models
Choosing Regression Models
 
7. logistics regression using spss
7. logistics regression using spss7. logistics regression using spss
7. logistics regression using spss
 
Correcting for missing data, measurement error and confounding
Correcting for missing data, measurement error and confoundingCorrecting for missing data, measurement error and confounding
Correcting for missing data, measurement error and confounding
 
Logistic regression (blyth 2006) (simplified)
Logistic regression (blyth 2006) (simplified)Logistic regression (blyth 2006) (simplified)
Logistic regression (blyth 2006) (simplified)
 
Algorithm based medicine
Algorithm based medicineAlgorithm based medicine
Algorithm based medicine
 
Basic survival analysis
Basic survival analysisBasic survival analysis
Basic survival analysis
 
Multinomial Logistic Regression Analysis
Multinomial Logistic Regression AnalysisMultinomial Logistic Regression Analysis
Multinomial Logistic Regression Analysis
 
Machine learning versus traditional statistical modeling and medical doctors
Machine learning versus traditional statistical modeling and medical doctorsMachine learning versus traditional statistical modeling and medical doctors
Machine learning versus traditional statistical modeling and medical doctors
 
Logistic regression
Logistic regressionLogistic regression
Logistic regression
 
R workshop xiv--Survival Analysis with R
R workshop xiv--Survival Analysis with RR workshop xiv--Survival Analysis with R
R workshop xiv--Survival Analysis with R
 
Improving epidemiological research: avoiding the statistical paradoxes and fa...
Improving epidemiological research: avoiding the statistical paradoxes and fa...Improving epidemiological research: avoiding the statistical paradoxes and fa...
Improving epidemiological research: avoiding the statistical paradoxes and fa...
 
Survival analysis
Survival analysis  Survival analysis
Survival analysis
 
Binary Logistic Regression
Binary Logistic RegressionBinary Logistic Regression
Binary Logistic Regression
 
Five questions about artificial intelligence
Five questions about artificial intelligenceFive questions about artificial intelligence
Five questions about artificial intelligence
 
Machine learning in medicine: calm down
Machine learning in medicine: calm downMachine learning in medicine: calm down
Machine learning in medicine: calm down
 
Rage against the machine learning 2023
Rage against the machine learning 2023Rage against the machine learning 2023
Rage against the machine learning 2023
 
Multivariate analysis
Multivariate analysisMultivariate analysis
Multivariate analysis
 
Logistic Ordinal Regression
Logistic Ordinal RegressionLogistic Ordinal Regression
Logistic Ordinal Regression
 

Ähnlich wie Clinical prediction models

Developing and validating statistical models for clinical prediction and prog...
Developing and validating statistical models for clinical prediction and prog...Developing and validating statistical models for clinical prediction and prog...
Developing and validating statistical models for clinical prediction and prog...Evangelos Kritsotakis
 
Healthcare Predicitive Analytics for Risk Profiling in Chronic Care: A Bayesi...
Healthcare Predicitive Analytics for Risk Profiling in Chronic Care: A Bayesi...Healthcare Predicitive Analytics for Risk Profiling in Chronic Care: A Bayesi...
Healthcare Predicitive Analytics for Risk Profiling in Chronic Care: A Bayesi...MIS Quarterly
 
A plea for good methodology when developing clinical prediction models
A plea for good methodology when developing clinical prediction modelsA plea for good methodology when developing clinical prediction models
A plea for good methodology when developing clinical prediction modelsBenVanCalster
 
Therapeutic_Innovation_&_Regulatory_Science-2015-Tantsyura
Therapeutic_Innovation_&_Regulatory_Science-2015-TantsyuraTherapeutic_Innovation_&_Regulatory_Science-2015-Tantsyura
Therapeutic_Innovation_&_Regulatory_Science-2015-TantsyuraVadim Tantsyura
 
Dichotomania and other challenges for the collaborating biostatistician
Dichotomania and other challenges for the collaborating biostatisticianDichotomania and other challenges for the collaborating biostatistician
Dichotomania and other challenges for the collaborating biostatisticianLaure Wynants
 
Bias in covid 19 models
Bias in covid 19 modelsBias in covid 19 models
Bias in covid 19 modelsLaure Wynants
 
Statistics for DP Biology IA
Statistics for DP Biology IAStatistics for DP Biology IA
Statistics for DP Biology IAVeronika Garga
 
Development and evaluation of prediction models: pitfalls and solutions (Part...
Development and evaluation of prediction models: pitfalls and solutions (Part...Development and evaluation of prediction models: pitfalls and solutions (Part...
Development and evaluation of prediction models: pitfalls and solutions (Part...BenVanCalster
 
Real world modified
Real world modifiedReal world modified
Real world modifiedStephen Senn
 
Overall presentation Matram project
Overall presentation Matram project Overall presentation Matram project
Overall presentation Matram project RaphaelGirod
 
Lemeshow samplesize
Lemeshow samplesizeLemeshow samplesize
Lemeshow samplesize1joanenab
 
Sample Size Estimation and Statistical Test Selection
Sample Size Estimation  and Statistical Test SelectionSample Size Estimation  and Statistical Test Selection
Sample Size Estimation and Statistical Test SelectionVaggelis Vergoulas
 
Pacmed - Machine Learning in health care: opportunities and challanges in pra...
Pacmed - Machine Learning in health care: opportunities and challanges in pra...Pacmed - Machine Learning in health care: opportunities and challanges in pra...
Pacmed - Machine Learning in health care: opportunities and challanges in pra...BigDataExpo
 
The absence of a gold standard: a measurement error problem
The absence of a gold standard: a measurement error problemThe absence of a gold standard: a measurement error problem
The absence of a gold standard: a measurement error problemMaarten van Smeden
 
Interpreting Complex Real World Data for Pharmaceutical Research
Interpreting Complex Real World Data for Pharmaceutical ResearchInterpreting Complex Real World Data for Pharmaceutical Research
Interpreting Complex Real World Data for Pharmaceutical ResearchPaul Agapow
 
mHealth Israel_Roy Malka, Scientific Director, Sheba, Tel Ha Shomer, Innovati...
mHealth Israel_Roy Malka, Scientific Director, Sheba, Tel Ha Shomer, Innovati...mHealth Israel_Roy Malka, Scientific Director, Sheba, Tel Ha Shomer, Innovati...
mHealth Israel_Roy Malka, Scientific Director, Sheba, Tel Ha Shomer, Innovati...Levi Shapiro
 
Draft AMCP 2006 Model Quality 4-4-06
Draft AMCP 2006 Model Quality 4-4-06Draft AMCP 2006 Model Quality 4-4-06
Draft AMCP 2006 Model Quality 4-4-06Joe Gricar, MS
 

Ähnlich wie Clinical prediction models (20)

Developing and validating statistical models for clinical prediction and prog...
Developing and validating statistical models for clinical prediction and prog...Developing and validating statistical models for clinical prediction and prog...
Developing and validating statistical models for clinical prediction and prog...
 
Healthcare Predicitive Analytics for Risk Profiling in Chronic Care: A Bayesi...
Healthcare Predicitive Analytics for Risk Profiling in Chronic Care: A Bayesi...Healthcare Predicitive Analytics for Risk Profiling in Chronic Care: A Bayesi...
Healthcare Predicitive Analytics for Risk Profiling in Chronic Care: A Bayesi...
 
A plea for good methodology when developing clinical prediction models
A plea for good methodology when developing clinical prediction modelsA plea for good methodology when developing clinical prediction models
A plea for good methodology when developing clinical prediction models
 
Therapeutic_Innovation_&_Regulatory_Science-2015-Tantsyura
Therapeutic_Innovation_&_Regulatory_Science-2015-TantsyuraTherapeutic_Innovation_&_Regulatory_Science-2015-Tantsyura
Therapeutic_Innovation_&_Regulatory_Science-2015-Tantsyura
 
Dichotomania and other challenges for the collaborating biostatistician
Dichotomania and other challenges for the collaborating biostatisticianDichotomania and other challenges for the collaborating biostatistician
Dichotomania and other challenges for the collaborating biostatistician
 
Bias in covid 19 models
Bias in covid 19 modelsBias in covid 19 models
Bias in covid 19 models
 
Statistics for DP Biology IA
Statistics for DP Biology IAStatistics for DP Biology IA
Statistics for DP Biology IA
 
Biostatistics
BiostatisticsBiostatistics
Biostatistics
 
Development and evaluation of prediction models: pitfalls and solutions (Part...
Development and evaluation of prediction models: pitfalls and solutions (Part...Development and evaluation of prediction models: pitfalls and solutions (Part...
Development and evaluation of prediction models: pitfalls and solutions (Part...
 
Real world modified
Real world modifiedReal world modified
Real world modified
 
Overall presentation Matram project
Overall presentation Matram project Overall presentation Matram project
Overall presentation Matram project
 
Resrach hypothesis
Resrach hypothesisResrach hypothesis
Resrach hypothesis
 
Lemeshow samplesize
Lemeshow samplesizeLemeshow samplesize
Lemeshow samplesize
 
Sample Size Estimation and Statistical Test Selection
Sample Size Estimation  and Statistical Test SelectionSample Size Estimation  and Statistical Test Selection
Sample Size Estimation and Statistical Test Selection
 
Brmedj00047 0038
Brmedj00047 0038Brmedj00047 0038
Brmedj00047 0038
 
Pacmed - Machine Learning in health care: opportunities and challanges in pra...
Pacmed - Machine Learning in health care: opportunities and challanges in pra...Pacmed - Machine Learning in health care: opportunities and challanges in pra...
Pacmed - Machine Learning in health care: opportunities and challanges in pra...
 
The absence of a gold standard: a measurement error problem
The absence of a gold standard: a measurement error problemThe absence of a gold standard: a measurement error problem
The absence of a gold standard: a measurement error problem
 
Interpreting Complex Real World Data for Pharmaceutical Research
Interpreting Complex Real World Data for Pharmaceutical ResearchInterpreting Complex Real World Data for Pharmaceutical Research
Interpreting Complex Real World Data for Pharmaceutical Research
 
mHealth Israel_Roy Malka, Scientific Director, Sheba, Tel Ha Shomer, Innovati...
mHealth Israel_Roy Malka, Scientific Director, Sheba, Tel Ha Shomer, Innovati...mHealth Israel_Roy Malka, Scientific Director, Sheba, Tel Ha Shomer, Innovati...
mHealth Israel_Roy Malka, Scientific Director, Sheba, Tel Ha Shomer, Innovati...
 
Draft AMCP 2006 Model Quality 4-4-06
Draft AMCP 2006 Model Quality 4-4-06Draft AMCP 2006 Model Quality 4-4-06
Draft AMCP 2006 Model Quality 4-4-06
 

Mehr von Maarten van Smeden

A gentle introduction to AI for medicine
A gentle introduction to AI for medicineA gentle introduction to AI for medicine
A gentle introduction to AI for medicineMaarten van Smeden
 
Guideline for high-quality diagnostic and prognostic applications of AI in he...
Guideline for high-quality diagnostic and prognostic applications of AI in he...Guideline for high-quality diagnostic and prognostic applications of AI in he...
Guideline for high-quality diagnostic and prognostic applications of AI in he...Maarten van Smeden
 
Algorithm based medicine: old statistics wine in new machine learning bottles?
Algorithm based medicine: old statistics wine in new machine learning bottles?Algorithm based medicine: old statistics wine in new machine learning bottles?
Algorithm based medicine: old statistics wine in new machine learning bottles?Maarten van Smeden
 
Clinical prediction models for covid-19: alarming results from a living syste...
Clinical prediction models for covid-19: alarming results from a living syste...Clinical prediction models for covid-19: alarming results from a living syste...
Clinical prediction models for covid-19: alarming results from a living syste...Maarten van Smeden
 
Prediction models for diagnosis and prognosis related to COVID-19
Prediction models for diagnosis and prognosis related to COVID-19Prediction models for diagnosis and prognosis related to COVID-19
Prediction models for diagnosis and prognosis related to COVID-19Maarten van Smeden
 
Why the EPV≥10 sample size rule is rubbish and what to use instead
Why the EPV≥10 sample size rule is rubbish and what to use instead Why the EPV≥10 sample size rule is rubbish and what to use instead
Why the EPV≥10 sample size rule is rubbish and what to use instead Maarten van Smeden
 
Living systematic reviews: now and in the future
Living systematic reviews: now and in the futureLiving systematic reviews: now and in the future
Living systematic reviews: now and in the futureMaarten van Smeden
 
The statistics of the coronavirus
The statistics of the coronavirusThe statistics of the coronavirus
The statistics of the coronavirusMaarten van Smeden
 
COVID-19 related prediction models for diagnosis and prognosis - a living sys...
COVID-19 related prediction models for diagnosis and prognosis - a living sys...COVID-19 related prediction models for diagnosis and prognosis - a living sys...
COVID-19 related prediction models for diagnosis and prognosis - a living sys...Maarten van Smeden
 
ML and AI: a blessing and curse for statisticians and medical doctors
ML and AI: a blessing and curse forstatisticians and medical doctorsML and AI: a blessing and curse forstatisticians and medical doctors
ML and AI: a blessing and curse for statisticians and medical doctorsMaarten van Smeden
 
Measurement error in medical research
Measurement error in medical researchMeasurement error in medical research
Measurement error in medical researchMaarten van Smeden
 
The basics of prediction modeling
The basics of prediction modeling The basics of prediction modeling
The basics of prediction modeling Maarten van Smeden
 
Anatomy of a successful science thread
Anatomy of a successful science threadAnatomy of a successful science thread
Anatomy of a successful science threadMaarten van Smeden
 
Is it causal, is it prediction or is it neither?
Is it causal, is it prediction or is it neither?Is it causal, is it prediction or is it neither?
Is it causal, is it prediction or is it neither?Maarten van Smeden
 

Mehr von Maarten van Smeden (18)

Uncertainty in AI
Uncertainty in AIUncertainty in AI
Uncertainty in AI
 
UMC Utrecht AI Methods Lab
UMC Utrecht AI Methods LabUMC Utrecht AI Methods Lab
UMC Utrecht AI Methods Lab
 
A gentle introduction to AI for medicine
A gentle introduction to AI for medicineA gentle introduction to AI for medicine
A gentle introduction to AI for medicine
 
Associate professor lecture
Associate professor lectureAssociate professor lecture
Associate professor lecture
 
Guideline for high-quality diagnostic and prognostic applications of AI in he...
Guideline for high-quality diagnostic and prognostic applications of AI in he...Guideline for high-quality diagnostic and prognostic applications of AI in he...
Guideline for high-quality diagnostic and prognostic applications of AI in he...
 
Algorithm based medicine: old statistics wine in new machine learning bottles?
Algorithm based medicine: old statistics wine in new machine learning bottles?Algorithm based medicine: old statistics wine in new machine learning bottles?
Algorithm based medicine: old statistics wine in new machine learning bottles?
 
Clinical prediction models for covid-19: alarming results from a living syste...
Clinical prediction models for covid-19: alarming results from a living syste...Clinical prediction models for covid-19: alarming results from a living syste...
Clinical prediction models for covid-19: alarming results from a living syste...
 
Prediction models for diagnosis and prognosis related to COVID-19
Prediction models for diagnosis and prognosis related to COVID-19Prediction models for diagnosis and prognosis related to COVID-19
Prediction models for diagnosis and prognosis related to COVID-19
 
Why the EPV≥10 sample size rule is rubbish and what to use instead
Why the EPV≥10 sample size rule is rubbish and what to use instead Why the EPV≥10 sample size rule is rubbish and what to use instead
Why the EPV≥10 sample size rule is rubbish and what to use instead
 
Living systematic reviews: now and in the future
Living systematic reviews: now and in the futureLiving systematic reviews: now and in the future
Living systematic reviews: now and in the future
 
Voorspelmodellen en COVID-19
Voorspelmodellen en COVID-19Voorspelmodellen en COVID-19
Voorspelmodellen en COVID-19
 
The statistics of the coronavirus
The statistics of the coronavirusThe statistics of the coronavirus
The statistics of the coronavirus
 
COVID-19 related prediction models for diagnosis and prognosis - a living sys...
COVID-19 related prediction models for diagnosis and prognosis - a living sys...COVID-19 related prediction models for diagnosis and prognosis - a living sys...
COVID-19 related prediction models for diagnosis and prognosis - a living sys...
 
ML and AI: a blessing and curse for statisticians and medical doctors
ML and AI: a blessing and curse forstatisticians and medical doctorsML and AI: a blessing and curse forstatisticians and medical doctors
ML and AI: a blessing and curse for statisticians and medical doctors
 
Measurement error in medical research
Measurement error in medical researchMeasurement error in medical research
Measurement error in medical research
 
The basics of prediction modeling
The basics of prediction modeling The basics of prediction modeling
The basics of prediction modeling
 
Anatomy of a successful science thread
Anatomy of a successful science threadAnatomy of a successful science thread
Anatomy of a successful science thread
 
Is it causal, is it prediction or is it neither?
Is it causal, is it prediction or is it neither?Is it causal, is it prediction or is it neither?
Is it causal, is it prediction or is it neither?
 

Kürzlich hochgeladen

GenAI talk for Young at Wageningen University & Research (WUR) March 2024
GenAI talk for Young at Wageningen University & Research (WUR) March 2024GenAI talk for Young at Wageningen University & Research (WUR) March 2024
GenAI talk for Young at Wageningen University & Research (WUR) March 2024Jene van der Heide
 
Base editing, prime editing, Cas13 & RNA editing and organelle base editing
Base editing, prime editing, Cas13 & RNA editing and organelle base editingBase editing, prime editing, Cas13 & RNA editing and organelle base editing
Base editing, prime editing, Cas13 & RNA editing and organelle base editingNetHelix
 
PROJECTILE MOTION-Horizontal and Vertical
PROJECTILE MOTION-Horizontal and VerticalPROJECTILE MOTION-Horizontal and Vertical
PROJECTILE MOTION-Horizontal and VerticalMAESTRELLAMesa2
 
Vision and reflection on Mining Software Repositories research in 2024
Vision and reflection on Mining Software Repositories research in 2024Vision and reflection on Mining Software Repositories research in 2024
Vision and reflection on Mining Software Repositories research in 2024AyushiRastogi48
 
Topic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptxTopic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptxJorenAcuavera1
 
Harmful and Useful Microorganisms Presentation
Harmful and Useful Microorganisms PresentationHarmful and Useful Microorganisms Presentation
Harmful and Useful Microorganisms Presentationtahreemzahra82
 
Pests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdfPests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdfPirithiRaju
 
《Queensland毕业文凭-昆士兰大学毕业证成绩单》
《Queensland毕业文凭-昆士兰大学毕业证成绩单》《Queensland毕业文凭-昆士兰大学毕业证成绩单》
《Queensland毕业文凭-昆士兰大学毕业证成绩单》rnrncn29
 
Pests of jatropha_Bionomics_identification_Dr.UPR.pdf
Pests of jatropha_Bionomics_identification_Dr.UPR.pdfPests of jatropha_Bionomics_identification_Dr.UPR.pdf
Pests of jatropha_Bionomics_identification_Dr.UPR.pdfPirithiRaju
 
GLYCOSIDES Classification Of GLYCOSIDES Chemical Tests Glycosides
GLYCOSIDES Classification Of GLYCOSIDES  Chemical Tests GlycosidesGLYCOSIDES Classification Of GLYCOSIDES  Chemical Tests Glycosides
GLYCOSIDES Classification Of GLYCOSIDES Chemical Tests GlycosidesNandakishor Bhaurao Deshmukh
 
User Guide: Orion™ Weather Station (Columbia Weather Systems)
User Guide: Orion™ Weather Station (Columbia Weather Systems)User Guide: Orion™ Weather Station (Columbia Weather Systems)
User Guide: Orion™ Weather Station (Columbia Weather Systems)Columbia Weather Systems
 
Citronella presentation SlideShare mani upadhyay
Citronella presentation SlideShare mani upadhyayCitronella presentation SlideShare mani upadhyay
Citronella presentation SlideShare mani upadhyayupadhyaymani499
 
Environmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial BiosensorEnvironmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial Biosensorsonawaneprad
 
Ai in communication electronicss[1].pptx
Ai in communication electronicss[1].pptxAi in communication electronicss[1].pptx
Ai in communication electronicss[1].pptxsubscribeus100
 
Manassas R - Parkside Middle School 🌎🏫
Manassas R - Parkside Middle School 🌎🏫Manassas R - Parkside Middle School 🌎🏫
Manassas R - Parkside Middle School 🌎🏫qfactory1
 
bonjourmadame.tumblr.com bhaskar's girls
bonjourmadame.tumblr.com bhaskar's girlsbonjourmadame.tumblr.com bhaskar's girls
bonjourmadame.tumblr.com bhaskar's girlshansessene
 
OECD bibliometric indicators: Selected highlights, April 2024
OECD bibliometric indicators: Selected highlights, April 2024OECD bibliometric indicators: Selected highlights, April 2024
OECD bibliometric indicators: Selected highlights, April 2024innovationoecd
 
Quarter 4_Grade 8_Digestive System Structure and Functions
Quarter 4_Grade 8_Digestive System Structure and FunctionsQuarter 4_Grade 8_Digestive System Structure and Functions
Quarter 4_Grade 8_Digestive System Structure and FunctionsCharlene Llagas
 
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...Universidade Federal de Sergipe - UFS
 

Kürzlich hochgeladen (20)

AZOTOBACTER AS BIOFERILIZER.PPTX
AZOTOBACTER AS BIOFERILIZER.PPTXAZOTOBACTER AS BIOFERILIZER.PPTX
AZOTOBACTER AS BIOFERILIZER.PPTX
 
GenAI talk for Young at Wageningen University & Research (WUR) March 2024
GenAI talk for Young at Wageningen University & Research (WUR) March 2024GenAI talk for Young at Wageningen University & Research (WUR) March 2024
GenAI talk for Young at Wageningen University & Research (WUR) March 2024
 
Base editing, prime editing, Cas13 & RNA editing and organelle base editing
Base editing, prime editing, Cas13 & RNA editing and organelle base editingBase editing, prime editing, Cas13 & RNA editing and organelle base editing
Base editing, prime editing, Cas13 & RNA editing and organelle base editing
 
PROJECTILE MOTION-Horizontal and Vertical
PROJECTILE MOTION-Horizontal and VerticalPROJECTILE MOTION-Horizontal and Vertical
PROJECTILE MOTION-Horizontal and Vertical
 
Vision and reflection on Mining Software Repositories research in 2024
Vision and reflection on Mining Software Repositories research in 2024Vision and reflection on Mining Software Repositories research in 2024
Vision and reflection on Mining Software Repositories research in 2024
 
Topic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptxTopic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptx
 
Harmful and Useful Microorganisms Presentation
Harmful and Useful Microorganisms PresentationHarmful and Useful Microorganisms Presentation
Harmful and Useful Microorganisms Presentation
 
Pests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdfPests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdf
 
《Queensland毕业文凭-昆士兰大学毕业证成绩单》
《Queensland毕业文凭-昆士兰大学毕业证成绩单》《Queensland毕业文凭-昆士兰大学毕业证成绩单》
《Queensland毕业文凭-昆士兰大学毕业证成绩单》
 
Pests of jatropha_Bionomics_identification_Dr.UPR.pdf
Pests of jatropha_Bionomics_identification_Dr.UPR.pdfPests of jatropha_Bionomics_identification_Dr.UPR.pdf
Pests of jatropha_Bionomics_identification_Dr.UPR.pdf
 
GLYCOSIDES Classification Of GLYCOSIDES Chemical Tests Glycosides
GLYCOSIDES Classification Of GLYCOSIDES  Chemical Tests GlycosidesGLYCOSIDES Classification Of GLYCOSIDES  Chemical Tests Glycosides
GLYCOSIDES Classification Of GLYCOSIDES Chemical Tests Glycosides
 
User Guide: Orion™ Weather Station (Columbia Weather Systems)
User Guide: Orion™ Weather Station (Columbia Weather Systems)User Guide: Orion™ Weather Station (Columbia Weather Systems)
User Guide: Orion™ Weather Station (Columbia Weather Systems)
 
Citronella presentation SlideShare mani upadhyay
Citronella presentation SlideShare mani upadhyayCitronella presentation SlideShare mani upadhyay
Citronella presentation SlideShare mani upadhyay
 
Environmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial BiosensorEnvironmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial Biosensor
 
Ai in communication electronicss[1].pptx
Ai in communication electronicss[1].pptxAi in communication electronicss[1].pptx
Ai in communication electronicss[1].pptx
 
Manassas R - Parkside Middle School 🌎🏫
Manassas R - Parkside Middle School 🌎🏫Manassas R - Parkside Middle School 🌎🏫
Manassas R - Parkside Middle School 🌎🏫
 
bonjourmadame.tumblr.com bhaskar's girls
bonjourmadame.tumblr.com bhaskar's girlsbonjourmadame.tumblr.com bhaskar's girls
bonjourmadame.tumblr.com bhaskar's girls
 
OECD bibliometric indicators: Selected highlights, April 2024
OECD bibliometric indicators: Selected highlights, April 2024OECD bibliometric indicators: Selected highlights, April 2024
OECD bibliometric indicators: Selected highlights, April 2024
 
Quarter 4_Grade 8_Digestive System Structure and Functions
Quarter 4_Grade 8_Digestive System Structure and FunctionsQuarter 4_Grade 8_Digestive System Structure and Functions
Quarter 4_Grade 8_Digestive System Structure and Functions
 
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...
REVISTA DE BIOLOGIA E CIÊNCIAS DA TERRA ISSN 1519-5228 - Artigo_Bioterra_V24_...
 

Clinical prediction models

  • 1. Clinical prediction models Regression analyses, calibration, internal and external validation Maarten van Smeden, PhD Leiden University Medical Center The Netherlands Department of Clinical Epidemiology 15 Feb 2019 Clinical Trials and Research Methodology in Cardiology Meeting Turkish Society of Cardiology Istanbul Sides available: https://www.slideshare.net/MaartenvanSmeden Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 2. Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 3. Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 4. Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 5. Prediction algorithms SPAM • $20 billion annual expenses • Spam filter algorithms since 1990s • Prediction (!) • Noise filter function • Dominated by classification Rao, Journal of Economic Perspectives, 2012. doi: 10.1257/jep.26.3.87 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 6. Prediction algorithms SPAM • $20 billion annual expenses • Spam filter algorithms since 1990s • Prediction (!) • Noise filter function • Dominated by classification Healthcare • $7, 200 billion expenses (WHO, 2015) • Algorithms since at least 1950s Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 7. Apgar score Apgar, JAMA, 1958. doi: 10.1001/jama.1958.03000150027007 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 8. Prediction algorithms SPAM • $20 billion annual expenses globally • Spam filter algorithms since 1990s • Prediction (!) • ”Noise” filtering function • Dominated by classification Healthcare • $7, 200 billion expenses (WHO, 2015) • Algorithms popularized since 1950s • Prediction (!) • Informing medical decision making • Dominated by explicit risk prediction It is important to distinguish [risk] prediction and classification. In many decision making contexts, classification represents a premature decision, because classification combines prediction and decision making..." Frank Harrell source: http://www.fharrell.com/post/classification/ (last accessed: Feb 12, 2019); bold-facing and [risk] added for clarity Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 9. Risk estimation example: SCORE 10 year fatal cardiovascular disease risk Conroy, European Heart Journal, 2003. doi: 10.1016/S0195-668X(03)00114-3 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 10. Development of a risk prediction model Probability of outcome = f (predictor variables) Pr(Y = 1) = f (X) Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 11. Development of a risk prediction model Pr(10 year coronary heart disease risk) = f (age, cholesterol, SBP, diabetes, smoking) Pr(Y = 1) = f (X) Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 12. Development of a risk prediction model Pr(10 year coronary heart disease risk) = f (age, cholesterol, SBP, diabetes, smoking) Pr(Y = 1) = f(β1age + β2cholesterol+ β3SBP + β4diabetes + β5smoking) Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 13. Development of a risk prediction model Pr(10 year coronary heart disease risk) = f (age, cholesterol, SBP, diabetes, smoking) Pr(Y = 1) = f (β1age + β2cholesterol+ β3SBP + β4diabetes + β5smoking) These are the building blocks (simplified) of the Framingham risk score. Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 14. Framingham risk score 10 year CVD risk To online calculator D’Agostino, Circulation, 2008. doi: 10.1161/CIRCULATIONAHA.107.699579 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 15. Model specification f (X) → linear predictor (lp) Simplest case: lp = β0 + β1x1 + . . . + βP xP (only ”main effects”) Note: in practice this simplest case is often too simple linear regression Y = lp + logistic regression ln{Pr(Y = 1)/(1-Pr(Y = 1))} = lp Pr(Y = 1) = 1/(1+exp{-lp}) Cox regression h(t)=h0(t)exp(lp) Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 16. Logistic function Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 17. Discrimination • Sensitivity/specificity trade-off • Arbitrary choice threshold → many possible sensitivity/specificity pairs • All pairs in 1 graph: ROC curve • Area under the ROC-curve: probability that a random individual with event has a higher predicted probability than a random individual without event • Area under the ROC-curve: the c- statistic (for logistic regression) takes on values between 0.5 (no better than a coin-flip) and 1.0 (perfect discrimination) Read more: Sedgwick, BMJ, 2015, doi: 10.1136/bmj.h2464 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 18. Calibration plot Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 19. The curse of statistical modeling: overfitting What you see is not what you get1 Idiosyncrasies in the data are fitted rather than generalizable patterns. A model may hence not be applicable to new patients, even when the setting of application is very similar to the development setting2 Note: prediction models are developed for new patients 1Babyak, Psychosomatic Medicine, 2004, PMID: 15184705; 2 Steyerberg, 2009, Springer, ISBN 978-0-387-77244-8. Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 21. Overfitting causes and consequences Steyerberg, 2009, Springer, ISBN 978-0-387-77244-8. Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 22. Overfitting: typical calibration plot • Low probabilities are predicted too low • high probabilities are predicted too high Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 23. How to avoid overfitting? Be conservative selecting/removing variable predictor variables • Avoid univariable, stepwise and forward selection • When using backward elimination use conservative p-values (e.g. p = 0.10 or 0.20) Figure: Steyerberg, JCE, 2018, doi: 10.1016/j.jclinepi.2017.11.013; Read more: Heinze, Biometrical J, 2018, doi: 10.1002/bimj.201700067 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 24. How to avoid overfitting? Apply penalized regression • Ridge regression (penalizes high regression coefficients) • Lasso regression (penalizes high regression coefficients + automatic variable selection) See: https://www.slideshare.net/MaartenvanSmeden/improving-predictions-lasso-ridge-and-steins-paradox-91544782 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 25. How to avoid overfitting? Apply penalized regression • Ridge regression (penalizes high regression coefficients) • Lasso regression (penalizes high regression coefficients + automatic variable selection) See: https://www.slideshare.net/MaartenvanSmeden/improving-predictions-lasso-ridge-and-steins-paradox-91544782 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 26. How to avoid overfitting? • Adequate sample size: sufficient number of ”events” relative to number of variables (considered) in the prediction model • Traditional rule of thumb (10 events per variable) has been shown to have no theoretical basis and perform poorly in simulation studies1; in many cases too lenient for development of prediction models • Alternative and more formal sample size calculations have recently been proposed van Smeden, BMC med res meth, 2016, doi: 10.1186/s12874-016-0267-3 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 27. Optimism Optimsm Predictive performance evaluations are too optimistic when estimated on the same data where the risk prediction model was developed. This is therefore called apparent performance of the model • Optimism can be large, especially in small datasets and with a large number of predictors • To get a better estimate of the predictive performance: - Internal validation (same data sample) - External validation (other data sample) Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 28. Internal validation • Evaluate performance of risk prediction model on data from the same population from which model was developed • Say that we start with one dataset with all data available: the original data • Option 1: Splitting original data - One portion to develop (’training set’); one portion to evaluate (’test set’) - Non-random vs random split - Generates 1 test of performance • Option 2: Resampling from original data - Cross-validation - Bootstrapping - Generates a distribution of performances • General advice: avoid splitting (option 1) because - Inefficient → especially when original data is small - Usually leads to a too small test set See: Steyerberg, JCE, 2001, doi: 10.1016/S0895-4356(01)00341-9 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 29. External validation • Study of the predictive performance of the risk prediction model in data of new subjects that were not used to develop it • The larger the difference between development and validation data, the more likely the model will be useful in (as yet) untested populations - Case-mix (distributions of predictors and outcome) • External validation is the strongest test of a prediction model - Different time period (’temporal’) - Different areas/centres (’geographical’) - Ideally by independent investigators See: Collins, BMJ, 2012, doi: 10.1136/bmj.e3186 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 30. External validation is not It is not repeating model development steps • Whether the same predictors, regression coefficients and predictive performance would be found in new data is not in question It is not re-estimating a previously developed model • Updating regression coefficients is sometimes done when the performance at external validation is unsatisfactory. This can be viewed as model (model revision) and calls for new external validation Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 31. What to expect at external validation • Decreased predictive performance compared to development is expected • Many possible causes: - Overfitting of the model at development - Different type of patients (case mix) - Different outcome occurrence - Differences in care over time - Differences in treatments - Improvement in measurements over time (e.g.previous CTs less accurate than spiral CT for PE detection) - . . . • When predictive performance is judged too low → consider model updating Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 32. Model updating • Recalibration in the large: re-estimate the intercept • Recalibration: re-estimate the intercept + additional factor that multiplies all coefficients with same factor (calibration slope) Table from Vergouwe, Stat Med, 2017, doi: 10.1002/sim.7179 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 33. Discrimination vs calibration • Discrimination: extent to which risks differentiate between cases on non-cases • Calibration: extent to which estimated risks are valid • Discrimination is usually the no. 1 performance measure - Risk models are typically compared on discriminative performance; not calibration - A risk prediction model with no discriminative performance is uninformative - A risk prediction model that is poorly calibrated is misleading Read more: Van Calster, JCE, 2016, doi: 10.1016/j.jclinepi.2015.12.005 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 34. Books Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 35. Books Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 36. TRIPOD statement TRIPOD, Ann Int Med, 2016, doi: 10.7326/M14-0697 and 10.7326/M14-0698 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 37. Final remarks Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 38. Toward machine learning and artificial intelligence? Source: Topol, Nature Medicine, 2019. doi: 10.1038/s41591-018-0300-7 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 39. Value of ML and AI for clinical prediction models? A systematic review of 282 direct comparisons between machine learning and logistic regression: ”We found no evidence of superior performance of ML over LR for clinical prediction modeling...” Christodoulou, J Clin Epi, 2019, doi: 10.1016/j.jclinepi.2019.02.004 Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 40. Do we even need new clinical prediction models at all? • > 110 models for prostate cancer (Shariat 2008) • > 100 models for traumatic brain injury (Perel 2006) • 83 models for stroke (Counsell 2001) • 54 models for breast cancer (Altman 2009) • 43 models for type 2 diabetes (Collins 2011; Dieren 2012) • 31 models for osteoporotic fracture (Steurer 2011) • 29 models in reproductive medicine (Leushuis 2009) • 26 models for hospital readmission (Kansagara 2011) • > 25 models for length of stay in cardiac surgery (Ettema 2010) • > 350 models for cardiovascular disease outcomes (Damen 2016) • What if your model becomes number 300-something? • What about the clinical benefit/utility of number 300-something? Courtesy of KGM Moons and GS Collins for this overview Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 41. Flow diagram Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 42. Flow diagram in Turkish Courtesy of Prof Ibrahim Halil Tanboga Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019
  • 43. This presentation is available at https://www.slideshare.net/MaartenvanSmeden Twitter: @MaartenvSmeden Clinical prediction models 15 Feb 2019