PMID- 8668867
OWN - NLM
STAT- MEDLINE
DCOM- 19960806
LR  - 20230808
IS  - 0277-6715 (Print)
IS  - 0277-6715 (Linking)
VI  - 15
IP  - 4
DP  - 1996 Feb 28
TI  - Multivariable prognostic models: issues in developing models, evaluating 
      assumptions and adequacy, and measuring and reducing errors.
PG  - 361-87
AB  - Multivariable regression models are powerful tools that are used frequently in 
      studies of clinical outcomes. These models can use a mixture of categorical and 
      continuous variables and can handle partially observed (censored) responses. 
      However, uncritical application of modelling techniques can result in models that 
      poorly fit the dataset at hand, or, even more likely, inaccurately predict 
      outcomes on new subjects. One must know how to measure qualities of a model's fit 
      in order to avoid poorly fitted or overfitted models. Measurement of predictive 
      accuracy can be difficult for survival time data in the presence of censoring. We 
      discuss an easily interpretable index of predictive discrimination as well as 
      methods for assessing calibration of predicted survival probabilities. Both types 
      of predictive accuracy should be unbiasedly validated using bootstrapping or 
      cross-validation, before using predictions in a new data series. We discuss some 
      of the hazards of poorly fitted and overfitted regression models and present one 
      modelling strategy that avoids many of the problems discussed. The methods 
      described are applicable to all regression models, but are particularly needed 
      for binary, ordinal, and time-to-event outcomes. Methods are illustrated with a 
      survival analysis in prostate cancer using Cox regression.
FAU - Harrell, F E Jr
AU  - Harrell FE Jr
AD  - Division of Biometry, Duke University Medical Center, Durham, North Carolina 
      27710, USA.
FAU - Lee, K L
AU  - Lee KL
FAU - Mark, D B
AU  - Mark DB
LA  - eng
GR  - HL-17670/HL/NHLBI NIH HHS/United States
GR  - HL-29436/HL/NHLBI NIH HHS/United States
GR  - HL-36587/HL/NHLBI NIH HHS/United States
GR  - etc.
PT  - Journal Article
PT  - Research Support, Non-U.S. Gov't
PT  - Research Support, U.S. Gov't, P.H.S.
PT  - Review
PL  - England
TA  - Stat Med
JT  - Statistics in medicine
JID - 8215016
SB  - IM
MH  - Clinical Trials as Topic/methods
MH  - Computer Graphics
MH  - Computer Simulation
MH  - Data Interpretation, Statistical
MH  - Discriminant Analysis
MH  - Humans
MH  - Linear Models
MH  - Male
MH  - Mathematical Computing
MH  - *Models, Statistical
MH  - Multivariate Analysis
MH  - Prostatic Neoplasms/drug therapy/mortality
MH  - Regression Analysis
MH  - Software
MH  - Survival Analysis
MH  - *Treatment Outcome
RF  - 72
EDAT- 1996/02/28 00:00
MHDA- 2000/06/20 09:00
CRDT- 1996/02/28 00:00
PHST- 1996/02/28 00:00 [pubmed]
PHST- 2000/06/20 09:00 [medline]
PHST- 1996/02/28 00:00 [entrez]
AID - 10.1002/(SICI)1097-0258(19960229)15:4<361::AID-SIM168>3.0.CO;2-4 [pii]
AID - 10.1002/(SICI)1097-0258(19960229)15:4<361::AID-SIM168>3.0.CO;2-4 [doi]
PST - ppublish
SO  - Stat Med. 1996 Feb 28;15(4):361-87. doi: 
      10.1002/(SICI)1097-0258(19960229)15:4<361::AID-SIM168>3.0.CO;2-4.