Prognostic models
Practical NeurologyPublished 21 July 2008
Peter M. Rothwell
Citations13
SJR quartileQ2
SJR score0.65
SNIP1.04
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
How prognostic models are best derived, and how their reliability should be assessed, are considered, drawing on examples from various different neurological disorders.
Abstract
An appreciation of prognosis is essential for effective clinical practice, in neurology as in other fields. Predicting risk of a poor outcome in individuals, or at least in well-defined groups of individuals, is necessary to properly inform patients about what the future holds for them, and the likely benefits of treatment. Despite the various pitfalls in the derivation and validation of prognostic models (or scores), there are an increasing number of useful models available to help patients and clinicians in routine neurology practice. This article considers how such models are best derived, and how their reliability should be assessed, drawing on examples from various different neurological disorders.
Keywords
Medicine
Psychological MethodsMissing data: Our view of the state of the art.
10,918 Citations2002Joseph L. Schafer, J. A. Graham
2 general approaches that come highly recommended: maximum likelihood (ML) and Bayesian multiple imputation (MI) are presented and may eventually extend the ML and MI methods that currently represent the state of the art.
Statistics in MedicineMULTIVARIABLE PROGNOSTIC MODELS: ISSUES IN DEVELOPING MODELS, EVALUATING ASSUMPTIONS AND ADEQUACY, AND MEASURING AND REDUCING ERRORS
9,798 Citations1996Frank E. Harrell, Kerry L. Lee +1 more
JAMAThe TIMI Risk Score for Unstable Angina/Non–ST Elevation MI
3,246 Citations2000Elliott M. Antman, Marc Cohen +8 more
In patients with UA/NSTEMI, the TIMI risk score is a simple prognostication scheme that categorizes a patient's risk of death and ischemic events and provides a basis for therapeutic decision making.
Statistics in MedicineDichotomizing continuous predictors in multiple regression: a bad idea
2,228 Citations2005Patrick Royston, Douglas G. Altman +1 more
It is argued that the simplicity achieved is gained at a cost; dichotomization may create rather than avoid problems, notably a considerable loss of power and residual confounding.
The LancetEffect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomised trials by the Coronary Artery Bypass Graft Surgery Trialists Collaboration
2,146 Citations1994Salim Yusuf, David M. Zucker +12 more
A strategy of initial CABG surgery is associated with lower mortality than one of medical management with delayed surgery if necessary, especially in high-risk and medium- risk patients with stable coronary heart disease, and in low-risk patients, the limited data show a non-significant trend towards greater mortality with CABGs.
Archives of Internal MedicineRisk Factors for Stroke and Efficacy of Antithrombotic Therapy in Atrial Fibrillation
2,070 Citations1994
The LancetMorbidity and mortality in the Swedish Trial in Old Patients with Hypertension (STOP-Hypertension)
1,888 Citations1991Björn Dahlöf, L. Hansson +4 more
It is concluded that antihypertensive treatment in hypertensive men and women aged 70-84 confers highly significant and clinically relevant reductions in cardiovascular morbidity and mortality as well as in total mortality.
New England Journal of MedicineUnruptured Intracranial Aneurysms — Risk of Rupture and Risks of Surgical Intervention
1,875 Citations1998The International Study of Unruptured Intracranial Aneurysms Investigators
The likelihood of rupture of unruptured intracranial aneurysms that were less than 10 mm in diameter was exceedingly low among patients in group 1 and was substantially higher among those in group 2, and the risk of morbidity and mortality related to surgery greatly exceeded the 7.5-year risk of rupture.
Statistics in MedicineRegression modelling strategies for improved prognostic prediction
1,836 Citations1984Frank E. Harrell, Kerry L. Lee +3 more
A general index of predictive discrimination is used to measure the ability of a model developed on training samples of varying sizes to predict survival in an independent test sample of patients suspected of having coronary artery disease.
Psychosomatic MedicineWhat You See May Not Be What You Get: A Brief, Nontechnical Introduction to Overfitting in Regression-Type Models
1,750 Citations2004Michael A. Babyak
The notion of overfitting is presented in terms of asking too much from the available data, and three common practices—automated variable selection, pretesting of candidate predictors, and dichotomization of continuous variables—are shown to pose a considerable risk for spurious findings in models.
BMJMRC trial of treatment of mild hypertension: principal results. Medical Research Council Working Party.
1,503 Citations1985Stanley Peart, Greta Barnes +74 more
Comparison of the two active drugs showed that the reduction in stroke rate on bendrofluazide was greater than that on propranolol and this difference between the responses to the two drugs was significant.
Statistics in MedicineWhat do we mean by validating a prognostic model?
1,425 Citations2000Douglas G. Altman, Patrick Royston
How to validate a model is considered and it is suggested that it is desirable to consider two rather different aspects - statistical and clinical validity - and some general approaches to validation are examined.
The LancetValidation and refinement of scores to predict very early stroke risk after transient ischaemic attack
1,338 Citations2007S. Claiborne Johnston, Peter M. Rothwell +5 more
Existing prognostic scores for stroke risk after TIA validate well on multiple independent cohorts, but the unified ABCD(2) score is likely to be most predictive.
Annals of Internal MedicineAssessing the Generalizability of Prognostic Information
1,170 Citations1999Amy C. Justice, Kenneth E. Covinsky +1 more
Annals of Internal MedicineThe Risk of Determining Risk with Multivariable Models
1,112 Citations1993John Concato, Alvan R. Feinstein +1 more
The purpose in the current research was to note the frequency with which multivariable analyses now appear in general medical journals, to identify some common problems and desirable precautions in the analyses, and to determine how well the challenges are being met.
Journal of Clinical EpidemiologyInappropriate use of bivariable analysis to screen risk factors for use in multivariable analysis
903 Citations1996Guo‐Wen Sun, Thomas Shook +1 more
This article uses both hypothetical and actual data to show how a nonsignificant risk factors in bivariable analysis may actually be a significant risk factor in multivariableAnalysis if confounding is properly controlled.
Journal of Neurology Neurosurgery & PsychiatryMalformation risks of antiepileptic drugs in pregnancy: a prospective study from the UK Epilepsy and Pregnancy Register
873 Citations2005James Morrow
Only 4.2% of live births to women with epilepsy had an MCM, and polytherapy exposure was greater than for monotherapy exposure, and carbamazepine was associated with the lowest risk of MCM.
Annals of Internal MedicineTranslating Clinical Research into Clinical Practice: Impact of Using Prediction Rules To Make Decisions
806 Citations2006Brendan M. Reilly, Arthur T. Evans
A fifth level of evidence is proposed because it is believed that broad verification of a prediction rule's clinical impact is no less important than that of the prediction rule on which it is based, and progressive evidentiary standards emphasize that a prediction rules rises to the level of a decision rule only if clinicians use its predictions to help make decisions for patients.
JAMAClinical Prediction Rules<subtitle>A Review and Suggested Modifications of Methodological Standards</subtitle>
770 Citations1997Andreas Laupacis
Although clinical prediction rules comply with some methodological criteria, for other criteria, better compliance is needed.
CirculationSelecting Patients With Atrial Fibrillation for Anticoagulation
756 Citations2004Brian F. Gage, Carl van Walraven +5 more
Patients with AF who have high and low rates of stroke when given aspirin can be reliably identified, allowing selection of antithrombotic prophylaxis to be individualized.
The LancetA simple score (ABCD) to identify individuals at high early risk of stroke after transient ischaemic attack
627 Citations2005PM Rothwell, MF Giles +5 more
Although further validations and refinements are needed, the ABCD score can be used in routine clinical practice to identify high-risk individuals who need emergency investigation and treatment.
Journal of Neurology Neurosurgery & PsychiatrySyndromes of transient amnesia: towards a classification. A study of 153 cases.
597 Citations1990J. R. Hodges, Charles Warlow
Of 153 patients presenting with acute transient amnesia, 114 fulfilled the proposed strict diagnostic criteria for transient global amnesia (TGA), and the group of 39 patients who did not meet the criteria for TGA had a significantly worse prognosis with a high incidence of major vascular events.
StrokePredictors of Depression after Stroke
460 Citations2005Maree L. Hackett, Craig S. Anderson
There is a paucity of well-designed studies of sufficient size to allow stable multivariate predictive models of depression after stroke to be developed, and current evidence does not allow for ready identification of patients most at risk of developing this important complication of stroke.
Statistics in MedicineA new measure of prognostic separation in survival data
456 Citations2004Patrick Royston, Willi Sauerbrei
A well‐founded and reliable measure of the prognostic ability of a model would be valuable to help define the separation between patients or prognostic groups that the model could provide, and to act as a benchmark of model performance in a validation setting.
HeartAspirin for primary prevention of coronary heart disease: safety and absolute benefit related to coronary risk derived from meta-analysis of randomised trials
413 Citations2001P. S Sanmuganathan, P Ghahramani +3 more
The upper 95% CI for this estimate suggests that harm from aspirin is unlikely to outweigh benefit provided the cardiovascular event risk is 0.8%/year, equivalent to a coronary risk of 0.6%/ year.
The LancetPrediction of benefit from carotid endar terectomy in individual patients: a risk-modelling study
408 Citations1999P M Rothwell, C P Warlow
A prognostic score is developed to identify patients with a high risk of stroke on medical treatment but a low operative risk and risk-factor modelling could be useful to identify those patients in whom endarterectomy will be beneficial.
HepatologyMultivariate survival analysis using Cox's regression model
373 Citations1987Erik Christensen
The Cox model can be used to adjust for imbalance in variables influencing prognosis and to identify variables being associated with the treatment effect (therapeutic variables) in controlled clinical trials.
BMJCommentary: Prognostic models: clinically useful or quickly forgotten?
372 Citations1995Jeremy C Wyatt, D. G Altman
The main reasons why doctors reject published prognostic models are lack of clinical credibility and lack of evidence that a prognostic model can support decisions about patient care (that is, evidence of accuracy, generality, and effectiveness).
The Lancet NeurologyA clinical prognostic scoring system for Guillain-Barré syndrome
368 Citations2007Rinske van Koningsveld, Ewout W. Steyerberg +4 more
A simple scoring system for patients with GBS, based on three clinical characteristics, accurately predicts outcome at 6 months, which could be used to counsel individual patients and identify high-risk groups to guide future trials.
Annals of Internal MedicinePrevention of First Bleeding in Cirrhosis
320 Citations1992Luigi Pagliaro, Gennaro D’Amico +7 more
Beta-blockers may be recommended for prevention of first bleeding in cirrhotic patients with varices who have a high risk for bleeding and the effectiveness of sclerotherapy remains undetermined.
The Lancet NeurologyPrediction of risk of seizure recurrence after a single seizure and early epilepsy: further results from the MESS trial
320 Citations2006Lois G. Kim, Tony Johnson +2 more
BrainTHE AETIOLOGY OF TRANSIENT GLOBAL AMNESIA
291 Citations1990John R. Hodges, Charles Warlow
A case-control comparison of 114 transient global amnesia patients using both normal community-based controls and transient ischaemic attack controls showed no evidence of an increased risk of TGA associated with any of the conventional risk factors for cerebrovascular disease, but migraine was significantly more common in the TGA patients than in either control group.
Statistics in MedicineCross‐validation in survival analysis
272 Citations1993Paul E. Verweij, Hans C. van Houwelingen
A measure of the predictive value of the Cox proportional hazards model, computed from the leave-one-out regression coefficients, is constructed, which can also be used to calculate a shrinkage factor which can be applied to improve the predictions and that can be used in R2-type measures of the proportion of explained variation.
BMJEvaluation of computer based clinical decision support system and risk chart for management of hypertension in primary care: randomised controlled trial
272 Citations2000Alan Montgomery
The computer based clinical decision support system did not confer any benefit in absolute risk reduction or blood pressure control and requires further development and evaluation before use in clinical care can be recommended.
Cerebrovascular DiseasesSystematic Review of Prognostic Models in Patients with Acute Stroke
257 Citations2001Carl Counsell, Martin Dennis
None of the existing prognostic models have been sufficiently well developed and validated to be useful in either clinical practice or research, and better quality models must be produced to enable adequate case-mix correction when comparing outcome among different groups of stroke patients.
British Journal of CancerMissing covariate data within cancer prognostic studies: a review of current reporting and proposed guidelines
222 Citations2004Andrea Burton, Douglas G. Altman
Journal of Clinical EpidemiologyDeveloping a prognostic model in the presence of missing data
209 Citations2003Taane G. Clark, Douglas G. Altman
Missing values can be imputed in cases where the reason for the data being missing is known, and this will increase the power of an analysis and may produce models that are more statistically reliable and applicable within clinical practice.
BMJDo doctors accurately assess coronary risk in their patients? Preliminary results of the coronary health assessment study
182 Citations1995Steven A. Grover, Ilka Lowensteyn +4 more
This study shows that doctors in primary care accurately assess the relative risk of coronary disease in individual patients They systematically overestimate, however, the absolute risk of heart disease inindividual patients.
Journal of Clinical EpidemiologyMeasuring the accuracy of prognostic judgments in oncology
178 Citations1997William J. Mackillop, Carol F. Quirt
It was concluded that oncologists' estimates of probability of cure are surprisingly accurate, but estimates of the duration of survival of incurable patients are much less accurate, and subject to a "horizon effect" similar to that recognized in weather forecasting.
Statistics in MedicineMethodological and statistical problems in the construction of composite measurement scales: A survey of six medical and epidemiological journals
145 Citations1995Joël Coste, J. Fermanian +1 more
Recommendations are proposed for the construction and the presentation of CMS, to help authors and investigators to report and choose, respectively, measurement instruments for a complex phenomenon.
Statistics in MedicineResampling and cross-validation techniques: a tool to reduce bias caused by model building?
142 Citations1997Martin Schumacher, Norbert Holländer +1 more
This paper illustrates the phenomenon of over-optimism with respect to the predictive ability of the 'final' regression model in a simple cutpoint model and explores to what extent bias can be reduced by using cross-validation and bootstrap resampling.
Journal of Neurology Neurosurgery & PsychiatryTransient ischaemic attacks: which patients are at high (and low) risk of serious vascular events?
134 Citations1992Graeme J. Hankey, J Slattery +1 more
Journal of Medical EthicsIndividual survival time prediction using statistical models
118 Citations2005Robin Henderson, N. Keiding
It is argued that models and statistical indices can be useful at the group or population level, but that human survival is so uncertain that even the best statistical analysis cannot provide single-number predictions of real use for individual patients.
Statistics in MedicineSample size considerations for the evaluation of prognostic factors in survival analysis
112 Citations2000Claudia Schmoor, Willi Sauerbrei +1 more
An approximate formula for sample size and power is provided to detect an interaction between the interesting and a second correlated factor and is extended to the situation where a correlated factor is included in the analysis.
The American Journal of MedicineAssessment of three schemes for stratifying stroke risk in patients with nonvalvular atrial fibrillation
105 Citations2000Lesly A. Pearce, Robert G. Hart +1 more
When tested in a large cohort of patients with atrial fibrillation who were treated with aspirin, available risk-stratification schemes successfully identified patients with low rates of ischemic stroke, but less consistently identified high-risk patients.
The LancetSheffield risk and treatment table for cholesterol lowering for primary prevention of coronary heart disease
96 Citations1995Iftikhar Ul Haq, Peter Jackson +2 more
A table is presented that refines previously proposed methods of risk prediction and highlights the predominant effect of age on coronary risk; a person who is free of vascular disease and younger than 52 years is unlikely to have the specified degree of risk.
Statistics in MedicineConstruction, validation and updating of a prognostic model for kidney graft survival
74 Citations1995Hans C. van Houwelingen, J. Thorogood
The construction, validation and updating of a prognostic model for kidney graft survival is reported using data from the Eurotransplant database, and recommendations are made to place less emphasis on the selection of variables and cut-off points, and more focus on the fine-tuning of the prognostic models by means of low-dimensional parametric models in independent data sets.
Liver TransplantationSystematic review and validation of prognostic models in liver transplantation
67 Citations2005Mathew Jacob, James Lewsey +4 more
It is demonstrated that currently available prognostic models of mortality after liver transplantation can have only a limited role in clinical practice, audit, and research.
Journal of Cardiovascular PharmacologyAn Effect Model for the Assessment of Drug Benefit
59 Citations1993Jean‐Pierre Boissel, Jean‐Paul Collet +2 more
An effect model that considers both iatrogenic effects of these drugs, i.e., depression of ventricular function and arrhythmogenic effects, is developed and found that these drugs may be beneficial in high-risk patients and that the background lethal i atrogenic effect is likely to affect low- and very low- risk patients.
BMJCryptogenic Drop Attacks: An Affliction of Women
56 Citations1973David Stevens, W.B. Matthews
The sex incidence and the circumstances of the falls suggest that the cause may lie in differences between the two sexes in the mechanism of walking rather than in any central disturbance.
BMJRisk assessment in primary prevention of coronary heart disease: randomised comparison of three scoring methods
52 Citations2000Chris Isles
Statistics in MedicineSOME ISSUES IN ESTIMATING THE EFFECT OF PROGNOSTIC FACTORS FROM INCOMPLETE COVARIATE DATA
38 Citations1997Werner Vach
This paper makes use of the traditional maximum likelihood principle to analyse incomplete data in an efficient manner and considers two further issues: evaluation of differences between estimates from a complete case analysis and analyses using all subjects and assessment of the predictive value of missing values.
BMJAnalysis of predicted coronary heart disease risk in England based on Framingham study risk appraisal models published in 1991 and 2000
13 Citations2002Kiran Nanchahal, ohn R Duncan +2 more
This programme includes use of coronary risk appraisal models from the Framingham study to help identify patients eligible for drug treatment and compares the predicted risks calculated using the two models.