Statistical methods for the identification and use of prognostic factors

This is an expository paper which reviews the rationale for determining prognostic factors and the statistical methods for finding and allowing for such factors in the design and analysis of clinical studies. The delineation of prognostic factors in clinical studies is useful: in possibly providing insight into the mechanism of disease; in the determination of stratifications of patients for planning clinical trials; in facilitating the comparison between the outcomes of disease in different groups of patients; in assisting in the allocation of treatment to an individual patient and in permitting remedial action. A response variable is a measure of the future health or illness of the patient and its value is usually dependent on one or more prognostic variables. Statistical methodology for determining prognostic factors is reviewed for the case in which response is dichotomous, continuous or a measure of time (with the possibility of censored observations) and when the predictor variables are discrete or continuous (or a combination of both types). Methods of allowing for known prognostic variables in the analysis of a study are reviewed. These include: grouping of patients according to prognostic variables and comparing treatments separately within each group; covariance analysis and maximum likelihood. Knowledge of prognostic factors is useful in defining a control group which is to be compared with a treated group in non‐randomized studies. Three examples of studies designed to elucidate prognostic factors are described, one each in breast cancer, cancer of the prostate and Hodgkin's disease.

[1]  S. Radhakrishna,et al.  Combination of results from several 2 X 2 contingency tables , 1965 .

[2]  G. W. Snedecor STATISTICAL METHODS , 1967 .

[3]  E. Gehan,et al.  Protected environment-prophylactic antibiotic program in the chemotherapy of acute leukemia , 1971, The American journal of the medical sciences.

[4]  Alvan R. Feinstein,et al.  XVI. The process of prognostic stratification (Part 2) , 1972 .

[5]  Peter Armitage,et al.  Use of Concomitant Variables and Incomplete Survival Information in the Estimation of an Exponential Survival Parameter , 1966 .

[6]  B. Hankey,et al.  Evaluating differences in survival between two groups of patients. , 1971, Journal of chronic diseases.

[7]  J. Aitchison,et al.  Statistical Problems of Treatment Allocation , 1970 .

[8]  E. Gehan A GENERALIZED WILCOXON TEST FOR COMPARING ARBITRARILY SINGLY-CENSORED SAMPLES. , 1965, Biometrika.

[9]  E. Gehan,et al.  Combined cyclophosphamide, vincristine, and prednisone therapy of malignant lymphoma , 1971, Cancer.

[10]  M. Halperin,et al.  Estimation of the multivariate logistic risk function: a comparison of the discriminant function and maximum likelihood approaches. , 1971, Journal of chronic diseases.

[11]  E. Gehan,et al.  Comparison of survival curves, 1956 versus 1962, in children with Wilms' tumor and neuroblastoma. Report of the Subcommittee on Childhood Solid Tumors, Solid Tumor Task Force, National Cancer Institute. , 1970, Pediatrics.

[12]  M Zelen,et al.  Estimation of exponential survival probabilities with concomitant information. , 1965, Biometrics.

[13]  Matched Samples in Medical Investigations , 1964, British journal of preventive & social medicine.

[14]  M. Hills,et al.  Discrimination and Allocation with Discrete Data , 1967 .

[15]  M. Merrell,et al.  Determination of prognosis in chronic disease, illustrated by systemic lupus erythematosus. , 1955, Journal of chronic diseases.

[16]  Nathan Mantel,et al.  Problems of Convergence of Maximum Likelihood Iterative Procedures in Multiparameter Situations , 1971 .

[17]  P. Armitage,et al.  Statistical studies of prognosis in advanced breast cancer. , 1969, Journal of chronic diseases.

[18]  J. Bailar,et al.  An exponential model relating censored survival data and concomitant information for prostatic cancer patients. , 1974, Journal of the National Cancer Institute.

[19]  M. Myers,et al.  Prognostic indicators in Hodgkin's disease , 1972, Cancer.

[20]  David R. Cox The analysis of binary data , 1970 .

[21]  D. Laurence,et al.  Value of a large dose of antitoxin in clinical tetanus. , 1960, Lancet.

[22]  N. Breslow Covariance analysis of censored survival data. , 1974, Biometrics.

[23]  Alvan R. Feinstein,et al.  XIV. The purposes of prognostic stratification , 1972 .

[24]  M Zelen,et al.  The use of prognostic factors in predicting survival for breast cancer patients. , 1966, Journal of chronic diseases.

[25]  J. Cornfield,et al.  A multivariate analysis of the risk of coronary heart disease in Framingham. , 1967, Journal of chronic diseases.

[26]  E. Gehan,et al.  Non-randomized controls in cancer clinical trials. , 1974, The New England journal of medicine.

[27]  S. George,et al.  Factors influencing survival in pediatric acute leukemia. The SWCCSG experience, 1958–1970 , 1973, Cancer.

[28]  M. Zelen,et al.  Application of Exponential Models to Problems in Cancer Research , 1966 .

[29]  E. Gehan,et al.  Interrupted vs. continued maintenance therapy in childhood acute leukemia , 1975, Cancer.

[30]  J. Peto,et al.  Asymptotically Efficient Rank Invariant Test Procedures , 1972 .