[ Neo Home | New | Jobs | Technology | Sociology & Ethics | History | Gallery | Careers | About ]

Retrolental Fibroplasia: A Modern Parable
by William A. Silverman, M.D.

Appendix D

Randomized Clinical Trials


Sackett has proposed ten methodologic standards for randomized clinical trials:

A. Criteria for inclusion in the trial

Clinical and demographic characteristics of the study population, as well as the fashion in which it was generated, must be stated clearly . . . to determine the ability to generalize the results to other groups of patients in other locations and situations . .

B. Random allocation of study patients

Allocation of study patients to alternative treatments must be random, preferably by use of random numbers tables . . .

C. Prognostic stratification

The underlying risk of developing, and dying from, the target outcomes must be identified for members of the study populations. If the subgroups in the study vary widely in the risk of target outcomes [e.g. mortality risk in newborn infants with mild to severe respiratory distress syndrome] consideration must be given to stratifying patients with respect to these risk factors before randomization, so that the comparability of the treatment and the control group is insured. In addition, in describing the results of such a trial, the effects of this prognostic stratification on the subsequent clinical outcomes should be described in detail . . .

D. Description of the therapeutic maneuver

The nature of the clinical administration and management of . . . therapy must be stated in sufficient detail to permit the reader to replicate the therapeutic program with precision . . .

E. Co-morbidity

Because the presence of a second disease may markedly affect both the ability of patients to function and their subsequent survival, the presence of relevant co-morbid conditions must be identified . . . If co-morbidity is known to affect clinical outcomes or survival . . . in a fashion which may rival the magnitude of the anticipated effectiveness of the [therapy under test], this fact should be taken into account in developing a prognostic stratification of patients before randomization.

F. Compliance

The extent to which study population members actually follow therapeutic instructions must be determined and described.

G. Co-intervention

The performance of additional screening, diagnostic, and therapeutic procedures upon the experimental group must be avoided unless these same procedures are performed with equal vigor upon members of the comparison group . . .

H. Contamination

There must be an attempt to minimize external contamination of the comparison group through the inadvertent administration of the same or related therapeutic agents as those received within the trial by the active therapy group . . .

1. Diagnostic criteria

The criteria for determining outcomes of interest, including individual causes of death, must be stated in sufficient detail to permit their application elsewhere . . .

J. Total morbidity reporting

Mortality from all causes, as well as that due to the target clinical conditions, must be reported.


Return to the Table of Contents Page
Return to the Classics Page

Created 7/31/2002 / Last modified 8/1/2002
Copyright © 1980 by Grune & Stratton, Inc.
Neonatology on the Web / webmaster@neonatology.org