To the Editor.—
The recent article by Manning et al1 in the July 11 issue of The Journal, entitled "Changing Prescribing Practices Through Individual Continuing Education," illustrates once again the problem of testing educational interventions with seriously deficient research designs. Unfortunately, the major threats to internal and external validity in this study undermine its optimistic conclusions of improvements in prescribing patterns. The study subjects were volunteers, representing only 12% of a 1700-person reference population of internists or family practitioners. Of these 122 physicians, fully 59% dropped out of the study. Nowhere is this questionable generalizability and loss to follow-up discussed. There is no description of procedures for experimental vs control group selection, leaving open a great potential for bias.Even more serious, the final control group consisted of only nine individuals. With such a small sample size, the influence of chance alone as a possible explanation for differences between
Soumerai SB, Avorn J. Changing Prescribing Practices Through Individual Continuing Education. JAMA. 1987;257(4):487. doi:10.1001/jama.1987.03390040103023
Customize your JAMA Network experience by selecting one or more topics from the list below.