Intended for healthcare professionals

Research

A meta-analysis of the association between adherence to drug therapy and mortality

BMJ 2006; 333 doi: https://doi.org/10.1136/bmj.38875.675486.55 (Published 29 June 2006) Cite this as: BMJ 2006;333:15

Intergrating biomedical & Social Science research collaboratively may yield a better answer

Scot H. Simpson et al in their article under the heading 'A Meta-
analysis of the association between adherence to drug therapy and
mortality' conclude that 'Good adherence to drug therapy is associated
with positive health outcomes'(BMJ, Vol. 333, 15-18). In response to this
statement, a cynic might feel justified in stating that this is music to
pharmaceutical companies!

A major problem with this meta-analysis is that virtually all the
papers cited had as their bedrock biomedical science. To an extent, this
is understandable since to date there is little or no collaboration
between biomedicine and social science when randomised drug trials are
carried out. This type of research remains the unchallenged domain of
biomedical science. By far, the majority of authors of published papers on
compliance are social scientists - Nicky Britten, David Armstrong, Myfawny
Morgan, etc. It is noted that a large number of the references on
compliance in this article relied extensively on social or behavioral
sciences - it could not be otherwise. Among this large number of social
scientists quoted is P. Conrad. Amongst the conclusions in his study of
compliance or non-compliance, Conrad stated that "higher compliance rates
are associated with physicians giving explicit and appropriate
instructions, more clear information, and more and better feedback." (1)
Conrad's study also showed that social class or educational status were
irrelevant factors in compliance with medical regimens.

It is clear from an analysis of the designs and conclusions of papers
on compliance with a social science perspective that mortality is not
considered. Equally, the biomedical scientists fail to seriously address
patient behaviours during randomised trials. In these disparate approaches
of biomedicine and social science, it is easy for biomedical scientists to
conclude that good adherence with drug regimens is associated with
positive health outcomes. This necessarily implores the following
question: Why? In virtually all well designed studies, the populations
that are studied are usually highly motivated and so are the researchers.
Such populations consequently receive 'better' clinical care. This is
called the "Hawthorne effect"(2).

Ref. (1)Peter Conrad: The Meaning of Medications: Another look at
Compliance. (Social Science Medicine Vol.20, No.1, 1985

(2)Robert Clarke & Peter Croft: Critical Reading for the
Reflective Practitioner: Butterworth/Heinemann 1998

Competing interests:
None declared

Competing interests: No competing interests

13 July 2006
Sam W Mhlongo
Professor of Family Medicine
Medunsa 0204, South Africa
University of Limpopo, Medunsa Campus,