gms | German Medical Science

Background: The work of both health authorities and of professional organisations comprises issuing guidance for usage of pharmaceutical substances or substance classes, following insight from pharmacovigilance. In particular, discovery of medication-related risks will lead to calls for more cautious use of the drugs at issue, including boxed warnings, if not withdrawal from the market. It is natural to ask for the impact of such guidance on changing prescribing patterns in order to assess whether regulatory and professional structures are coping efficiently with the issues of patient safety involved.

Materials and Methods: We are currently conducting an exemplary study of such effects in the field of combined hormonal contraceptives (CHC). While it is still too early to present the results of this study, which is still in progress, we describe the methodological challenges we have encountered in assessing the practical consequences of official guidance and/or warnings on CHC prescribing patterns. The time series of CHC usage data are, amongst others, analysed using joinpoint analysis, this being a well established method for detecting and describing changing patterns time.

Results: It proves to be very difficult to detect reliably the impact of guidance issued on prescribing patterns. One main reason for this is, in general, a poor signal-to-noise ratio in prescribing structures, which are influenced by a number of different factors, in addition to random fluctuations. In addition, typically guidance is not issued suddenly and surprisingly, but is the consequence of professional discussions culminating in the action of regulatory bodies. Hence, the onset of changes in prescribing behaviour tends to be gradual, which further complicates the task of identifying the impact of particular measures taken.

Conclusion: Unequivocally identifying the impact of guidance issued by regulatory or professional bodies drawing on insight from drug utilisation studies is hard to accomplish in practice. Nevertheless, the process leading to such issuance is an important contributor to change, and such change should not be expected to be directly attributable to the final step of such processes only. In addition, past experience suggests that the codification of the concerns about specific treatments is an important step in making change binding and persistent.