To the Editor: The observational study by Henderson and colleagues1 is important, as it is the first to look at the effect of advertising in clinical software on prescribing behaviour in Australia. However, the stated conclusion — that “exposure to advertisements in clinical software has little influence on the prescribing behaviour of [general practitioners]” — requires qualification, as there were potentially important confounders including the effects of other forms of pharmaceutical promotion that were not evaluated.

Evaluating the effect of a single advertising medium in isolation from concurrent activities (eg, print advertising, pharmaceutical detailing, provision of sample packs) is problematic. Pharmaceutical promotion relies on a range of activities to influence prescribing — multiple activities are synergistic, even if a single strategy shows little or no effect.2 Similarly, from a quality prescribing perspective, numerous studies show that changing prescriber behaviour is most effective when multiple strategies are used.3

There is evidence to suggest that providing information in clinical software at the point of decision making influences prescribing behaviour; for example, prompts are effective in reminding clinicians about preventive interventions such as Pap smears and immunisations.4,5

Since the study was done, another form of promotion by pharmaceutical companies has been introduced in the “advertising software” — “drug support” prompts. These prompts are linked to specific drugs and contain sponsored information from a pharmaceutical company. Users of the software may find it difficult to identify these prompts as a form of advertising, because their format and design are similar to clinical decision support prompts such as drug interaction alerts.

Clinical software has become an essential tool, with a potentially powerful influence on prescribing. If promotional messages are to be permitted in clinical software, at the very least they should be clearly labelled as such, so that the user can distinguish them from genuine decision support.

Prescribing should be based on the clinician’s sound knowledge of a drug’s indications and its relative benefits and harms, and patient treatment preferences and value for money. Advertising of medicines in any form may not be in the patient’s best interest.