Regular assessments of readers’ opinions on the quality and usefulness of articles are not difficult and should be undertaken by all bulletins. At the simplest level, information about subscribers and their opinion of the bulletin can be collected from the subscription renewal form, by analysing spontaneous mail from readers and by asking questions about why people stop subscribing. Bulletins that receive a large number of letters (several thousands per year) can retrieve much relevant information from this mail, with the quality of the letters’ content increasing with the development of the bulletin. The richer the content of the bulletin is, the more varied and interesting the exchanges with readers are.

Simple readership surveys can be carried out annually and more in-depth surveys, focus group discussions, etc. every few years. It may involve surveying everyone on the mailing list, a random sample of the mailing list or a specific professional group (e.g. primary care doctors). This type of evaluation is useful for finding out if a bulletin is read, easily understood, kept for future reference, and how readers value the information in it.

This can be done through:

• a readership survey;

• focus group discussion;

• discussions involving 'spontaneous' groups of readers (for example, groups of readers with common practices, such as refusing visits from pharmaceutical company representatives, or generic prescribers, or readers’ clubs simply meeting together to critically discuss each issue of the bulletin);

• in-depth interviews with individuals (in person or by telephone);

• 'target audience' contact. You sometimes want to study the views of non-readers in your target audience as well as the views of your loyal readers.

Sometimes two or more separate, but complementary, methods are used, such as a postal questionnaire and interviews. A telephone survey can be used to validate the postal survey because it can be done without identifying who the survey is being conducted for; this can eliminate some of the response bias which occurs when it is obvious which publication the survey is about. For bulletins that are available electronically, surveys by e-mail may be feasible.

Using one of these methods it is possible to find out what readers think of a bulletin and how it can be improved. They are relatively easy and inexpensive to conduct, especially simple surveys, and do not require an external infrastructure, such as one that collects prescription data or adverse drug reaction reports. In addition, a single assessment can cover a range of issues. For example, a Prescriber Update survey¹ included questions about how well the bulletin was read, usefulness of various types of articles, perceived influences on prescribing, readability, preferred authors, design improvements and whether electronic distribution was an acceptable alternative to paper copies. However, when designing survey questions, it is important to maintain a balance between how much is asked and the likelihood of the respondents completing the survey.