The aim of the project was to develop a toolkit for professionally led library services to assess the impact of the various services provided, including provision of electronic information, on outcomes that related to patient care. The objectives were to: • pilot and validate survey tools in a variety of services (including clinical librarian services as well as more traditional health library settings) • evaluate the research processes, to provide sound advice in the toolkit. KEY MESSAGES Planning the project Impact studies may usually be conducted as audit projects, and this avoids the need to obtain research ethics approval. Advice from research governance staff is recommended, and research governance practice varies locally. Human resources will do the random sampling of acute staff for you – but primary care staff need to be targeted in other ways. Research design Keep questionnaires short, check that the terms on any validated questionnaires will be interpreted as you expect. Use a mix of questionnaire and interview. Calculate the sample size required for your population and the confidence interval you wish for the questionnaire survey. If you do not have time to do a large survey, focus your impact survey on one or two staff groups – better to get a good response from one staff group than a poor response from all staff groups. Use interviews to collect the details, and the stories, about the way in which information obtained via library service contributed to policy, guidelines, patient care, governance. Target teams, units, and practice development (in the widest sense) to make best use of interview time. Conducting the research Ensure that the accompanying documentation is clear, and indicates any benefits to the potential respondent of participation. Simple descriptive statistics are adequate for assessing the impact, but when planning enhancement of services some synthesis and interpolation of the findings may be necessary. Leave space for some open-ended comments on the questionnaire. Respondents find estimating the time spent on a search difficult, but such estimates help – together with other evidence – to justify the cost-effectiveness of a service. Impact does not just affect the individual searcher as ‘advice to colleagues’, impacts on service delivery, and advice to patients indicate that information obtained may have long-term effects on practice and patient care.