If your research involves hospital in-patients the first step is to decide which wards are most likely to be caring for the patients you are interested in. This could be obvious: if you are interested in an obstetric issue you will focus on these wards; similarly critical care patients will be on intensive care. However, recruitment could be more general, so for example, if you are examining referrals to rehabilitation wards, patients may come from any number of wards. Similarly, patients throughout the hospital could have the problem you are interested in, such as poor mobility, cognitive impairment, pain, malnutrition and so on. If your patient group is diverse and could come from many wards try to focus on a limited number, if you try to cover all wards it will be time consuming and unproductive. Ask the ward staff if they think their ward will provide suitable patients; they will know best what type of patients they care for. Then it is a matter of going round the wards and locating possible patients and approaching them.

Before you start, don’t forget to inform the consultants and their teams that you will be recruiting patients under their care. Explain the study briefly, what is involved for their patients, and state that you have all the necessary approvals. They are responsible for each patient’s care while they are in hospital so it is essential that they know what you are doing. Occasionally, some consultants will object to the inclusion of their patients and you will have to respect this; however, in my experience this is rare.

The study I ran looking at patients who were taking an antibiotic is a good example of how a recruiting strategy can develop. Clearly, patients taking antibiotics were spread throughout the hospital and in theory I could recruit from any ward. I started by thinking through the exclusion criteria and working out if certain wards would not be fruitful. Because immune compromised patients were excluded I avoided the intensive care wards. I also had to recruit patients within two days of starting an antibiotic so long stay wards were also of little use. I needed to exclude patients who had recent previous antibiotic treatment and after a short time I found vascular surgery patients often had a history of frequent antibiotic use so I avoided this ward too. After several months I had refined my recruiting strategy to a limited number of wards, including admission wards, care of the elderly and respiratory wards. I could then really focus on building relationships with the staff on these wards to make recruitment as efficient and rapid as possible. The key was to check all new admissions as soon as possible to find out if they were going on antibiotics. I tried numerous approaches including obtaining the previous day’s admissions lists, liaising with emergency department doctors, asking medical teams on each ward to contact me if they prescribed antibiotics, and working with the ward pharmacists to highlight new patients. Many of these strategies failed, principally because people who work in hospitals are very busy and although they might want to help with research, it usually comes a long way down their priority list. The best way I found was to go to the wards and check personally with the staff there, and in the end this was the most time efficient and effective strategy.

In another study I worked on I was interested in recruiting patients just before they were discharged from hospital. Anyone who has worked in an acute hospital environment will know that predicting discharge dates is extremely difficult. Again the best way was to focus on particular wards, build up relationships with the ward staff and visit as often as possible. Finding time for a friendly chat or to help someone out paid huge dividends when it came to getting timely notification of someone going home.

Much health care research will involve people using the health care system, and hence volunteers can be recruited using the methods described in the previous section. Nevertheless, some research requires healthy volunteers or volunteers who are at risk of a particular condition but are currently not accessing the health care system. For this type of volunteer you will have to go out and recruit more widely from the community. There are several possible approaches that can be used individually or in combination, including advertising, writing letters, or publicising the research you are doing.

Advertising in newspapers or with posters sited in work or public spaces can be an effective means of recruiting volunteers. You will need to investigate what is the best way locally and this may be a process of trial and error. If you know other researchers in the area, ask for advice.

Local papers may carry adverts; these will usually be at a cost, although some papers will provide free space for projects seeking volunteers (for example, the London Evening Standard). Ring your local papers and ask what they can offer you.

If you are using posters think about placing them in large offices, GP surgeries, sports centres, colleges, universities, or libraries. Before placing any adverts make sure you get the correct permissions to advertise and remember your advert or poster must have been approved by the ethics committee.

The group you are targeting will dictate the best places to advertise. For example, in the past I have looked for people who wanted to lose weight and a newspaper advert produced a large response. However, other studies requiring men with high cholesterol were harder to recruit for and we had to focus adverts on organisations with a high proportion of male employees, such as the police or fire service.

The following questions are worth asking when recruiting with adverts: