Participation in clinical studies among patients infected with HIV-1 in a single treatment centre over 12 years

To examine a complete population of clinic attenders in order to compare the demographics of patients who participated in a clinical study with those who had not. These were subdivided into trials of antivirals, trials for drugs used in opportunistic infections or symptomatic HIV and epidemiological studies. The setting was an established London teaching hospital. All patients diagnosed HIV-positive and attending between July 1983 and 1 January 1999 with one measured CD4 count and at least one follow-up visit were included.

The demographics of those participating in a clinical study were compared to those not enrolling using chi2 tests and Wilcoxon tests. Cox models were used to determine factors related to participation in clinical studies.

Data from 2703 patients representing 5342.7 person-years' follow-up were assessed. Median time of follow-up was 23.6 months. Six hundred and eighty-seven (33%) patients had ever participated in a clinical study. After adjustment for demographic factors in multivariate analysis using Cox models, homosexuals were more likely to participate compared with heterosexuals or injecting drug users (IDU) (P = 0.0035 and P = 0.0001, respectively). Women were more likely to enter a study (P = 0.02) and there was no difference between Caucasians and black Africans (P = 0.35). Between the three types of studies few differences were seen.

High rates of participation in clinical trials and epidemiological studies were seen in this cohort. In keeping with other studies, homosexual men were well represented but IDU were under-represented. However, women and black African patients showed good uptake of all clinical studies. Hence in this population there is some success in targeting representative groups to participate in clinical studies, but more effort needs to be made with IDU.