A large US randomized control trial of the impact on STI risk of HIV risk reduction counselling in conjunction with the HIV test (Metsch & Colfax) challenges the view that pre-test counselling constitutes an efficient use of limited resources – at least in the US setting.
Counselling has long had an important place in the struggle against HIV/AIDS. The established form of individual patient-centred counselling focussing on the discussion of the patient’s specific behaviours and the negotiation of achievable risk-reduction steps goes back to 1993. Its effectiveness was apparently confirmed by the 1998 Project RESPECT study (STIs/Rietmeijer). More recently (2008), however, a systematic review by the US Preventive Services Task Force (Lin & Bauer) could cite only one study – namely, RESPECT – as showing an effect on subsequent STI acquisition. A systematic review of 14 studies (of which 12 in the US), published 2007 in STIs, reached no very definitive conclusions (STIs/Ward & Radcliffe), though at about that time (2007) counselling was included in the UK National Institute of Clinical Excellence (NICE) guidelines (STIs/Ward). Now, in the US, the resource implications of the National Strategy goal of increasing the percentage of HIV infected who know their status from 79% to 90% by 2015 has raised the question whether the resources expended on pre-test counselling could not be better spent.
The recently reported Project AWARE study (Metsch & Colfax) is comparable in scale to Project RESPECT (5,012 participants over nine clinics). It revisits the same question of the effectiveness of pre-test counselling, though in the contemporary context of the existence of effective therapies. It also includes among its participants, men who have sex with men, an important group who were not included in Project RESPECT. Project AWARE divided participants between a “counselling” arm, who received c. 20’ of individual patient-centred counselling (as described above) along with their point-of-care HIV test, and a control arm, who were given information only. Tests for STIs, plus a computer interview, were administered at base-line, and at six months follow-up.
The results showed no significant difference in six months STI incidence between the intervention and the control arm: for the 2,039 participants in the counselling arm, there were 250 (12.3%) incident cases; for the 2,032 in the information only arm there were 226 (11.1% incident) cases. Furthermore, there was no significant difference in sexual behaviour at six months between the two groups in respect to number of sex acts, number of unprotected sex acts, and number of unprotected partners: though there was a small difference in the reported number of partners (2.7 in “counselling” as against 3 in “information only”). The authors conclude that if pre-test counselling resulted in changes in sexual behaviour, such changes were not sufficient to affect cumulative STI incidence.
An important limitation, as stated by the authors, is that the results are not “generalizable to international settings”. STI journal has featured a number of studies indicating the effectiveness of voluntary testing and counselling (VCT) in limited resource settings, whether changes of behaviour are the result of the counselling itself, or a change of attitude prior to the decision to attend VCT: see (Kenya) STIs/Arthur & Gilks; (South Africa) STIs/Nshiniyimana & Bruyn; (Zambia) STIs/Sikasote & Murray. Another clearly stated limitation is that the study applies only to the established type of 20’ individual patient-centred intervention they describe – not to other types, such as, for example, the couple-based intervention very positively evaluated by LaCroix & Johnson (STIs/LaCroix & Johnson).