Concordance. A 10-item scale adapted from Elwyn et al. [11] and based on the

concordance model was developed to capture the overall shared decision-making process around treatment change in an HIV clinical situation. Respondents were asked to indicate the extent to which the doctor carried out the following: (a) described issues behind the need to change treatment; (b) clarified s/he had a balanced view on their options; (c) outlined options available; (d) provided information in their preferred format; (e) checked their understanding of issues and their preferred role in the decision-making; (f) explored their concerns, expectations check details and options; (g) gave them time to talk to others before reaching a decision; (h) made and reviewed a final decision. selleck inhibitor Each item was coded as: 1 (did not happen), 2 (not very good), 3 (adequate) and 4 (very good). A concordance score was then generated by summing the 10 item

scores. It ranged from 10 (low) to 40 (high). Sexual behaviour. Information on partnership and sexual risk behaviour in the preceding 3 months was recorded. HIV sexual risk behaviour was defined as unprotected sexual intercourse with someone of unknown or discordant HIV status during the previous 3 months. Treatment switching. The use of HAART and whether such treatment had been switched once, twice or more, or stopped, were recorded. Symptom and pain levels. The Memorial Symptom Assessment Short Form (MSAS) Florfenicol inventory, a multiple symptom inventory measuring the 7-day prevalence of physical and psychological symptoms, and their associated burden, was used [23]. Three subscales (physical, psychological and global distress

indices) were calculated. Two additional items (feeling optimistic and suicidal thoughts) were also included and independently analysed. Quality of life. EuroQol 5D, which includes five dimensions of quality of life (mobility, self-care, usual activities, pain/discomfort and anxiety/depression) and a quality of life visual analogue scale (VAS), was used [24]. Satisfaction. Five-point Likert type rating scales were employed to assess satisfaction in relation to medical treatment and care. Perceived involvement in decision-making and doctor–patient agreement. Five-point Likert type rating scales were used. Adherence. Patient self-report recall over the preceding 7 days was used to assess antiretroviral adherence. Full adherence was coded as no missed doses and all taken within 1 h of the correct time and in accordance with any dietary requirements. Partial adherence was coded as those who had taken all doses, but had not been fully adherent to dose timing and/or requirements [25]. Nonadherence included all other responses – where doses had been missed and timing/circumstance had been inconsistent. For a subset of patients who provided consent, questionnaire data were linked to clinical information which provided the VL and CD4 cell count at the time of the questionnaire and 6–12 months afterwards.