Journal of International AIDS Society, 19 (Suppl 7), 100: P125
Poster presented at the HIV Drug Therapy Conference Glasgow 23rd – 26th October 2016
BACKGROUND: Given recent developments in HIV treatment it is important that, in addition to health status and symptoms, a more holistic view of the impact of HIV on quality of life is obtained. The HIV Dependent Quality of Life (HIVDQoL) questionnaire is an individualised condition-specific QoL questionnaire based on a template developed by CB for the ADDQoL[1] (Audit of Diabetes-Dependent QoL) and –DQoL measures for other conditions [e.g.2] Qualitative work to design the item content of the HIVDQoL is reported elsewhere[3] This abstract reports the psychometric evaluation of the HIVDQoL.
METHODS: The study employed a survey design with participants (N=255) recruited from the UK (N=128) and the US (N=127), via the internet, by Opinion Health. Mean age of participants was 49 years (SD=10.64), mean time since diagnosis was 15 years (SD=9.43), 203 participants were male, 49 female. Participants chose to complete and return the questionnaire individually (via post) or with a researcher (via phone). The HIVDQoL included two overview items which measure generic ‘present QoL’ and ‘HIV-specific QoL’ and 26 items that measure the impact of having HIV on specific aspects of life (e.g. family, physical appearance) and measures the importance of these aspects of life for QoL. ‘Not applicable’ options are provided for items that do not apply to everyone (e.g. work). Exploratory Factor Analysis (EFA) was used to examine scale structure and reliability using Cronbach’s alpha coefficient of internal consistency.
RESULTS: EFA was conducted in two stages. Principal components analysis, Eigenvalues >1, scree plot and parallel analysis guided the number of factors to extract, and principal axis factoring was used to determine the underlying structure. The analysis revealed a one-factor structure which included 24 of the 26 items. Two items were dropped (religious/spiritual life; having children) due to low communality and low loadings. The 24 items explained 40 percent of the variance. The factor matrix revealed the lowest loading item loaded at 0.442 and included 7 excellent items (loading >0.71), 5 very good items (>0.63), 6 good items (>0.55) and 5, fair items (>0.45). Reliability was strong: alpha=.939 for the 24 items.
CONCLUSIONS: The HIVDQoL is here shown to have sound psychometric properties including excellent reliability. It is suitable for use in clinical trials, other research and in routine clinical practice to evaluate the impact of HIV and its treatment on quality of life with a view to identifying treatments that optimise quality of life.
References
1. Bradley C, et al. The development of an individualized questionnaire measure of perceived impact of diabetes on QoL: the ADDQoL. Quality of Life Research, 1999; 8: 79-91.
2. Peach G, et al., Design of new patient-reported outcome measures to assess QoL, symptoms and treatment satisfaction in patients with abdominal aortic aneurysms. British Journal of Surgery, 2016; 8: 79-91.
3. Romaine J, et al. Design of the HIV Dependent Quality of Life (HIVDQoL) questionnaire and HIV Symptom Rating Questionnaire (HIVSRQ). Quality of Life Research 2015, 24 (suppl 1) 162: Abstract #3035.
