High individual pain variability in people living with HIV: A graphical analysis

People living with HIV (PLWH) frequently experience pain. Following calls to analyse individual-level data in addition to group-level data in pain studies, we compared individual and group-level changes in pain prevalence, intensity and number of pain sites over 48 weeks in a large cohort of PLWH. This is the largest ever cohort study of pain in PLWH, and is the first to report pain at the level of the individual.

Participants included all participants with complete pain records from a randomized clinical trial (RCT) for the treatment of HIV (n = 787/1053). At weeks 0, 12, 24, 36 and 48 we assessed participants' pain in the last week; presence of pain, and if present, the intensity and locations of the pain. We used standard averaging methods to describe data at the group level, and unique graphical reporting methods to analyse data at the level of the individual.

Group-level data demonstrated a trend for pain prevalence to decline over time (19% week 0, 12% week 48). Worst pain intensity remained stable (median between 4/10 and 5/10), as did the number (median = 1) and common sites of pain across the 48 weeks. In contrast, individual-level data demonstrated high intra-individual variability with regards to the presence of pain, and the intensity and location of the pain.

While our group-level data were similar to previous longitudinal studies, an apparent reduction in pain over 48 weeks, the individual-level data showed large variability within individuals in that same time frame.

This graphical analysis highlights the high variability in pain (pain prevalence, intensity and body sites) across time in people living with HIV, and how presenting averaged data hides this important variability. Our data support the reporting of individual-level data in human experimental and observational studies.