JMIR Publications

ABSTRACT

Background:

People with chronic pain have highlighted a need to forecast variability in their pain severity. We propose a forecasting model for both short-term variability (e.g. daily fluctuations) and longer-term variability (e.g. weekly patterns). For development of this model, clusters of weekly trajectories of pain severity are required, so that future work can predict between-cluster movement and within-cluster variability in pain severity.

Objective:

This paper aims to understand clusters of common weekly patterns as a first stage in the development of a pain-forecasting model.

Methods:

Data from a population-based mobile health (mHealth) study were used to compile weekly pain trajectories (n = 21,919) and clustered using a k-medoids algorithm. Sensitivity analyses tested the impact of assumptions related to the ordinal and longitudinal structure of the data. The characteristics of people within clusters were examined and a transition analysis was conducted to understand the movement of people between consecutive weekly clusters.

Results:

Four clusters of weekly pain severity were identified representing trajectories of no or low pain (n = 1714), mild pain (n = 8246), moderate pain (n = 8376), and severe pain (n = 3583). Sensitivity analyses indicated a four-cluster solution remained suitable under changing assumptions, and resulting clusters were similar to the main analysis, with at least 85% of trajectories belonging to the same cluster as the main analysis. Men spent longer (7.9% of weeks) in the “no or low pain” cluster than women (6.5% of weeks). Younger people (17–24 year olds) spent longer (28.3% of weeks) in the “severe pain” cluster than those aged 65–86 years (9.8% of weeks). People with fibromyalgia (31.5% of weeks) and neuropathic pain (31.1% of weeks) spent longer in the “severe pain” cluster than other conditions, and people with rheumatoid arthritis spent longer (7.8% of weeks) in the “no or low pain” cluster than other conditions. There were 12,267 pairs of consecutive weeks which contributed to the transition analysis. The empirical percentage remaining in the same cluster across consecutive weeks was 66%. When movement between clusters occurred, the highest percentage of movement was to an adjacent cluster.

Conclusions:

The clusters of pain severity identified in this study provide a parsimonious description of the weekly experiences of people with chronic pain. These clusters could be used for future study of between-cluster movement and within-cluster variability, in order to develop accurate and stakeholder-informed pain forecasting tools.