Advocates for community participation and quality of life may benefit from a better understanding of how pain intensity and environmental barriers influence participation outcomes. Unfortunately, little evidence exists on how the interaction between personal factors (e.g., pain) and environmental factors (e.g., physical accessibility) influence participation. To address this gap, we studied Pain Interference Patterns (PIP) by collecting both longitudinal and ecological momentary assessment (EMA or real-time) data to explore these factors and outcomes. Through better understanding of these interactions, we hope to inform interventions, policy, and services that can promote full participation in community life.
Domains of Participation
For this project looked at participation in three categories.
- Visits to Doctors or Health Care Providers
- Grocery Stores
- Shopping Malls
- Large Box Stores
- Public Parks
- Exercise Facilities
- Active Recreation
- Socializing Outside the Home
- Religious Activities
- Community Activities
The regression models for medical and discretionary participation were significant, while shopping participation was not. It appears that shopping (including grocery shopping, malls, and big box stores) is not sensitive to variation in personal and environmental factors. This is somewhat expected in that household shopping and in particular grocery shopping constitutes a basic need that cannot be ignored.
Higher sum of secondary conditions scores were associated with increased medical participation, while higher reported pain levels were associated with reduced medical participation. This finding may reflect the debilitating nature of pain such that pain at higher levels decreases the likelihood of leaving home to access or use medical services. Interestingly, higher rates of participation barriers were associated with increased medical care use. It is possible that individuals who encounter more barriers have more limiting health conditions and the participation barriers scale is acting as a proxy for disability severity. Respondents who were employed PT or FT engaged in significantly less medical participation. Employment may be capturing variance of a factor not specified in the model, like complexity of medical needs, which affects both employment and medical service utilization. Alternatively, it could simply reflect that employed people have less time for pursuing medical appointments.
Interestingly, respondents who were married or had a partner participated in almost four fewer discretionary outings per week. This may indicate that single people go into the community for socialization, while couples have this need met at home. Pleasure was associated with more discretionary participation, where respondents who scored higher on the pleasure domain of the orientation to happiness scale participated in more activities. This finding seems to reflect that people who participate more experience more pleasure. However, the opposite could also be true; individuals who experience more pleasure tend to participate more. Understanding the directionality of this result could be important for developing interventions to increase participation.
Resting for pain management was associated with reduced discretionary participation. The literature indicates that those who use resting as a coping strategy for pain management are more likely to have worse physical functioning (Jensen, Turner, Romano, & Strom, 1995). Rest may serve as a proxy variable for disability severity, which in turn acts as a mediating variable between pain management and discretionary participation.
Overall, the models highlight the need to explore specific domains of participation, as facilitators in one domain may actually serve as barriers in another. Additional participation domains likely deserve attention as well. For instance, we used employment as an independent variable for the medial, shopping and discretionary domains, but taking part in the workforce is also a participation outcome. We did not explore employment because it is typically analyzed using logistic regression and was not directly comparable to the other participation domains.
The Next Steps
By looking at change scores over time with longitudinal data, we can begin to understand causal factors related to participation. These hypothesized relationships can be further analyzed among participants who also completed EMA real-time data (n = 100). EMA respondents answered six mini-surveys per day for 14 consecutive days about what they were doing, how they were feeling, and what barriers they encountered since the last data prompt (i.e., approximately over the past two hours). Recording events as they occur reduces recall bias and allows us to explore causality and patterns. For instance, high pain episodes may follow high rates of barriers encountered in the community. Likewise, participation may have lagged patterns related to changes in pain levels. How variables interact provides evidence for addressing pain and participation outcomes and can be used to inform behavioral interventions.