BACKGROUND

A strong body of evidence supports the value of peer mentoring and peer-led education in increasing self-efficacy among individuals with health conditions, leading to improved self-management skills and health outcomes. However, this work has not been translated into interventions for patients in the early stages of recovery, who may lack insight about the importance of self-management. Our research attempted to address this important evidence gap.

OBJECTIVES

The project sought to identify effective and sustainable approaches to support the transition from hospital to home for patients completing inpatient rehabilitation after spinal cord injury (SCI). Goals were to (1) design, build, and test refinements to transition supports; (2) implement system changes that appear to be effective; and (3) evaluate effectiveness of the system changes in reducing hospital readmissions and improving patient-reported outcomes.

METHODS

The study population was SCI patients admitted for inpatient rehabilitation (n = 1147). We evaluated the impact of 3 interventions: (1) peer-led education in self-management, (2) one-to-one peer mentoring, and (3) an internet patient portal to provide ready access to information and resources to support self-management. The primary outcome for evaluation of the effects of peer-led education was patients' positive engagement in classroom activities (eg, asking questions, participating in discussion). The primary outcome of interest for the other 2 aims was unplanned hospital readmissions 30 days postdischarge from inpatient rehabilitation. Secondary outcomes included a reduction in unplanned hospital readmissions 90 and 180 days postdischarge as well as improved patient-reported self-efficacy, satisfaction with care, depression, injury intrusiveness, and overall life satisfaction. We conducted the research over a 4-year period in a large, private, nonprofit hospital specializing in medical rehabilitation of patients with SCI, acquired brain injury, and other serious neurological conditions. We conducted 3 studies. In study 1, we replaced conventional nurse-led self-management classes with peer-led classes incorporating approaches to promote transformative learning. Using a multiple-baseline design, we implemented the revised classes across 3 subject areas (skin care, bladder management, special concerns and documented the effects on positive engagement responses of patients attending the classes. In study 2, a randomized trial compared the effectiveness of one-to-one peer mentoring with general peer support. In study 3, we used interrupted time-series analysis (ITSA) to examine the combined and individual effects of the 3 interventions on unplanned readmissions and patient-reported outcomes once implemented into standard practice.

RESULTS

In study 1, the average (per patient per class) engagement responses were higher with the peer-led classes across each of 3 subject areas ( = 0.008; mean difference [Dif] = 15.5; CI, 10.3-20.7). In study 2, no significant differences occurred between peer-led mentoring and general peer support in the percentage of patients with unplanned hospital readmissions 30, 90, and 180 days postdischarge. However, experimental participants receiving peer mentoring had significantly fewer unplanned hospital days than control-group participants 30 days ( = 0.018, Dif = 26, CI = 16-36), 90 days ( < 0.001, Dif = 95, CI = 76-114), and 180 days ( < 0.001, Dif = 112, CI = 91-133) postdischarge. We also found the growth rate for self-efficacy in the 6 months postdischarge to be significantly higher for peer-led mentored participants ( = 0.015). Quality-of-life outcomes (depressive symptoms, community participation, injury intrusiveness, and overall satisfaction with life quality) were the same in the peer-led mentoring and control groups at any follow-up interval. Following implementation of the peer-led education and peer mentoring interventions as standard practice (study 3), our ITSA showed a significant decrease in the level ( < 0.001) but not slope ( = 0.095) of the percentage of patients rehospitalized and both the level ( < 0.001) and slope ( = 0.004) of unplanned hospital days 30 days postdischarge. Reduced hospital days were associated with the amount of peer mentoring ( = 0.007) received but not with the number of peer-led education classes attended ( = 0.763). We also observed significant improvements in self-efficacy ( < 0.001) 180 days postdischarge associated with exposure to peer mentoring and a significant relationship between improved self-efficacy and reduced hospital readmissions ( < 0.001). Peer mentored and control groups had the same satisfaction with care, depression, injury intrusiveness, and overall life quality. In evaluating the impact of the patient portal on unplanned hospital readmissions, we observed a decrease (from 7% to 3%) in the percentage of patients readmitted and days rehospitalized (ratio change from 0.012 to 0.0083) 30 days postdischarge. However, the study was underpowered (because of too few observation intervals) to draw valid inferences concerning the observed change. Portal adoption was low—19% of eligible patients—and insufficient data were available to examine the relationship between portal use and hospital readmissions.

CONCLUSIONS

Our findings suggest a relationship between exposure to one-to-one peer mentoring to improve SCI patient self-efficacy and fewer days of unplanned hospital readmissions postdischarge. Peer-led education is associated with improved patient engagement in class activities, but its relationship to patient-reported outcomes is equivocal.

STUDY LIMITATIONS

These include (1) no verification of external validity of findings; (2) no controlled evaluation of the effects of peer-led education on patient-reported outcomes; and (3) limited power to evaluate the impact of the patient portal on study outcomes.