Abstract
Objective:
Behavioral cancer pain interventions are efficacious for improving important pain outcomes; yet, traditional in-person delivery limits patient access. This study compared videoconference-delivered Pain Coping Skills Training (mPCST) to in-person Pain Coping Skills Training (PCST-traditional).
Methods:
This study was a randomized, noninferiority trial with cancer patients. Participants (N=178) were randomly assigned to four, 45-minute sessions of mPCST or PCST-traditional. Session content focused on evidence-based cognitive and behavioral pain management skills. Assessments were completed at baseline, post-treatment, and 3-months post-treatment, and included measures of primary intervention outcomes (i.e., pain severity and pain interference) and secondary intervention outcomes (i.e., physical symptoms, psychological distress, physical well-being, and self-efficacy). The main study aim tested whether mPCST was more accessible (defined as feasibility, acceptability, patient burden, and engagement) than PCST-traditional. The second aim tested whether mPCST was noninferior to PCST-traditional.
Results:
mPCST demonstrated significantly greater feasibility (i.e., attrition, adherence, time to completion) than PCST-traditional. Both groups reported similar patient burden and engagement as well as a high degree of acceptability. All intervention outcomes demonstrated noninferiority at post-treatment and, with the exception of physical symptoms, 3 months post-treatment. Concerning the primary intervention outcomes, 95% CIs for the mean differences (d) were below the noninferiority margin of 1 for pain severity (post-treatment d=0.09, 95% CI −0.63 to 0.81; 3 month d=−0.43 95% CI −1.22 to 0.36) and pain interference (post-treatment d=−0.11, 95% CI −0.99 to 0.76; 3 month d=−0.26 95% CI −1.14 to 0.62).
Conclusion:
mPCST is highly accessible and noninferior to PCST-traditional.
Keywords: cancer, pain, behavioral pain intervention, mobile health, pain coping skills training
Background
Pain is one of the most common and distressing symptoms for cancer patients.1 Behavioral pain interventions can decrease pain, pain-related symptoms, and distress in cancer patients and serve as an efficacious adjuvant therapy to analgesic regimens.2 Behavioral interventions teach patients behavioral and cognitive strategies for managing factors that contribute to pain.3 Unfortunately, most cancer patients do not have access to behavioral cancer pain interventions due to persistent barriers.
Traditionally, behavioral cancer pain interventions have been delivered in-person at major medical centers. This delivery modality presents challenges for cancer patients. Cancer patients often receive care many miles from their home and additional in-person appointments are difficult due to illness, cost, time constraints, transportation, and travel distance.4,5 Others receive care in community clinics where therapists trained in behavioral interventions are unavailable.
Mobile health technologies offer opportunities to expand the reach of efficacious behavioral pain interventions. Pain Coping Skills Training (PCST) is a psychosocial and behavioral intervention that includes efficacious cognitive-behavioral pain management strategies. It was developed and has been extensively tested by our group, including pilot work supporting this trial,3,6,7 and provided a good model of intervention to translate to videoconferencing. We developed a Mobile Health Pain Coping Skills Training (mPCST) protocol that is delivered through videoconferencing to cancer patients in their own environment (e.g., home, work) by a therapist at the medical center. mPCST includes four 45-minute videoconferencing sessions teaching cognitive-behavioral pain management strategies. mPCST also includes a website that provides patients with mPCST materials and information, social networking, and daily assessments used to personalize sessions. In two small pilot trials (i.e., single arm, randomized controlled trial [RCT]), mPCST was found to be feasible, acceptable, engaging, low burden, and demonstrated pre- to post-treatment improvements in pain and other pain-related outcomes.6,7 The RCT (N=30) compared mPCST to a traditional in-person intervention and found that participants in both groups reported improvements in pain, physical symptoms, and self-efficacy for pain management suggesting mPCST, a potentially highly accessible intervention, may be at least as efficacious as in-person.
This study used a large RCT to compare mPCST to a traditional, in-person behavioral cancer pain intervention (i.e., Pain Coping Skills Training Traditional [PCST-traditional]). Aim 1 tested whether mPCST was more accessible than PCST-traditional. We hypothesized mPCST would improve intervention access, assessed by feasibility (i.e., attrition, adherence, study completion time), patient burden (i.e., physical, emotional, financial), engagement (i.e., skills practice, understanding), and acceptability. Aim 2 tested whether mPCST was noninferior to PCST-traditional when examining primary (i.e., pain severity, pain interference) and secondary (i.e., physical well-being, psychological distress, self-efficacy for pain management) intervention outcomes. We hypothesized mPCST would be noninferior to PCST-traditional.
Methods
Study Design
Participants (N=178) were randomized to mPCST or PCST-traditional. Participants completed baseline, post-treatment, and 3-month follow-up assessments.
Participants
Participants were diagnosed with breast, lung, prostate, or colorectal cancer within the last two years. Eligibility criteria included: 1) ≥18 years old; 2) life expectancy of ≥6 months; and 3) two clinical pain ratings of ≥3 (0–10 scale, current pain) gathered as part of clinical oncology visits at least 3 weeks apart but no more than 12 months apart, with one of the ratings on the day of recruitment. Exclusion criteria included: 1) cognitive impairment or severe psychiatric condition based on chart review; 2) brain metastases; or 3) receiving pain coping skills training ≤6 months.
Procedure
Procedures complied with ethical guidelines and received Duke University institutional review board approval (Pro00054792). Recruitment took place 2014–2017 at the Duke Cancer Center. Participants were identified using electronic medical records (EMR). A letter signed by the oncologist and principal investigator (PI) was sent to the patient. Patients were contacted and met with study staff for consent.
After completing the baseline assessment, participants were randomized with a 1:1 allocation ratio to receive mPCST or PCST-traditional. A random number assignment procedure was conducted by a study member who did not interact with participants. Participants were blinded to study hypotheses. Assessments were completed online to reduce demand characteristics and eliminate assessor bias. Study statisticians were not involved in data collection. Analyses were conducted after completion of data collection.
Intervention Conditions
mHealth Pain Coping Skills Training (mPCST).
The mPCST protocol was four 45-minute videoconferencing sessions involving: 1) efficacious theory-based skills and home practice assignments; 2) real-time pre-session assessment of symptoms, coping skills use, and preferences; 3) personalized session content based on the pre-session assessment; and 4) a study website. mPCST participants were given a tablet computer (iPad) with data plans for Internet access to engage in study procedures including videoconferencing sessions.
Session content was based on Pain Coping Skills Training (PCST). PCST enhances participants’ abilities to cope with pain by teaching skills to change their behaviors, thoughts, and feelings about pain. mPCST skills (e.g., progressive muscle relaxation [PMR], activity pacing/planning, cognitive restructuring, imagery) were designed to enhance participants’ confidence in their abilities to: 1) distract from pain and distress; 2) manage pain to increase activities; and 3) use cognitive restructuring to decrease negative pain related thoughts. Homework was given to incorporate skills into daily life.
Before each session, participants used the study website to complete self-report assessments (i.e., pain, pain self-efficacy, physical disability, distress, coping skills use) and session preferences (i.e., skills review, PMR practice) that were provided to the therapist in real-time. Based on these data, therapists tailored sessions to participants’ needs. Decision algorithms were used to guide therapists in personalizing content. For example, graphs of the trajectory of assessment scores for pain and self-efficacy for pain management coupled with participants’ reported use of pain coping skills allowed therapists to problem solve and make a plan for skills use in the case of increased pain (30%) or decreased self-efficacy (30%). Skills review and use of PMR in session was also tailored based on patient preferences.
The study website provided a centralized place that included intervention information, skills practice tracking, videos modeling skills, patient skills use stories, social networking, and study assessments. The social networking component prompted participants to write about their skills use experiences; study staff moderated networking.
Pain Coping Skills Training Traditional (PCST-traditional).
PCST-traditional participants received the same sessions and coping skill content as mPCST participants; however, sessions were conducted in-person at the medical center. Participants had access to the study website and intervention sessions were also personalized using pre-session assessments. PCST-traditional participants were provided with an iPad, as needed, to access the website.
Study Therapists, Training, and Treatment Fidelity
Therapists were PhD-level clinical psychologists or advanced clinical psychology students. Therapists were trained in the manualized protocol by the PI (Somers, PhD). Therapists role played each session with an actor and sessions were rated for protocol adherence and delivery competence by the PI using a standard 11-item form. Eighty percent adherence and competence for each session was required to certify therapists for intervention delivery. Sessions were recorded and a subset reviewed with the therapist by the PI or a senior clinician.
Measures
Access Variables
Feasibility.
Feasibility was assessed by study attrition (i.e., withdrawing before completing), adherence (i.e., calculated by total number of completed sessions [4] and assessments [3] with a possible range of 0–7), and completion time (i.e., days from session 1 to 4). No time limit was placed on the participants’ intervention completion as long as they expressed willingness to remain enrolled, as this was a primary outcome question.
Acceptability.
Acceptability was assessed post-treatment with the Client Satisfaction Questionnaire, 10-item version.8 Example items include, “How would you rate the quality of the program?” and “To what extent did this program meet your needs?” Items were rated from 1=low acceptability to 4=high acceptability and were averaged (α=0.92).
Patient burden.
Patient burden was assessed post-treatment.7 Participants rated how difficult it was to complete sessions from 1=not at all to 4=very much. Example items included feeling fatigued, time and cost, transportation, and distance from the medical center. Items were summed (α=0.72).
Engagement.
At post-treatment, participants were asked how many days in the past 7 they had used each coping skill. They were also asked to rate the degree to which the study information was presented in an understandable way (0=none to 7=fully). Each of these items were included separately in analyses.
Primary and Secondary Intervention Outcomes
Intervention outcomes were assessed using validated measures of pain and pain-related outcomes at baseline, post-treatment, and 3-month follow-up. See Supplemental Information for intervention outcome measure details. Primary intervention outcomes included pain severity and pain interference referencing the last 7 days using the Brief Pain Inventory (BPI).9 Secondary intervention outcomes included physical symptoms, physical well-being, and psychological distress assessed with the Patient Care Monitor (PCM), v2.0,10,11 and self-efficacy for pain management assessed with the Chronic Pain Self-Efficacy Scale.12
Patient Characteristic and Medical Variables
Participants’ medical variables (i.e., cancer type, stage, diagnosis date) were collected through EMR. Other participant characteristics (e.g., marital status, education level) were collected through self-report at baseline.
Statistical Analyses
Aim 1 tested whether mPCST was more accessible than PCST-traditional. The hypothesis was tested by comparing between-group differences in access variables, including feasibility (i.e., attrition, adherence, completion), intervention acceptability, patient burden, and patient engagement (i.e., coping skills practice, information understanding). Between-group differences in access were assessed using independent samples t-test, chi-square analysis, and Mann-Whitney U test.
Aim 2 tested whether mPCST was noninferior to PCST-traditional. We used a linear mixed modeling approach13 that included fixed effects for group, time, and a group-by-time interaction term. Models controlled for baseline values of the outcome and parameters were estimated using restricted maximum likelihood. A two-sided 95% confidence interval (CI) approach was used.14 Noninferiority was demonstrated when the 95% CI for the mean difference did not cross the margin of noninferiority. Noninferiority margins for the primary intervention outcomes of pain severity and interference were selected based on literature that suggests a 1-point change on an 11-point scale is clinically meaningful.15 Consistent with prior noninferiority trials,16 we conservatively set the noninferiority margins for the secondary intervention outcomes as the smallest change a participant could endorse. Primary noninferiority analyses were conducted on the per-protocol sample, as using the full intent-to-treat sample can increase the risk of incorrectly claiming noninferiority.17 A secondary sensitivity analysis was then conducted using the intent-to-treat sample.
Sample Size
The sample size was determined based on the primary study aim. An a priori power analysis with α=0.05 (one-tailed) suggested a sample size of 156 would provide >80% power to detect an effect size of d=0.40 (i.e., a medium-sized effect) for between-group differences in all access variables.
Results
Study Flow and Baseline Characteristics
Of 707 patients approached, 529 were ineligible (n=216) or declined participation (n=313). Overall, 36% of eligible patients consented (N=178; Figure 1). Table 1 presents participant demographics and medical variables. At baseline, pain severity on the BPI was rated as mild (0–3) by 50%, moderate by 40% (4–6), and severe (7–10) by 10% (PCST-traditional mean=3.98, range=0 to 10; mPCST mean=3.86, range=0 to 9.25).
Figure 1.
Study flow chart
Table 1.
Baseline characteristics
| Characteristic | Overall N=178 |
mPCST n=89 |
PCST-traditional n=89 |
|---|---|---|---|
| Female | 128 (71.91) | 60 (67.42) | 68 (76.40) |
| Age | 56.47 (10.53) | 56.08 (10.92) | 56.85 (10.18) |
| Race | |||
| Caucasian | 122 (68.54) | 61 (68.54) | 61 (68.54) |
| African American | 49 (27.53) | 23 (25.84) | 26 (29.21) |
| Other | 7 (3.93) | 5 (5.62) | 2 (2.25) |
| Marital Status | |||
| Married | 106 (59.55) | 49 (55.06) | 57 (64.04) |
| Single | 37 (20.79) | 20 (22.47) | 17 (19.10) |
| Divorced or separated | 29 (16.29) | 17 (19.10) | 12 (13.48) |
| Widowed | 6 (3.37) | 3 (3.37) | 3 (3.37) |
| Education | |||
| Less than high school | 14 (7.87) | 7 (7.87) | 7 (7.87) |
| High school diploma | 20 (11.24) | 7 (7.87) | 13 (14.61) |
| Some college | 54 (30.34) | 27 (30.34) | 27 (30.34) |
| Bachelor’s degree | 38 (21.35) | 22 (24.72) | 16 (17.98) |
| Graduate degree | 45 (25.28) | 24 (26.97) | 21 (23.60) |
| Missing | 7 (3.93) | 2 (2.25) | 5 (5.62) |
| Cancer Type | |||
| Breast | 86 (48.31) | 46 (51.69) | 40 (44.94) |
| Lung | 42 (23.60) | 20 (22.47) | 22 (24.72) |
| Colorectal | 41 (23.03) | 18 (20.22) | 23 (25.84) |
| Prostate | 9 (5.06) | 5 (5.62) | 4 (4.49) |
| Cancer Stage at Recruitment | |||
| I | 25 (14.04) | 13 (14.61) | 12 (13.48) |
| II | 54 (30.34) | 27 (30.34) | 27 (30.34) |
| III | 39 (21.91) | 23 (25.84) | 16 (17.98) |
| IV | 60 (33.71) | 26 (29.21) | 34 (38.20) |
| Months since initial diagnosis | 11.81 (23.87) | 14.05 (32.51) | 9.48 (7.86) |
| Months since current diagnosis | 8.95 (7.05) | 8.88 (6.53) | 9.02 (7.58) |
| Days of pain medication usage | 4.41 (2.87) | 4.51 (2.90) | 4.31 (2.85) |
Abbreviations: mPCST, Pain Coping Skills Training delivered using iPad+Skype; PCST-traditional, Pain Coping Skills Training delivered in person. Data presented are number (percent) or mean (standard deviation). Days (0–7) of pain medication usage in past week was self-reported. There were no significant between-group differences in any variables (ps>0.05).
Aim 1: Accessibility
mPCST demonstrated greater feasibility than PCST-traditional (Table 2), specifically: 1) study attrition was lower in the mPCST group than the PCST-traditional group (16% vs. 30% attrition, X2=5.36, p=0.02); 2) adherence to the intervention protocol was better in the mPCST group (difference=0.72, p=0.03); and 3) session completion time was lower for mPCST participants (difference=−38.41 days, p<0.001). Overall patient burden was similar in both groups. Important group differences emerged when examining individual burden items. PCST-traditional participants reported more burden due to the distance lived from the medical center (Mann-Whitney U=1925.00, p=0.007) and difficulty getting to sessions (difference=−0.36, p=0.004); mPCST participants reported more difficulty remembering to attend sessions (difference=0.23, p=0.04). mPCST participants reported greater usage of brief relaxation (difference=1.91, p<0.001). Other skills also showed trends toward greater use in the mPCST group, though were not statistically significant. Both groups reported high intervention acceptability.
Table 2.
Patient access and mean skills practice
| Group | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| mPCST (n=89) | PCST-traditional (n=89) | ||||||||||
| Mean | SD | n | Mean | SD | n | Mean difference | 95% CI | p | |||
| Feasibility and Acceptability | |||||||||||
| Adherence | 6.02 | 2.09 | 89 | 5.30 | 2.31 | 89 | 0.72 (0.33) | 0.07 to 1.37 | 0.03* | ||
| Completion days | 27.11 | 10.05 | 74 | 65.52 | 42.77 | 58 | −38.41 (5.74) | −49.88 to −26.94 | <0.00001* | ||
| Patient burden | 17.56 | 3.56 | 71 | 17.85 | 5.39 | 62 | −0.29 (0.80) | −1.89 to 1.30 | 0.72 | ||
| Patient acceptability | 3.55 | 0.42 | 73 | 3.52 | 0.50 | 62 | 0.02 (0.08) | −0.13 to 0.18 | 0.77 | ||
| Engagement | |||||||||||
| PMR with audio | 1.97 | 2.30 | 72 | 1.90 | 2.25 | 61 | 0.07 (0.40) | −0.71 to 0.85 | 0.86 | ||
| PMR no audio | 3.74 | 2.44 | 72 | 3.08 | 2.51 | 61 | 0.65 (0.43) | −0.20 to 1.51 | 0.13 | ||
| Activity-rest cycles | 3.67 | 2.61 | 72 | 3.42 | 2.45 | 59 | 0.24 (0.45) | −0.64 to 1.12 | 0.59 | ||
| Pleasant activities | 4.36 | 2.04 | 72 | 3.98 | 2.13 | 60 | 0.38 (0.36) | −0.34 to 1.10 | 0.30 | ||
| Positive thinking | 4.92 | 2.12 | 72 | 4.33 | 2.38 | 60 | 0.58 (0.39) | −0.19 to 1.36 | 0.14 | ||
| Guided imagery | 4.03 | 2.39 | 72 | 3.60 | 2.57 | 60 | 0.43 (0.43) | −0.43 to 1.28 | 0.32 | ||
| Brief relaxation practice | 3.81 | 2.84 | 72 | 1.90 | 2.34 | 60 | 1.91 (0.45) | 1.01 to 2.80 | 0.00004* | ||
| Information understanding | 6.74 | 0.95 | 72 | 6.44 | 1.35 | 61 | 0.29 (0.21) | −0.11 to 0.70 | 0.16 | ||
Abbreviations and measures: mPCST, Mobile Pain Coping Skills Training; PCST-traditional, Pain Coping Skills Training delivered in person; Mean difference represents mPCST minus PCST-traditional (standard error).
p<0.05 for independent samples t-test.
Aim 2: Noninferiority
Consistent with our hypotheses, mPCST was noninferior to PCST-traditional on all outcomes at both follow-up timepoints, with one exception (Table 3). Specifically, physical symptoms were only noninferior immediately post-treatment. All outcomes improved over time, as evidenced by significant effects of time in each linear mixed model (supplemental Table 4). None of the group-by-time effects were significant: mPCST was not superior to PCST-traditional. Results from the sensitivity analyses (data not shown) using the intent-to-treat sample demonstrated a similar pattern, except that physical symptoms were noninferior at both timepoints.
Table 3.
Noninferiority analyses (n=132)
| Means at each time point | Noninferiority analyses | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
| ||||||||||||||||
| Baseline | Post-treatment | 3 month follow-up | Post-treatment | 3 month follow-up | ||||||||||||
| Outcome | mPCST | PCST-traditional | mPCST | PCST-traditional | mPCST | PCST-traditional | Mean difference | 95% CI | Mean difference | 95% CI | Noninferiority margin | |||||
|
| ||||||||||||||||
| Pain severity | 3.75 | 3.84 | 3.38 | 3.29 | 2.98 | 3.41 | 0.09 | −0.63 | −0.43 | −1.22 to | 1 | |||||
| (0.26) | (0.29) | (0.24) | (0.27) | (0.27) | (0.30) | (0.36) | to | (0.40) | 0.36† | |||||||
| 0.81† | ||||||||||||||||
|
| ||||||||||||||||
| Pain | 3.92 | 3.86 | 3.09 | 3.21 | 2.89 | 3.15 | −0.11 | −0.99 | −0.26 | −1.14 to | 1 | |||||
| interference | (0.27) | (0.31) | (0.29) | (0.33) | (0.29) | (0.34) | (0.44) | to | (0.44) | 0.62† | ||||||
| 0.76† | ||||||||||||||||
|
| ||||||||||||||||
| Physical | 4.54 | 3.84 | 3.56 | 3.49 | 3.65 | 3.39 | 0.07 | −0.71 | 0.26 | −0.58 to | 1 | |||||
| symptoms | (0.23) | (0.26) | (0.26) | (0.30) | (0.28) | (0.32) | (0.39) | to | (0.42) | 1.09 | ||||||
| 0.85† | ||||||||||||||||
|
| ||||||||||||||||
| Psychological | 2.91 | 2.86 | 2.15 | 2.20 | 2.28 | 2.72 | −0.06 | −0.88 | −0.44 | −1.27 to | 1 | |||||
| Distress | (0.27) | (0.30) | (0.27) | (0.31) | (0.28) | (0.32) | (0.42) | to | (0.42) | 0.39† | ||||||
| 0.77† | ||||||||||||||||
|
| ||||||||||||||||
| Physical well- | 2.40 | 2.50 | 2.62 | 2.62 | 2.71 | 2.60 | 0.01 | −0.29 | 0.10 | −0.21 to | −1 | |||||
| being | (0.09) | (0.11) | (0.10) | (0.11) | (0.11) | (0.12) | (0.15) | to | (0.16) | 0.42† | ||||||
| 0.30† | ||||||||||||||||
|
| ||||||||||||||||
| Self-efficacy | 57.61 | 54.10 | 67.31 | 66.09 | 68.21 | 63.53 | 1.22 | −6.56 | 4.69 | −4.10 to | −10 | |||||
| (2.35) | (2.65) | (2.58) | (2.96) | (2.92) | (3.34) | (3.93) | to | (4.44) | 13.48† | |||||||
| 9.00† | ||||||||||||||||
Abbreviations and measures: mPCST, Mobile Pain Coping Skills Training; PCST-traditional, Pain Coping Skills Training delivered in-person; Analyses were conducted on the per-protocol sample (i.e., those who completed all intervention sessions) using linear mixed models. n=128 for the pain severity analysis due to missing data at baseline. Means are estimated marginal means controlling for the baseline value of the outcome (standard error). The mean difference represents mPCST minus PCST-traditional (standard error). For pain severity, pain interference, physical symptoms, and psychological distress, noninferiority is evidenced by the upper limit of the 95% CI being below the noninferiority margin. For physical well-being and self-efficacy, noninferiority is evidenced by the lower limit of the 95% CI being above the noninferiority margin.
Mean difference is noninferior. Supplemental Information provides additional Table 3 details.
Conclusion
This work demonstrates that a videoconference-delivered behavioral cancer pain protocol is highly accessible, feasible, and acceptable. Compared to PCST-traditional participants, mPCST participants were much more likely to adhere to the intervention, complete the intervention in a timely manner, and report that it was easier to attend sessions. Further, mPCST participants reported greater daily use of skills compared to PCST-traditional participants, a critical PCST component. These results are the first to provide evidence of the high level of accessibility of a videoconferencing mHealth behavioral cancer pain intervention with wide implementation potential.
Past work has shown efficacy for behavioral pain management interventions for cancer patients with pain,2 but the burden of receiving such an intervention via in-person delivery has stymied intervention reach. PCST-traditional participants reported significantly greater burden of travel and distance to participate compared to mPCST participants. This burden is also evidenced by our recruitment data: 18% of patients who declined participation indicated the possible travel requirement was their reason for not participating. Even among enrolled participants, overall attrition was greater in PCST-traditional compared to mPCST (30% vs. 16%).
mPCST was more accessible to participants and reduced intervention burden. These findings suggest a behavioral cancer pain intervention delivered by videoconferencing has the potential to dramatically increase patient access. Data from this study suggests intervention efficacy can be maintained when using new mobile technology modalities, as videoconference delivery was noninferior to a traditional face-to-face intervention. Our data showed improvement in both groups on pain and pain-related outcomes that were maintained at three months post-treatment. Importantly, self-efficacy for pain management continued to increase in the mPCST group but started to decline in the PCST-traditional group at the 3-month follow-up. The between-group difference was not statistically significant, but this pattern could suggest when the intervention is delivered with a high level of fidelity (i.e., in 4–6 weeks) the long-term impact is greater.
Study Limitations
This study has limitations. First, there was differential group attrition. This finding was consistent with our hypothesis about the accessibility of mPCST; however, attrition may have affected the noninferiority analyses. Additionally, the study was powered for the primary aim of testing accessibility; the noninferiority analyses may have been underpowered. Coping skills practice was assessed using retrospective recall and thus may have been influenced by recall bias. Pain levels were predominately in the mild-to-moderate range, which may have limited the potential intervention impact. Future studies should consider enrolling patients with more severe pain and assessing pain duration, frequency, and pain treatments to better characterize the sample. Participants were provided with the technology (i.e., iPad, data plan) needed to complete the study; in clinical practice, some patients may not be comfortable or familiar with technology and/or able to access the intervention in their home. We are exploring alternatives to providing hardware and connectivity to patients, including patients using their own devices and implementing iPad kiosks in community clinics.
Clinical Implications
This study is one of the first to directly compare a videoconferencing mobile-based intervention to a traditional, in-person mode of delivery. Findings strongly support the use of videoconferencing for behavioral cancer pain intervention delivery as a highly accessible and efficacious option. This finding is of central importance as videoconferencing dramatically increases the potential for successful dissemination and implementation of an efficacious pain management strategy for cancer patients. As intervention delivery mode options increase, providing patients with a choice of delivery mode may further assist with improving patient access to important pain coping strategies and ultimately improve outcomes.
Supplementary Material
Acknowledgments
Funding was provided by grants from the American Cancer Society (RSG-14–061-01-PCSM, PI: Somers; and PF-17–054-01-PCSM, PI: Winger).
Footnotes
Conflict of Interest Disclosures: None
Data Availability Statement:
The data that support the findings of this study are available from the corresponding author upon reasonable request.
References
- 1.Swarm R, Abernethy AP, Anghelescu DL, et al. Adult cancer pain. J Natl Compr Canc Netw 2010;8:1046–1086. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Syrjala KL, Jensen MP, Mendoza ME, Jean CY, Fisher HM, Keefe FJ. Psychological and behavioral approaches to cancer pain management. J Clin Oncol 2014;32:1703–1711. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Keefe FJ, Abernethy AP, L CC. Psychological approaches to understanding and treating disease-related pain. Annu Rev Psychol 2005;56:601–630. [DOI] [PubMed] [Google Scholar]
- 4.Selove R, Kroll T, Coppes M, Cheng Y. Psychosocial services in the first 30 days after diagnosis: results of a web-based survey of Children’s Oncology Group (COG) member institutions. Pediatr Blood Cancer 2012;58:435–440. [DOI] [PubMed] [Google Scholar]
- 5.Greenberg DB. Barriers to the treatment of depression in cancer patients. J Natl Cancer Inst Monogr 2004:127–135. [DOI] [PubMed]
- 6.Somers TJ, Abernethy AP, Edmond SN, et al. A pilot study of a mobile health pain coping skills training protocol for patients with persistent cancer pain. J Pain Symptom Manage 2015;50:553–558. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Somers TJ, Kelleher SA, Westbrook KW, et al. A small randomized controlled pilot trial comparing mobile and traditional pain coping skills training protocols for cancer patients with pain. Pain Res Treat 2016. [DOI] [PMC free article] [PubMed]
- 8.Attkisson CC, Zwick R. The Client Satisfaction Questionnaire: psychometric properties and correlations with service utilization and psychotherapy outcome. Eval Program Plann 1982;5:233–237. [DOI] [PubMed] [Google Scholar]
- 9.Cleeland CS, Ryan KM. Pain assessment: global use of the Brief Pain Inventory. Ann Acad Med Singapore 1994;23:129–138. [PubMed] [Google Scholar]
- 10.Abernethy AP, Zafar SY, Uronis H, et al. Validation of the Patient Care Monitor (Version 2.0): a review of system assessment instrument for cancer patients. J Pain Symptom Manage 2010;40:545–558. [DOI] [PubMed] [Google Scholar]
- 11.Cella DF, Tulsky DS, Gray G, et al. The Functional Assessment of Cancer Therapy scale: Development and validation of the general measure. J Clin Oncol 1993;11:570–579. [DOI] [PubMed] [Google Scholar]
- 12.Anderson KO, Dowds BN, Pelletz RE, Edwards WT, Peeters-Asdourian C. Development and initial validation of a scale to measure self-efficacy beliefs in patients with chronic pain. Pain 1995;63:77–83. [DOI] [PubMed] [Google Scholar]
- 13.Hox JJ, Moerbeek M, van de Schoot R. Multilevel analysis: techniques and applications Routledge; 2017. [Google Scholar]
- 14.Piaggio G, Elbourne DR, Pocock SJ, Evans SJ, Altman DG, Group C. Reporting of noninferiority and equivalence randomized trials: extension of the CONSORT 2010 statement. JAMA 2012;308:2594–2604. [DOI] [PubMed] [Google Scholar]
- 15.Hui D, Shamieh O, Paiva CE, et al. Minimal clinically important differences in the Edmonton Symptom Assessment Scale in cancer patients: a prospective, multicenter study. Cancer 2015;121:3027–3035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Schwartz MD, Valdimarsdottir HB, Peshkin BN, et al. Randomized noninferiority trial of telephone versus in-person genetic counseling for hereditary breast and ovarian cancer. J Clin Oncol 2014;32:618. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Jones B, Jarvis P, Lewis J, Ebbutt A. Trials to assess equivalence: the importance of rigorous methods. BMJ 1996;313:36–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.