A Pilot Study of a Mobile Health Pain Coping Skills Training Protocol for Patients with Persistent Cancer Pain

Abstract

Context

Pain coping skills training (PCST) interventions have shown efficacy for reducing pain and providing other benefits in patients with cancer. However, their reach is often limited because of a variety of barriers (e.g., travel, physical burden, cost, time).

Objectives

This study examined the feasibility and acceptability of a brief PCST intervention delivered to patients in their homes using mobile health (mHealth) technology. Pre-to-post intervention changes in pain, physical functioning, physical symptoms, psychological distress, self-efficacy for pain management, and pain catastrophizing also were examined.

Methods

Patients with a diagnosis of breast, lung, prostate, or colorectal cancer who reported persistent pain (N=25) participated in a four-session intervention delivered using mHealth technology (video-conferencing on a tablet computer). Participants completed measures of pain, physical functioning, physical symptoms, psychological distress, self-efficacy for pain management, and pain catastrophizing. We also assessed patient satisfaction.

Results

Participants completed an average of 3.36 (SD=1.11) of the four intervention sessions for an overall session completion rate of 84%. Participants reported that the program was of excellent quality and met their needs. Significant pre- to post-intervention differences were found in pain, physical symptoms, psychological distress, and pain catastrophizing.

Conclusion

The use of mHealth technology is a feasible and acceptable option for delivery of PCST for patients with cancer. This delivery mode is likely to dramatically increase intervention access for cancer patients with pain compared to traditional in-person delivery. Preliminary data also suggest that the program is likely to produce pre- to post-treatment decreases in pain and other important outcomes.

Introduction

Patients with cancer report pain to be one of their most distressing symptoms.¹ A meta-analysis (N=52) reported that cancer pain prevalence is greater than 50%.² A multi-country study found that 50% or more of cancer patients (N=5084) reported moderate-to-severe pain.³ Pain in cancer patients is related to higher levels of physical disability and psychological distress.^4,5 Higher levels of cancer pain have been shown to be related to decreased survival time across cancers,⁵ and pain interference can be a predictor of overall survival time.⁶

Analgesics are the primary therapy for treating cancer-related pain;⁷ behavioral cancer pain interventions can be an efficacious adjuvant therapy.^8,9 These intervention protocols teach behavioral and cognitive strategies for pain management.¹⁰ Behavioral cancer pain interventions have led to significant reductions in pain in 65% to 85% of trials.^8,11 The National Institutes of Health (NIH) guidelines recommend that behavioral pain interventions be integrated into cancer treatment.

Intervention barriers limit the use of behavioral cancer pain interventions.¹⁰ To date, most behavioral cancer pain interventions are conducted through in-person sessions at a medical center. Barriers to in-person interventions include time constraints, transportation difficulties, and distance.¹² Cost factors also pose barriers to in-person interventions.^13,14 Cancer patients who reported higher levels of time, effort, and cost barriers to treatments reported lower levels of functional and psychological well-being.¹⁵.

Advances in mobile health (mHealth) technologies can decrease barriers that limit access to behavioral pain interventions. Studies examining mHealth strategies to manage chronic medical illnesses (i.e., diabetes) and psychiatric disorders have found that mHealth delivery is feasible, acceptable, and efficacious.^16-18 Prior to widespread implementation of novel mHealth behavioral cancer pain interventions they need to be tested for feasibility, acceptability, and efficacy.

We piloted an mHealth behavioral cancer pain intervention that capitalizes on the advantages of technologies by increasing the reach of the intervention and potentially enhancing intervention efficacy. Our work builds on and extends work done by Keefe et al.^19,20 that has developed and tested Pain Coping Skills Training (PCST) protocols for patients with chronic disease. PCST was designed to help patients with persistent pain acquire mastery of skills that can enhance their pain management.²¹

The novel study intervention (mobile PCST [mPCST]) was delivered in patients' homes through live video-conferencing (i.e., Skype) with a therapist using a tablet computer. mPCST focuses on enhancing the efficacy of pain coping skills training by using technologies to address social cognitive factors that may influence patients' confidence to manage their pain (i.e., mastery, vicarious learning, verbal encouragement, negative responses to skills). mPCST allows the patient to acquire, practice, and master pain coping skills in their natural environment. The therapist models skills and explains skills use by others (i.e., vicarious learning). As the patient practices skills in their home, the therapist provides real-time feedback and problem solving (i.e., verbal encouragement, addressing negative responses).

The primary study aim was to examine whether mPCST would be feasible and acceptable. The secondary study aim was to examine the initial efficacy of mPCST by examining intervention-related changes in pain, physical functioning, physical symptoms, psychological distress, self-efficacy for pain management, and pain catastrophizing.

Methods

Participants (N=25) had a diagnosis of breast, lung, colorectal, or prostate cancer. Participants were 18 years of age or older and had two clinical pain ratings of ≥3 (on a 0 to 10 scale). Participants were excluded if they were cognitively impaired or were unable to speak English. Recruitment took place between May 29, 2012 and August 9, 2013.

A study team member explained study procedures and administered the baseline assessment. Participants were given an iPad to take home to complete the intervention. All tablet computers had data plans for Internet access. Two therapists conducted the intervention – a licensed clinical psychologist and an advanced clinical psychology doctoral student supervised closely by the psychologist. The post-treatment assessment was completed approximately one week following intervention completion.

Duke University Medical Center's Institutional Review Board approved study procedures; participants provided informed consent. As part of the informed consent process, participants were informed in writing of the risk of confidentiality associated with using video-conferencing via the Internet. They were encouraged to read the security information associated with the video-conferencing program used (i.e., Skype) and provided with the Internet address that detailed security for the program.

Measures

Demographic and medical data were collected through electronic medical records (EMR). Patient self-report measures were collected with a secure web-based assessment website.

Patient Demographic and Medical Variables

Age, race, gender, cancer diagnosis and date, and comorbid medical disorders were collected through EMRs and/or self-report. Participants' home zip codes were used to calculate mileage to the medical center.

Pain Severity

Pain severity was assessed with the Brief Pain Inventory (BPI).¹⁵ The BPI assesses pain at its “worst”, “least”, “average”, and “now” (0 = no pain to 10 = pain as bad as you can imagine). Items were referenced to the last seven days. A composite score was used. This measure is recommended for use in all pain clinical trials.^22,23 Reliability was excellent (0.89, 0.87).

Physical Functioning

The physical functioning scale of the Patient Care Monitor (PCM) v2.0 was used.^16,17 The PCM has been validated against other symptom inventories and quality of life (QOL) scales.^16,17 The four items on this scale ask about patients' ability to run, do light physical work or fun activities, do hard physical work or fun activities, and to function normally, referencing the last seven days. The response scale is 0 = not a problem to 10 = as bad as possible (Cronbach's alpha = 0.80 and 0.88).

Physical Symptoms

The physical symptom scale of the PCM was used.^16,17 This subscale has five items asking about fatigue, concentration, pain, sleepiness, and insomnia (0 = not a problem to 10 = as bad as possible). Cronbach's alphas were 0.70 and 0.83.

Psychological Distress

Psychological distress was measured with four PCM items assessing crying or feeling like crying, being worried, feeling nervous, tense, or anxious, and feeling sad or depressed in the last seven days. The response scale was 0 = not a problem to 10 = as bad as possible (Cronbach's alpha = 0.83).

Pain Self-Efficacy

The self-efficacy for pain management subscale of the Chronic Pain Self-Efficacy Scale was used.²⁴ This subscale contains five items that inquire about patients' certainty regarding degree of pain control, pain during daily activities, controlling pain during sleep, and making pain reductions without extra medication. The response scale was 10 = very uncertain to 100 = very certain. This scale has shown good reliability²⁴ and has been used with cancer patients.²⁵ Cronbach's alphas were 0.76 and 0.60.

Pain Catastrophizing

The six-item pain catastrophizing subscale of the Coping Strategies Questionnaire was used.²⁶ Items ask about participants' tendency to catastrophize when faced with pain (e.g., “When I feel pain it is awful and it overwhelms me”); the response scale is 0 = never to 6 = always. Items are summed; this scale has good reliability in cancer patients.²⁷ Cronbach's alphas were excellent (0.95 and 0.92).

Mobile Pain Coping Skills

mPCST comprised four 30-45 minute sessions delivered via video-conferencing (i.e., Skype) to the participant at their home. The protocol skills were designed to enhance patients' self-efficacy to: a) use attention diversion techniques to decrease pain, b) manage their pain to engage in activities that will decrease their pain, disability, and distress, and c) use cognitive restructuring to decrease pain catastrophizing, which can result in decreased pain, disability, and distress. In protocol design, careful attention was given to using the mHealth technology to capitalize on enhancing session factors proposed by social cognitive theory to be related to self-efficacy (i.e., mastery, vicarious learning, verbal encouragement, decreasing negative reactions).

In session, skills practice was reviewed, new skills were learned, practiced and reviewed, and new home practice assignments were given. In Session 1, participants learned about the Gate Control Theory²⁸ and were taught progressive muscle relaxation (PMR) through didactics and therapist modeling (i.e., vicarious learning). In Session 2, activity-rest cycle and pleasant activity scheduling were taught. Activity-rest cycle teaches patients how to schedule activity to be productive and active, but avoid increased pain from overactivity. Pleasant activity scheduling provides patients with pain distraction, improves their pain by increasing their activity, and is associated with lower distress. In Session 3, participants were taught the cognitive model and asked to identify negative thoughts that increase their pain and negative mood and generate more neutral or positive self-statements that lead to better outcomes. In Session 4, participants learned mini-relaxation practices and guided imagery.

Statistical Analyses

Data were analyzed using SPSS 21.0 (IBM Corp., Armonk, NY). Descriptive statistics were calculated for demographic, medical, and study-related variables. Paired samples t-tests were used to examine differences from pre- to post-intervention.

Results

Table 1 displays participant demographic and medical data.

Table 1.

Participant Characteristics and Summary of Self-Report Data (N=25).

	M	SD	%	N
Age	53.88	12.59
Gender (% female)			76.0	19
Race
White			76.0	19
Black			24.0	6
Relationship Status
Married/life partner			58.3	14
Single			16.7	4
Divorced/Widowed			12.5	3
Unknown			12.5	3
Cancer Type
Breast			48.0	12
Lung			16.0	4
Prostate			16.0	4
Colorectal			20.0	5
Time since initial diagnosis (months)	42.52	60.91
Active cancer treatment in the last 7 days a			60.0	15
Pill/anticancer drug			36.0	9
Chemotherapy			32.0	8
Hormone therapy			7.7	2
Radiation			4.0	1
Surgery			4.0	1
Vaccine			4.0	1
Other			4.0	1
Comorbidities b			44.0	11
COPD			8.0	2
Hypertension			28.0	7
Osteoarthritis			12.0	3
Rheumatoid Arthritis			8.0	2
Crohn's disease or IBS			8.0	2
Clinical Pain Score 1	5.28	2.28
Clinical Pain Score 2	5.35	2.12
Distance from Medical Center (miles)	68.70	68.39

Note.

^a7 participants reported two or more treatments in the last week.

^b4 participants reported more than one comorbid disorder.

Feasibility

Participants completed an average of 3.36 (SD=1.11; 84%) of the four intervention sessions; 72% of participants completed all sessions. All participants completed the pre-intervention assessment and 84% completed the post-intervention assessment. Among non-completers, one never started the study, one completed three sessions, and two completed one session. Reasons for non-completion were: lost to contact (2), illness (1), and unknown (1).

Acceptability

Participants rated the program quality as good or excellent (62% excellent). All participants reported that they received the kind of information and skills that they wanted, and 67% and 33% respectively said they would definitely or probably recommend the program to a friend. All participants were either very satisfied (67%) or mostly satisfied (33%) with the program, and the majority of them (95.2%) said that it helped their pain management. Qualitative feedback from participants was positive. One participant said: “This program has taught me how to cope with pain and how to relax my body…when you think positively about what is going on then you can change how you feel.”

Baseline Characteristics

Participants' baseline pain severity score was in the moderate range (M=4.75, SD=1.97). They reported moderate difficulty with physical functioning (M=5.09, SD=2.33) and a moderate level of physical symptoms (M=4.29, SD=1.64). Participants' self-efficacy for managing pain was somewhat low (M=58.08, SD=17.17). Patients reported a moderate level of psychological distress (M=3.04, SD=2.28) and low levels of pain catastrophizing (M=1.59, SD=1.27, range 0-6).

Pre-and Post-Intervention Differences

Table 2 presents pre- and post-intervention scores (M, SD) and the results of the t-tests. Following the intervention, participants reported significantly decreased pain severity (t=2.92, P=0.009), physical symptoms (t=3.84, P=0.001), psychological distress (t=4.31, P<0.001), and pain catastrophizing (t=3.15, P=0.005). There were no statistically significant changes in pain self-efficacy; however, the mean score on pain self-efficacy did increase (M=58.08, SD=17.17 vs. M=62.57, SD=13.82, t=-1.34, P=0.19).

Table 2.

Means and Paired Samples T Tests of Study Variables.

	Pre-Treatment		Post-Treatment		t	p
	M	SD	M	SD
Pain Severity	4.75	1.97	3.37	1.63	2.92	0.009
Physical Functioning	5.09	2.33	4.63	2.70	1.20	0.24
Physical Symptoms	4.29	1.64	2.80	1.92	3.84	0.001
Distress	3.04	2.28	1.32	1.58	4.31	<0.001
Pain Self-Efficacy	58.08	17.17	62.57	13.82	-1.34	0.19
Pain Catastrophizing	1.59	1.27	0.85	0.84	3.15	0.005

Discussion

Cognitive-behavioral approaches to pain management, including PCST, have shown efficacy, yet persistent patient access barriers limit their routine use. The study intervention was designed to be a highly accessible and efficacious behavioral cancer pain intervention. To our knowledge, this was the first study to use video-conferencing technology to deliver a PCST intervention to patients with cancer pain. This intervention is unique because it was delivered using mHealth technology, was brief, and capitalized on the unique advantages of having patients participate in the intervention from home. The intervention had a high level of feasibility and acceptability and produced positive changes for patients' reports of pain and other outcomes. This intervention may hold strong promise for increasing routine implementation.

This pilot work found the mHealth PCST intervention to be feasible, with 84% of participants completing the study. Participants in this study were patients with cancer of whom almost half reported having a comorbid medical condition. Given the significant disease burden, an 84% completion rate suggests that mobile technologies increase participant access and improve participation in such an intervention. Further, participants lived on average 69 miles from the medical center where they received their cancer care. We suspect that participants living this distance from the cancer center would not have been able to complete an in-person PCST intervention, thus access to an mHealth intervention provided them with beneficial care they would not have otherwise received.

For most participants, this was the first time that they used a mobile device to receive a medical intervention. Overall, despite the novelty of delivery modality, participants rated the quality of the intervention as excellent and reported being satisfied with the intervention. Participants in this study said that the mPCST intervention met their needs; over 95% reported that the program helped them to understand the overall experience of pain, and over 90% of patients reported that the program taught them skills that improved their pain coping.

This brief intervention delivered via Skype produced significant changes in patients' symptoms and functioning. Participants reported significant reductions in pain, physical symptoms, psychological distress, and pain catastrophizing. Given that many of the participants in this trial were undergoing active treatment for cancer, these improvements in pain and other symptoms are particularly noteworthy.

Our small sample size limits our ability to statistically control for demographic or medical variables. A larger sample would provide us more power to look at other variables that may contribute to the efficacy of the intervention. Further, there was no control group; this is an important area of future work that we are currently undertaking by conducting a randomized controlled trial examining mPCST vs. an in-person condition.

We recruited a diverse sample of patients with multiple types of cancer. Participants lived as far as 238 miles from the medical center, encompassing a wide range of geographical locations, including many participants who lived in medically underserved areas. We also recruited patients who were experiencing significant pain, by ensuring that patients reported two clinical pain scores of 3 or higher. Further, we used clinical data from participants' medical providers to identify patients with high levels of pain to improve continuity of care and increase the feasibility of receiving patient referrals, as no additional assessment by providers was necessary.

Several areas of future work are indicated. First, it is important to re-emphasize the potential advantages of providing patients with this intervention in their own homes. This mode of delivery decreases persistent patient access barriers to behavioral cancer pain interventions, but also may increase the efficacy of the intervention. Delivery of this intervention in a patient's home allows acquisition, practice, and mastery of pain coping skills in the environment where they are expected to use the skills. This eliminates the assumption of generalization from the medical center to home. The therapist is able to observe skills practice in the home and provide real-time feedback and problem solving (e.g., addressing negative responses). Second, use of video-conferencing for an intervention could easily incorporate the patient's partner or other caregiver. Past work has found including family members in behavioral interventions for patients with cancer to be efficacious. Video-conferencing also could involve a team of health care providers (e.g., team meeting) that could reinforce positive changes patients are making or work to problem solve around challenges. For instance, PCST might be combined with pain medication management through video-conferencing between the patient, a behavioral therapist, and a physician. Finally, future work in mHealth applications should consider moderators of treatment response (e.g., cancer type, pain interference). Moderators can assist in intervention application to patients most likely to benefit.