Original Article Effect of a Virtual Pain Coach on Pain Management Discussions: A Pilot Study Deborah Dillon McDonald, RN, PhD,* Stephen Walsh, ScD,* Cunegundo Vergara, BS,† and Timothy Gifford, MD* ---
-
From the *University of Connecticut, Storrs; †Hartford Hospital, Hartford, Connecticut. Address correspondence to Deborah Dillon McDonald, RN, PhD, University of Connecticut School of Nursing, 231 Glenbrook Road, Storrs, CT 06269-2026. E-mail: [email protected] Received September 15, 2010; Revised March 20, 2011; Accepted March 24, 2011. 1524-9042/$36.00 Ó 2013 by the American Society for Pain Management Nursing http://dx.doi.org/10.1016/ j.pmn.2011.03.004
ABSTRACT:
There is a need to enhance patient and practitioner pain communications. A pain communication plus virtual pain coach intervention was tested in the primary care setting for the effect on communication of osteoarthritis pain information by older adults aged $60 years, on practitioners’ pain management changes, and on older adults’ reduced pain and depressive symptoms 1 month later. A randomized controlled pilot study design was used. Twenty-three older adults with osteoarthritis pain were randomly assigned to the pain communication plus virtual pain coach group or the pain communication–only group. Pain communication consisted of a video of important osteoarthritis pain information. The coach consisted of practicing out loud with a virtual pain coach via laptop computer. Pain and depressive symptoms were measured with, respectively, the Brief Pain Inventory Short Form and the Beck Depression Inventory II before intervention and 1 month later. Immediately after the intervention, older adults had their primary care visits, which were audiotaped, transcribed, and content analyzed for older adults’ communicated pain information and practitioners’ pain management changes. Older adults in the pain communication plus virtual pain coach group described significantly more pain source information and were prescribed significantly more osteoarthritis pain treatments than older adults in the pain communication–only group. A nonsignificant trend in pain intensity and depressive symptoms reduction resulted for older adults in the pain communication plus virtual pain coach group 1 month later. The virtual pain coach presents a possible strategy for increasing pain management discussions between practitioners and older adults with persistent pain. Ó 2013 by the American Society for Pain Management Nursing
Pain Management Nursing, Vol 14, No 4 (December), 2013: pp 200-209
Virtual Pain Coach
More than 48% of Americans aged $65 years have been diagnosed with arthritis (Lawrence, et al., 2008). Although use of acetaminophen along with exercise has been proven to be effective as the initial treatment for significantly reducing osteoarthritis pain (Zhang, Moskowitz, Nuki, Abramson, Altman, & Tugwell, 2008), many older adults do not use these proven treatments (Merkle & McDonald, 2009). Disuse of effective treatments might be due to lack of awareness of efficacious treatment options, insufficient trial time for treatments, or inadequate pain relief from treatments. To improve treatment, older adults need to be able to talk with their health care practitioners to achieve treatment regimens that significantly reduce their individual pain. The aim of the present study was to pilot test an innovative feasible pain communication plus virtual practitioner coaching intervention with older adults for the effect on: 1) older adults’ communication of clinically relevant osteoarthritis pain information during a health care visit; 2) practitioners’ pain management changes in response to the pain information; and 3) older adults’ reports of pain intensity, functional interference from the pain, and depressive symptoms. The importance of effective communication about pain with health care practitioners is underscored by results from a recent nationwide telephone survey. Whereas about 75% of participants reporting recent pain also reported talking with their health care practitioners, only a little more than one-half (56%) of adults who talked with their health care practitioner reported substantial pain relief. Those who reported that they did not believe their practitioners understood their pain were six times less likely to report adequate pain relief. Older adults were particularly vulnerable and were less than one-half as likely to report good pain relief as a result of their health care visit (Shi, Langer, Cohen, & Cleeland, 2007). Older adults might need additional support in talking about their pain with their health care practitioners. Interventions that only encourage patients to talk about their pain might be inadequate for promoting effective pain communication. Increased communication between patients and practitioners was not associated with increased pain relief, perhaps because communication was restricted to discussing pain treatments and to asking the patient to alert practitioners when pain occurred (Carlson, Youngblood, Dalton, Blau, & Lindley, 2003). Patients and family caregivers have clearly identified the need for improved communication with their health care practitioners (Kimberlin, Brushwood, Allen, Radson, & Wilson, 2004). Patients and practitioners need research-based support to help them communicate about pain in ways that lead to greater pain relief.
201
Promising coaching interventions to assist patients to describe their pain have been tested. Patients with cancer have been individually coached to talk with their health care practitioners during a 20-minute session immediately before their office visit by learning to identify treatment goals, formulate questions, and practice talking about their pain management issues (Oliver, Kravitz, Kaplan, & Meyers, 2001). Patients with cancer have also been taught about their pain management in weekly 6-week sessions by a specially trained oncology nurse and coached to talk with their health care practitioners (Miaskowski, Dodd, West, Paul & Koo, 2004). The significant pain reduction found in both of these studies suggests that patients can be assisted to communicate their pain effectively and to receive treatments that significantly reduce their pain. The cost of the individual coaching interventions might limit the widespread use of coaching interventions. And both studies involved patients with cancer pain; individual coaching interventions for patients with different pain etiologies might not be as effective in eliciting more responsive pain management from practitioners. Coaching via the internet by nurses has been conducted with chronic pain patients (Allen, Iezzoni, Huang, Huang, & Leveille, 2008). However, only onehalf of participants accessed the intervention website before their practitioner visit, and pain outcomes were not reported. Furthermore, participants were primarily white and college educated. Pain communication coaching interventions need to be linked to patients’ pain outcomes and to be accessed easily by all people, including economically disadvantaged persons and persons from diverse racial and ethnic backgrounds. Communication accommodation theory (CAT) has been used to guide causal research about communication behaviors with older adults (Ryan, Hamilton, & See, 1994). CAT describes the motivations and behaviors of people as they adjust their communication in response to their own needs and the perceived behavior of the other person (Fox & Giles, 1993; Giles, 1973). Observation of the other person’s behavior during communicating can be used to enhance communication. These observations, termed attuning strategies, include use of interpretability strategies and discourse management (Coupland, Coupland, Giles, & Henwood, 1988). Interpretability strategies involve use of terminology that clearly and explicitly relates important information to the practitioner. Examples of strategies to increase interpretability include older adults’ use of the 0-to-10 pain intensity scale and their description of current pain treatments and the effectiveness of the treatments. Discourse management
202
McDonald et al.
involves selecting the topic, contributing to the discussion by taking your turn, and maintaining topic focus. It follows that the more skillful that older adults are in using communication strategies, the more likely they will be to convey important osteoarthritis pain information to the practitioners and to be prescribed more effective pain management treatments. The following hypotheses were tested: 1. Older adults in the pain communication plus virtual pain coach group describe more pain information to practitioners than those in the pain communication– only group. 2. Older adults in the pain communication plus virtual pain coach group receive more pain regimen changes from practitioners than those in the pain communication– only group. 3. Older adults in the pain communication plus virtual pain coach group report reduced pain intensity, interference with functional status, and depressive symptoms compared with those in the pain communication–only group.
METHODS Design A randomized controlled pilot study design was used. Older adults were randomized within practitioner practice to control for individual practitioner practice effects. The pain communication plus virtual pain coach group was taught about pain communication and coached to communicate by the virtual practitioner, and the pain communication–only group was taught only about pain communication. Baseline measurements of pain intensity, functional interference from pain, and depressive symptoms were taken, and the participants viewed their respective intervention while waiting in the examination room to talk with their practitioner. Immediate outcome measures during the health care visit included the amount of important distinct pain information described by older adults to their practitioner and the number of appropriate pain management changes made by practitioners in response to the information. Short-term pain outcomes for the older adult included reductions in pain intensity, interference with functional status, and depressive symptoms at 1 month after intervention. Sample and Setting The initial two participants were recruited at an urban ambulatory care clinic of a large university-affiliated hospital. Owing to the need for English-speaking participants, the remaining participants were recruited from a suburban university primary care practice site. All participants included in the final analysis were from the second site owing to incomplete data
from the initial site. Community-dwelling adults, aged $60 years who spoke, read, and understood English and who had self-identified osteoarthritis pain at a level of $4 on a 0-to-10 scale most days of the preceding month were eligible. People with self-identified malignant pain or those scheduled for joint replacement surgery or actively considering joint replacement surgery within the next month were ineligible. The practitioner sample consisted of primary care physicians who consented to the audiotaping of their health care visit with older adults participating in the study. Measures Demographic Information. The demographic information included age, gender, race, ethnic group, highest completed education, income, marital status, the number of years they have had osteoarthritis pain most days of the month, the type, month, and year of any joint replacement surgery, if they were under the care of a practitioner for their osteoarthritis, if they were under the care of a practitioner for their pain management, if they were under the care of a rheumatologist, and if they were under the care of a pain management specialist. Brief Pain Inventory Short Form (BPI-SF). The BPI-SF consists of 15 questions that measure pain location, intensity, pain treatment, and the effect of pain on mood and everyday activities. Zalon (1999) compared the BPI-SF with the Short-Form McGill Pain Questionnaire (SF-MPQ) with a group of surgical patients. The correlation between the BPI-SF and the SF-MPQ for pain over the preceding 24 hours was 0.61 (p < .001), supporting concurrent validity. Cronbach alpha for the overall BPI-SF has been reported to be 0.77-0.87 (McDonald, Thomas, Livingston, & Severson, 2005; Zalon, 1999). Beck Depression Inventory II (BDI-II). The 21item BDI-II, based on the Diagnostic Symptoms Manual IV criteria for depression, measures the intensity of depressive symptoms. Internal consistency for the total BDI-II scale was 0.92 (Harris & d’Eon, 2008). Test-retest reliability over an average of 3 days was 0.96 (Sprinkle, Lurie, Insko, Atkinson, Jones, & Bissada, 2006). The BDI-II has been recommended for use in chronic pain clinical trials, with a decrease of $5 points considered to be clinically important (Dworkin, Turk, Wyrwich, Beaton, Cleeland, & Zavisic, 2008). Procedure After university Institutional Review Board approval, consents were obtained from the practitioners. Adults with upcoming clinic appointments were sent a letter from the director of the primary care site informing
Virtual Pain Coach
them of the study and directing interested people who met the preliminary eligibility criteria to telephone for further information. Interested participants were further screened for eligibility, and a study appointment was made immediately before their scheduled clinic appointment. Screening included asking participants to self-identify if they experienced pain from osteoarthritis during most days of the preceding month and to rate their pain on a 0-to-10 pain scale. Older adults with a pain level of $4 most days of the preceding month were eligible. Participants were also screened for exclusion criteria, which included current cancer pain or planned joint replacement surgery within the next month. The study visit began in a private room in the clinic with informed consent and Health Insurance Portability and Accountability Act authorization. Baseline measures (demographics, BPI-SF, and BDI-II) were obtained, after which participants were automatically randomized within the respective practitioners’ practice to one of the two conditions through use of the current clock and a Mersenne Twister (Matsumoto & Nishimura, 1998) random number generator program. The intervention was audiotaped to identify which condition participants were in and to provide a manipulation check that older adults in the virtual coach group actually practiced talking about their pain with the virtual coach. The audiotapes were transcribed and content analyzed. Practitioner consent to audiotape discussion during the health care visit was obtained before starting the study, and reaffirmed before each older adult clinic visit. Practitioners were also given a personal copy of the American Pain Society’s (2002) Guideline for the Management of Pain in Osteoarthritis, Rheumatoid Arthritis, and Juvenile Chronic Arthritis and a copy of the summary of the American Geriatrics Society’s (2009b) clinical practice guideline ‘ Pharmacological Management of Persistent Pain in Older Adults.’’ Practitioners were unaware to which condition their patients were randomized. Older adults’ communication of important pain information during the health care visit was audiotaped and transcribed for content analysis. Health care practitioner prescription of pain treatments was recorded on the same audiotape, transcribed, and content analyzed for appropriate treatment changes. One month after the clinic visit, older adults were telephoned and administered the BPI-SF and BDI-II. Older adults were financially compensated for the baseline and 1-month measures and received a pain management pamphlet from the Arthritis Foundation. Intervention Pain Communication Intervention. The pain communication intervention incorporated two attuning
203
strategies from CAT: interpretability and discourse management. Interpretability was increased by greater explicitness. By describing important types of osteoarthritis pain information that practitioners need to assess, older adults learn how to be more explicit. Interpretability was also increased by greater clarity through use of terminology familiar to the practitioner (e.g., pain intensity of 7 on a 0-10 scale), increased focus on the pain management topic, and increased complexity through greater breadth and depth of pain information described to the practitioner. Discourse management was enhanced by learning to select pain management as a topic to discuss with the practitioner. The pain communication–only intervention consisted of a video of a female health care practitioner describing the types of osteoarthritis pain information important to share with health care practitioners. Content for the pain communication education portion of the intervention was based on the American Pain Society osteoarthritis pain management guidelines. The 3-minute pain communication education videotape, professionally produced and displayed on a laptop computer, was tested and found to encourage older adults to describe a mean of 5.0 (SD 2.37) items of important pain information (McDonald, Gifford, & Walsh, 2011). Virtual Pain Coach Intervention. A computergenerated and -displayed representation of a female practitioner was used for the virtual pain coach to assist older adults to practice talking with their practitioner, because female practitioners encourage significantly more communication from their patients (Roter, Hall, & Aoki, 2002). The virtual pain coach orally instructed the older adults to practice talking with her about their pain. The coach asked the older adults to talk about their pain, detected and responded to pauses by encouraging the older adults to describe additional information, provided general positive feedback on the practice session, and concluded the coaching session by encouraging the older adults to share their important pain information with their practitioner. Older adults verbally interacted with the virtual pain coach as they practiced talking about their pain. The virtual pain coaching intervention reinforced the interpretability strategies taught in the videotape portion of the intervention, and introduced two important discourse management strategies: practice in selecting relevant pain management topics to discuss with the practitioner; and practice in turn taking. The virtual pain coach was an innovation developed by the authors and previously tested with older adults in a nonmedical setting (McDonald et al., 2011). Content Analysis. Older adults’ communication of important pain information during their health care visit
204
McDonald et al.
was content analyzed following Krippendorff’s (2004) approach for content analysis and using the same a priori criteria from the American Pain Society’s (2002) Guidelines for the Management of Pain in Osteoarthritis, Rheumatoid Arthritis, and Juvenile Chronic Arthritis as used in previous studies (McDonald et al., 2011; McDonald, Shea, Rose, & Fedo, 2009). Practitioner changes in pain treatment were analyzed using a priori pain treatment criteria from the same American Pain Society (2002) guidelines and the American Geriatrics Society’s (2009b) clinical practice guideline ‘ Pharmacological Management of Persistent Pain in Older Adults.’ The unit of analysis was any word or phrase that described one of the a priori criteria. One point was given for each word or phrase describing a criterion. Repeated use of the same word or phrase was counted only the initial time. Each distinctly different word or phrase about the same criterion was credited with one point. The a priori criteria for older adults’ pain communication included type of pain (nociceptive/neuropathic), quality of pain, pain source, location, intensity, duration/time course, affect, effect on personal lifestyle, functional interference, current pain treatments, effectiveness of prescribed treatments, analgesic side effects, weight management to ideal body weight, exercise regimen, or physical therapy and/or occupational therapy, and indications for surgery. The a priori osteoarthritis pain management criteria for the practitioners included prescriptions of or changes in structured patient education, acetaminophen, nonsteroidal antiinflammatory drug, exercise, physical therapy, occupational therapy, assistive device, weight loss, tramadol, adjunctive analgesic for neuropathic pain, intra-articular hyaluronic acid, glucocorticoid injection, massage, use of hot or cold application, distraction, mental imagery, cognitive restructuring, activity pacing, relaxation, pleasant activity and goal setting, stress management training, opioid, and surgical intervention. Two raters independently coded the data and compared results. Disagreements were resolved through discussion. A third rater, blind to participant condition, independently coded the data and compared results with the initial coding. Data resulting from the second coding was entered into an SPSS database. Older adults’ pain communication score was calculated by summing all of the important pain content described by participants while talking with their health care practitioners. Practitioner pain management changes were calculated by summing all of the recommended pain management changes prescribed by practitioners during the health care visit. The manipulation check for older adults in the virtual coach group consisted of reviewing the transcripts from the audiotaped intervention session. Participants
who orally responded to the virtual coach or to the pain communication videotape were coded as responders. Statistical Analyses Chi-square tests and independent-samples t tests were applied to the baseline values of the study variables to assess effectiveness of the randomization procedure at creating comparable samples of virtual pain coach and pain communication education–only participants. The primary dependent variables represented older adults’ communication of important pain content, practitioner prescription of recommended osteoarthritis pain treatments, and changes in older adult pain intensity, interference with functional status, and depressive symptoms during the study period. When values of these dependent variables within each group were consistent with a normal distribution, means were contrasted between the virtual pain coach and pain communication education-only groups using the independent-samples t test. When the dependent variables deviated from normality, but were consistent with a Poisson distribution, means were compared via a generalized linear model that incorporated a logarithmic link function and provided an estimate of the ratio of means. When a dependent variable was not consistent with either the normal or the Poisson distribution, the Mann-Whitney U test was applied to determine whether one condition was stochastically larger, i.e., tended to produce larger values, than the other.
RESULTS A total of 23 eligible older adults completed the study, with 11 older adults in the pain communication plus virtual coach group and 12 in the pain communication–only group. The older adults were primarily seen by five of the consenting primary care physicians. The majority of the older adults were white collegeeducated women with no previous joint replacement surgery. Most participants reported not being under the care of a generalist or a specialist for their arthritis or their arthritis pain. Table 1 presents demographic frequencies for the full sample and the pain communication plus virtual coach and pain communication– only groups. The mean age was 74.3 (SD 7.60) years. Older adults communicated a mean 11.1 (SD 7.58) items of important pain information, and practitioners prescribed a mean 1.0 (SD 1.19) pain treatment changes. Pain intensity and interference with functional status decreased a mean 0.2 (SD 1.82) and 0.2 (SD 1.86), respectively, and the BDI-II depressive symptoms score decreased a mean 2.0 (SD 4.78) from the pretest to the posttest. Means and SDs for age, older
205
Virtual Pain Coach
TABLE 1. Frequencies of Older Adult Sample Characteristics, n (%)
Variable
Full (n ¼ 23)
Virtual Coach (n ¼ 11)
Women 19 (82.6) 11 (100.0) Race Asian 2 (8.7) 0 (0.0) Black 1 (4.3) 1 (9.1) White 19 (82.6) 10 (90.9) Multiracial 1 (4.3) 0 (0.0) Education
-
From the *University of Connecticut, Storrs; †Hartford Hospital, Hartford, Connecticut. Address correspondence to Deborah Dillon McDonald, RN, PhD, University of Connecticut School of Nursing, 231 Glenbrook Road, Storrs, CT 06269-2026. E-mail: [email protected] Received September 15, 2010; Revised March 20, 2011; Accepted March 24, 2011. 1524-9042/$36.00 Ó 2013 by the American Society for Pain Management Nursing http://dx.doi.org/10.1016/ j.pmn.2011.03.004
ABSTRACT:
There is a need to enhance patient and practitioner pain communications. A pain communication plus virtual pain coach intervention was tested in the primary care setting for the effect on communication of osteoarthritis pain information by older adults aged $60 years, on practitioners’ pain management changes, and on older adults’ reduced pain and depressive symptoms 1 month later. A randomized controlled pilot study design was used. Twenty-three older adults with osteoarthritis pain were randomly assigned to the pain communication plus virtual pain coach group or the pain communication–only group. Pain communication consisted of a video of important osteoarthritis pain information. The coach consisted of practicing out loud with a virtual pain coach via laptop computer. Pain and depressive symptoms were measured with, respectively, the Brief Pain Inventory Short Form and the Beck Depression Inventory II before intervention and 1 month later. Immediately after the intervention, older adults had their primary care visits, which were audiotaped, transcribed, and content analyzed for older adults’ communicated pain information and practitioners’ pain management changes. Older adults in the pain communication plus virtual pain coach group described significantly more pain source information and were prescribed significantly more osteoarthritis pain treatments than older adults in the pain communication–only group. A nonsignificant trend in pain intensity and depressive symptoms reduction resulted for older adults in the pain communication plus virtual pain coach group 1 month later. The virtual pain coach presents a possible strategy for increasing pain management discussions between practitioners and older adults with persistent pain. Ó 2013 by the American Society for Pain Management Nursing
Pain Management Nursing, Vol 14, No 4 (December), 2013: pp 200-209
Virtual Pain Coach
More than 48% of Americans aged $65 years have been diagnosed with arthritis (Lawrence, et al., 2008). Although use of acetaminophen along with exercise has been proven to be effective as the initial treatment for significantly reducing osteoarthritis pain (Zhang, Moskowitz, Nuki, Abramson, Altman, & Tugwell, 2008), many older adults do not use these proven treatments (Merkle & McDonald, 2009). Disuse of effective treatments might be due to lack of awareness of efficacious treatment options, insufficient trial time for treatments, or inadequate pain relief from treatments. To improve treatment, older adults need to be able to talk with their health care practitioners to achieve treatment regimens that significantly reduce their individual pain. The aim of the present study was to pilot test an innovative feasible pain communication plus virtual practitioner coaching intervention with older adults for the effect on: 1) older adults’ communication of clinically relevant osteoarthritis pain information during a health care visit; 2) practitioners’ pain management changes in response to the pain information; and 3) older adults’ reports of pain intensity, functional interference from the pain, and depressive symptoms. The importance of effective communication about pain with health care practitioners is underscored by results from a recent nationwide telephone survey. Whereas about 75% of participants reporting recent pain also reported talking with their health care practitioners, only a little more than one-half (56%) of adults who talked with their health care practitioner reported substantial pain relief. Those who reported that they did not believe their practitioners understood their pain were six times less likely to report adequate pain relief. Older adults were particularly vulnerable and were less than one-half as likely to report good pain relief as a result of their health care visit (Shi, Langer, Cohen, & Cleeland, 2007). Older adults might need additional support in talking about their pain with their health care practitioners. Interventions that only encourage patients to talk about their pain might be inadequate for promoting effective pain communication. Increased communication between patients and practitioners was not associated with increased pain relief, perhaps because communication was restricted to discussing pain treatments and to asking the patient to alert practitioners when pain occurred (Carlson, Youngblood, Dalton, Blau, & Lindley, 2003). Patients and family caregivers have clearly identified the need for improved communication with their health care practitioners (Kimberlin, Brushwood, Allen, Radson, & Wilson, 2004). Patients and practitioners need research-based support to help them communicate about pain in ways that lead to greater pain relief.
201
Promising coaching interventions to assist patients to describe their pain have been tested. Patients with cancer have been individually coached to talk with their health care practitioners during a 20-minute session immediately before their office visit by learning to identify treatment goals, formulate questions, and practice talking about their pain management issues (Oliver, Kravitz, Kaplan, & Meyers, 2001). Patients with cancer have also been taught about their pain management in weekly 6-week sessions by a specially trained oncology nurse and coached to talk with their health care practitioners (Miaskowski, Dodd, West, Paul & Koo, 2004). The significant pain reduction found in both of these studies suggests that patients can be assisted to communicate their pain effectively and to receive treatments that significantly reduce their pain. The cost of the individual coaching interventions might limit the widespread use of coaching interventions. And both studies involved patients with cancer pain; individual coaching interventions for patients with different pain etiologies might not be as effective in eliciting more responsive pain management from practitioners. Coaching via the internet by nurses has been conducted with chronic pain patients (Allen, Iezzoni, Huang, Huang, & Leveille, 2008). However, only onehalf of participants accessed the intervention website before their practitioner visit, and pain outcomes were not reported. Furthermore, participants were primarily white and college educated. Pain communication coaching interventions need to be linked to patients’ pain outcomes and to be accessed easily by all people, including economically disadvantaged persons and persons from diverse racial and ethnic backgrounds. Communication accommodation theory (CAT) has been used to guide causal research about communication behaviors with older adults (Ryan, Hamilton, & See, 1994). CAT describes the motivations and behaviors of people as they adjust their communication in response to their own needs and the perceived behavior of the other person (Fox & Giles, 1993; Giles, 1973). Observation of the other person’s behavior during communicating can be used to enhance communication. These observations, termed attuning strategies, include use of interpretability strategies and discourse management (Coupland, Coupland, Giles, & Henwood, 1988). Interpretability strategies involve use of terminology that clearly and explicitly relates important information to the practitioner. Examples of strategies to increase interpretability include older adults’ use of the 0-to-10 pain intensity scale and their description of current pain treatments and the effectiveness of the treatments. Discourse management
202
McDonald et al.
involves selecting the topic, contributing to the discussion by taking your turn, and maintaining topic focus. It follows that the more skillful that older adults are in using communication strategies, the more likely they will be to convey important osteoarthritis pain information to the practitioners and to be prescribed more effective pain management treatments. The following hypotheses were tested: 1. Older adults in the pain communication plus virtual pain coach group describe more pain information to practitioners than those in the pain communication– only group. 2. Older adults in the pain communication plus virtual pain coach group receive more pain regimen changes from practitioners than those in the pain communication– only group. 3. Older adults in the pain communication plus virtual pain coach group report reduced pain intensity, interference with functional status, and depressive symptoms compared with those in the pain communication–only group.
METHODS Design A randomized controlled pilot study design was used. Older adults were randomized within practitioner practice to control for individual practitioner practice effects. The pain communication plus virtual pain coach group was taught about pain communication and coached to communicate by the virtual practitioner, and the pain communication–only group was taught only about pain communication. Baseline measurements of pain intensity, functional interference from pain, and depressive symptoms were taken, and the participants viewed their respective intervention while waiting in the examination room to talk with their practitioner. Immediate outcome measures during the health care visit included the amount of important distinct pain information described by older adults to their practitioner and the number of appropriate pain management changes made by practitioners in response to the information. Short-term pain outcomes for the older adult included reductions in pain intensity, interference with functional status, and depressive symptoms at 1 month after intervention. Sample and Setting The initial two participants were recruited at an urban ambulatory care clinic of a large university-affiliated hospital. Owing to the need for English-speaking participants, the remaining participants were recruited from a suburban university primary care practice site. All participants included in the final analysis were from the second site owing to incomplete data
from the initial site. Community-dwelling adults, aged $60 years who spoke, read, and understood English and who had self-identified osteoarthritis pain at a level of $4 on a 0-to-10 scale most days of the preceding month were eligible. People with self-identified malignant pain or those scheduled for joint replacement surgery or actively considering joint replacement surgery within the next month were ineligible. The practitioner sample consisted of primary care physicians who consented to the audiotaping of their health care visit with older adults participating in the study. Measures Demographic Information. The demographic information included age, gender, race, ethnic group, highest completed education, income, marital status, the number of years they have had osteoarthritis pain most days of the month, the type, month, and year of any joint replacement surgery, if they were under the care of a practitioner for their osteoarthritis, if they were under the care of a practitioner for their pain management, if they were under the care of a rheumatologist, and if they were under the care of a pain management specialist. Brief Pain Inventory Short Form (BPI-SF). The BPI-SF consists of 15 questions that measure pain location, intensity, pain treatment, and the effect of pain on mood and everyday activities. Zalon (1999) compared the BPI-SF with the Short-Form McGill Pain Questionnaire (SF-MPQ) with a group of surgical patients. The correlation between the BPI-SF and the SF-MPQ for pain over the preceding 24 hours was 0.61 (p < .001), supporting concurrent validity. Cronbach alpha for the overall BPI-SF has been reported to be 0.77-0.87 (McDonald, Thomas, Livingston, & Severson, 2005; Zalon, 1999). Beck Depression Inventory II (BDI-II). The 21item BDI-II, based on the Diagnostic Symptoms Manual IV criteria for depression, measures the intensity of depressive symptoms. Internal consistency for the total BDI-II scale was 0.92 (Harris & d’Eon, 2008). Test-retest reliability over an average of 3 days was 0.96 (Sprinkle, Lurie, Insko, Atkinson, Jones, & Bissada, 2006). The BDI-II has been recommended for use in chronic pain clinical trials, with a decrease of $5 points considered to be clinically important (Dworkin, Turk, Wyrwich, Beaton, Cleeland, & Zavisic, 2008). Procedure After university Institutional Review Board approval, consents were obtained from the practitioners. Adults with upcoming clinic appointments were sent a letter from the director of the primary care site informing
Virtual Pain Coach
them of the study and directing interested people who met the preliminary eligibility criteria to telephone for further information. Interested participants were further screened for eligibility, and a study appointment was made immediately before their scheduled clinic appointment. Screening included asking participants to self-identify if they experienced pain from osteoarthritis during most days of the preceding month and to rate their pain on a 0-to-10 pain scale. Older adults with a pain level of $4 most days of the preceding month were eligible. Participants were also screened for exclusion criteria, which included current cancer pain or planned joint replacement surgery within the next month. The study visit began in a private room in the clinic with informed consent and Health Insurance Portability and Accountability Act authorization. Baseline measures (demographics, BPI-SF, and BDI-II) were obtained, after which participants were automatically randomized within the respective practitioners’ practice to one of the two conditions through use of the current clock and a Mersenne Twister (Matsumoto & Nishimura, 1998) random number generator program. The intervention was audiotaped to identify which condition participants were in and to provide a manipulation check that older adults in the virtual coach group actually practiced talking about their pain with the virtual coach. The audiotapes were transcribed and content analyzed. Practitioner consent to audiotape discussion during the health care visit was obtained before starting the study, and reaffirmed before each older adult clinic visit. Practitioners were also given a personal copy of the American Pain Society’s (2002) Guideline for the Management of Pain in Osteoarthritis, Rheumatoid Arthritis, and Juvenile Chronic Arthritis and a copy of the summary of the American Geriatrics Society’s (2009b) clinical practice guideline ‘ Pharmacological Management of Persistent Pain in Older Adults.’’ Practitioners were unaware to which condition their patients were randomized. Older adults’ communication of important pain information during the health care visit was audiotaped and transcribed for content analysis. Health care practitioner prescription of pain treatments was recorded on the same audiotape, transcribed, and content analyzed for appropriate treatment changes. One month after the clinic visit, older adults were telephoned and administered the BPI-SF and BDI-II. Older adults were financially compensated for the baseline and 1-month measures and received a pain management pamphlet from the Arthritis Foundation. Intervention Pain Communication Intervention. The pain communication intervention incorporated two attuning
203
strategies from CAT: interpretability and discourse management. Interpretability was increased by greater explicitness. By describing important types of osteoarthritis pain information that practitioners need to assess, older adults learn how to be more explicit. Interpretability was also increased by greater clarity through use of terminology familiar to the practitioner (e.g., pain intensity of 7 on a 0-10 scale), increased focus on the pain management topic, and increased complexity through greater breadth and depth of pain information described to the practitioner. Discourse management was enhanced by learning to select pain management as a topic to discuss with the practitioner. The pain communication–only intervention consisted of a video of a female health care practitioner describing the types of osteoarthritis pain information important to share with health care practitioners. Content for the pain communication education portion of the intervention was based on the American Pain Society osteoarthritis pain management guidelines. The 3-minute pain communication education videotape, professionally produced and displayed on a laptop computer, was tested and found to encourage older adults to describe a mean of 5.0 (SD 2.37) items of important pain information (McDonald, Gifford, & Walsh, 2011). Virtual Pain Coach Intervention. A computergenerated and -displayed representation of a female practitioner was used for the virtual pain coach to assist older adults to practice talking with their practitioner, because female practitioners encourage significantly more communication from their patients (Roter, Hall, & Aoki, 2002). The virtual pain coach orally instructed the older adults to practice talking with her about their pain. The coach asked the older adults to talk about their pain, detected and responded to pauses by encouraging the older adults to describe additional information, provided general positive feedback on the practice session, and concluded the coaching session by encouraging the older adults to share their important pain information with their practitioner. Older adults verbally interacted with the virtual pain coach as they practiced talking about their pain. The virtual pain coaching intervention reinforced the interpretability strategies taught in the videotape portion of the intervention, and introduced two important discourse management strategies: practice in selecting relevant pain management topics to discuss with the practitioner; and practice in turn taking. The virtual pain coach was an innovation developed by the authors and previously tested with older adults in a nonmedical setting (McDonald et al., 2011). Content Analysis. Older adults’ communication of important pain information during their health care visit
204
McDonald et al.
was content analyzed following Krippendorff’s (2004) approach for content analysis and using the same a priori criteria from the American Pain Society’s (2002) Guidelines for the Management of Pain in Osteoarthritis, Rheumatoid Arthritis, and Juvenile Chronic Arthritis as used in previous studies (McDonald et al., 2011; McDonald, Shea, Rose, & Fedo, 2009). Practitioner changes in pain treatment were analyzed using a priori pain treatment criteria from the same American Pain Society (2002) guidelines and the American Geriatrics Society’s (2009b) clinical practice guideline ‘ Pharmacological Management of Persistent Pain in Older Adults.’ The unit of analysis was any word or phrase that described one of the a priori criteria. One point was given for each word or phrase describing a criterion. Repeated use of the same word or phrase was counted only the initial time. Each distinctly different word or phrase about the same criterion was credited with one point. The a priori criteria for older adults’ pain communication included type of pain (nociceptive/neuropathic), quality of pain, pain source, location, intensity, duration/time course, affect, effect on personal lifestyle, functional interference, current pain treatments, effectiveness of prescribed treatments, analgesic side effects, weight management to ideal body weight, exercise regimen, or physical therapy and/or occupational therapy, and indications for surgery. The a priori osteoarthritis pain management criteria for the practitioners included prescriptions of or changes in structured patient education, acetaminophen, nonsteroidal antiinflammatory drug, exercise, physical therapy, occupational therapy, assistive device, weight loss, tramadol, adjunctive analgesic for neuropathic pain, intra-articular hyaluronic acid, glucocorticoid injection, massage, use of hot or cold application, distraction, mental imagery, cognitive restructuring, activity pacing, relaxation, pleasant activity and goal setting, stress management training, opioid, and surgical intervention. Two raters independently coded the data and compared results. Disagreements were resolved through discussion. A third rater, blind to participant condition, independently coded the data and compared results with the initial coding. Data resulting from the second coding was entered into an SPSS database. Older adults’ pain communication score was calculated by summing all of the important pain content described by participants while talking with their health care practitioners. Practitioner pain management changes were calculated by summing all of the recommended pain management changes prescribed by practitioners during the health care visit. The manipulation check for older adults in the virtual coach group consisted of reviewing the transcripts from the audiotaped intervention session. Participants
who orally responded to the virtual coach or to the pain communication videotape were coded as responders. Statistical Analyses Chi-square tests and independent-samples t tests were applied to the baseline values of the study variables to assess effectiveness of the randomization procedure at creating comparable samples of virtual pain coach and pain communication education–only participants. The primary dependent variables represented older adults’ communication of important pain content, practitioner prescription of recommended osteoarthritis pain treatments, and changes in older adult pain intensity, interference with functional status, and depressive symptoms during the study period. When values of these dependent variables within each group were consistent with a normal distribution, means were contrasted between the virtual pain coach and pain communication education-only groups using the independent-samples t test. When the dependent variables deviated from normality, but were consistent with a Poisson distribution, means were compared via a generalized linear model that incorporated a logarithmic link function and provided an estimate of the ratio of means. When a dependent variable was not consistent with either the normal or the Poisson distribution, the Mann-Whitney U test was applied to determine whether one condition was stochastically larger, i.e., tended to produce larger values, than the other.
RESULTS A total of 23 eligible older adults completed the study, with 11 older adults in the pain communication plus virtual coach group and 12 in the pain communication–only group. The older adults were primarily seen by five of the consenting primary care physicians. The majority of the older adults were white collegeeducated women with no previous joint replacement surgery. Most participants reported not being under the care of a generalist or a specialist for their arthritis or their arthritis pain. Table 1 presents demographic frequencies for the full sample and the pain communication plus virtual coach and pain communication– only groups. The mean age was 74.3 (SD 7.60) years. Older adults communicated a mean 11.1 (SD 7.58) items of important pain information, and practitioners prescribed a mean 1.0 (SD 1.19) pain treatment changes. Pain intensity and interference with functional status decreased a mean 0.2 (SD 1.82) and 0.2 (SD 1.86), respectively, and the BDI-II depressive symptoms score decreased a mean 2.0 (SD 4.78) from the pretest to the posttest. Means and SDs for age, older
205
Virtual Pain Coach
TABLE 1. Frequencies of Older Adult Sample Characteristics, n (%)
Variable
Full (n ¼ 23)
Virtual Coach (n ¼ 11)
Women 19 (82.6) 11 (100.0) Race Asian 2 (8.7) 0 (0.0) Black 1 (4.3) 1 (9.1) White 19 (82.6) 10 (90.9) Multiracial 1 (4.3) 0 (0.0) Education
