Preparing to Implement a Self-Management Program for Back Pain in New York City Senior Centers: What Do Prospective Consumers Think?

Abstract

Objective

Prior to testing the feasibility/potential efficacy of a newly developed self-management pain program for seniors with back pain, this study sought to: 1) determine prospective consumers’ prior exposure to self-management pain programs, 2) determine their willingness to participate in the new program; and 3) ascertain perceived barriers/facilitators to program participation.

Design

Cross-sectional survey.

Setting

Six senior centers located in New York City.

Participants

We enrolled a race/ethnicity stratified (African American, Hispanic, or non-Hispanic White) sample of 90 subjects who were ages 60 years or older and had chronic back pain.

Results

While 60% of non-Hispanic Whites reported prior participation in a self-management pain program, fewer Hispanic (23%) and African Americans (20%) participants reported prior participation. Most participants (80%) were strongly willing to participate in the new program. Multivariate analyses revealed that only pain intensity had a trend toward significance (p=.07), with higher pain scores associated with greater willingness to participate. Few barriers to participation were identified, however, respondents felt that tailoring the course to best meet the needs of those with physical disabilities, providing flexibility in class timing, and informing individuals about program benefits prior to enrollment could help maximize program reach. No race/ethnicity differences were identified with respect to willingness to participate or program participation barriers.

Conclusions

These data support efforts to disseminate self-management pain programs in older populations, particularly minority communities. The recommendations made by participants can help to guide implementation efforts of the newly developed pain program and may help to enhance both their reach and success.

INTRODUCTION

Among older adults, chronic non-malignant pain is a common, costly, and often disabling condition and one that poses challenges for both health care providers and caregivers [1]. Prevalence estimates vary but indicate that up to 50% of all community-dwelling older adults live with chronic pain [2]. The deleterious consequences of inadequately treated chronic pain are far-reaching and include impaired quality of life, sleep disturbance, cognitive and physical disability, as well as depression, and social withdrawal [2–8]. Indeed, pain is by far the most frequently cited symptom causing activity of daily living disability in later life [9].

A variety of self-management programs have been developed in an effort to reduce the burden of suffering due to chronic pain. Prior research has documented the efficacy of these programs, mostly in non-Hispanic Caucasian populations [10–14]. Despite these successes, one estimate suggests that the programs have reached fewer than 1% of U.S. adults with arthritis and arthritis-related diseases [11]. In addition, other research indicates that exercise and particularly relaxation methods, which constitute key components of self-management programs for pain, are infrequently used by older persons with chronic pain [15–18]. These results suggest limited diffusion of the programs and infrequent use of core program components, and have led authorities to recommend that increased “attention be given to community-based strategies for improving the uptake of self-management strategies for chronic pain” [16].

In an effort to address the aforementioned problems, the investigators developed a self-management pain program targeting community-dwelling older adults with chronic back pain. The group-based program consists of eight, 90-minute sessions offered weekly and provides instruction in the use of specific physical exercises and cognitive (e.g., relaxation, visual imagery) techniques for managing pain. Developing a program for older adults with chronic back pain was deemed important given the substantial physical and psychological morbidity associated with this disorder [6,8,19] and a recent population-based study showing increased rates of back pain across all age groups, including older adults [20].

As a prelude to testing the feasibility and efficacy of the newly developed pain program, this study was conducted to: 1) determine prospective consumers’ prior exposure to self-management pain programs, 2) examine their willingness to participate in the course; and 3) ascertain perceived barriers and facilitators to program participation. Finally, because prior research has demonstrated race/ethnicity differences in types of self-care strategies used to manage pain [21–24], we sought to determine whether the above outcomes varied as a function of participants’ race/ethnicity status.

METHODS

Study Setting & Sample Assembly

We partnered with six senior centers located in New York City to assemble a race/ethnicity stratified sample of older adults with chronic back pain. Two centers provided services to predominantly African American older adults, two served largely Hispanic older adults, and the remaining two provided services to largely non-Hispanic white older adults. Our target goal was to enroll 30 individuals from each race/ethnicity group. Although the centers each served one predominant group, they also provided services to individuals of other race/ethnicities. Subject recruitment methods varied by site. Investigators provided formal presentations to senior center clientele in two centers (at the recommendation of center directors) to inform prospective participants about the research. In the remaining four centers, senior center personnel posted flyers about the research and made announcements about the project during regularly scheduled lunches. Individuals who responded to these advertisements were screened for eligibility by the research staff and were deemed eligible to participate if they: 1) were 60 years of age or older, 2) spoke English or Spanish, and 3) had chronic back pain, defined as providing an affirmative response to the following question: “During the past three months, have you been bothered by pain/discomfort in the back on most days of every month?” Individuals who answered yes were subsequently asked: “Is the back pain/discomfort due to cancer?” All persons who screened positive for chronic back pain provided a no response to the latter question. We did not inquire about the specific location (e.g., thoracic spine, lumbar spine) of the pain. Thus, the sample consisted of individuals with back pain in various locations of the spine and of undetermined etiology.

The research instruments were forward and back translated into Spanish by a professional translator and then pilot tested for comprehension with Spanish speaking older adults. All interviews with Latino participants were conducted by research assistants fluent in Spanish. In preparation for the study, the survey was also piloted in English speaking adults attending senior centers. Written consent was obtained from all participants. Participants were compensated 25 dollars for their time and the Weill Cornell Medical College Institutional Review Board approved the study.

Of the 100 prospective participants who expressed interest in the study, 5 were excluded because they did not have chronic back pain. The remaining 95 individuals enrolled in the study. The convenience sample was composed of 30 Hispanics, 30 African Americans, 30 non-Hispanic Whites, along with five participants who reported other race/ethnicities. The analytic sample for the current study was restricted to the 90 subjects who were Hispanic, African American, or non-Hispanic Caucasian.

Survey Administration

All participants underwent face-to-face interviews conducted by trained interviewers. The average time to complete each interview was 25 minutes (range = 18 to 46 minutes). Interviews with Latino participants were conducted by research assistants fluent in Spanish. The survey consisted of 31-items (individual items are described below) that included previously validated measures as well as measures employed in prior work conducted by the principal investigator [25–28]. Response scales for all quantitative questions included either simple yes/no or Likert style response sets, whereas for the qualitative survey questions open-ended responses were transcribed by the interviewers.

Data Collection

Pain-Related and Pain Management Variables

Average pain intensity using an 11-item numeric rating scale with 0 (no pain/discomfort at all) and 10 (worst pain/discomfort you can imagine) as anchors. Participants were asked to estimate the duration of back pain in either months or years, as appropriate [25]. Level of back pain-related interference experienced while performing activities of daily living [25] was assessed by asking “Over the past month, to what extent has your back pain interfered with your ability to: bathe; dress; transfer; get to places out of walking distance; shop for groceries; prepare meals; and do housework?” Response categories for all seven items consisted of ‘not at all,’ ‘somewhat,’ and ‘a lot.’ To estimate participants’ overall level of back pain-related functional interference, we summed each participant’s score for all seven items. Scores ranged from 0 (no back pain-related interference) to 14 (maximal interference).

The interviewer inquired about current use of pain medications, where response categories (by drug class) included acetaminophen, non-steroidal anti-inflammatory drugs (NSAIDs), opioids, over the counter pain medicines, alternative/complimentary agents, anticonvulsants, antidepressants, and dermal preparations. Patients were given examples of drugs from each class as recall prompts. Response items for each category consisted of either ‘yes’ or ‘no.’ Participants were next asked about their use of non-pharmacologic therapies to manage back pain, including physical agents such as heat/cold, as well as exercise, acupuncture, prayer, relaxation/meditation, and massage. Response items for each category included either ‘yes’ or ‘no.’ To determine participants’ perceived efficacy of the interventions, interviewers asked participants to rate the helpfulness of each strategy they employed using a five-item response scale that ranged from 0 (not helpful at all) to 4 (extremely helpful). Finally, participants’ perceived ability to self-manage pain successfully (pain self-efficacy) was ascertained using the Pain Self Efficacy Questionnaire [29]. Scores on this previously validated measure range from a low of 0 (no self-efficay) to a high of 60 (maximal self-efficacy).

Primary Outcome Variables

Prior exposure to self-management pain programs

To determine participants’ prior exposure(s), interviewers asked “Have you ever participated in a program that taught you techniques to help you better manage pain on your own?” Respondents could answer either ‘yes’ or ‘no.’ Those answering yes were asked to describe the type of program they participated in.

Willingness to participate in a subsequent self-management pain program

Interviewers asked respondents how willing they would be to participate in a new self-management program designed to teach them specific exercises and relaxation techniques in a group setting to help them better manage back pain. Response categories included ‘strongly willing,’ ‘somewhat willing,’ ‘undecided,’ ‘somewhat unwilling,’ and ‘strongly unwilling.’

Perceived barriers and facilitators to program participation

Using barriers identified in previous research [30], participants were asked to rate the extent to which any of the following would make it difficult for them to participate: 1) language/culture problems, 2) transportation concerns, 3) time conflicts, 4) health problems, 5) lack of confidence in one’s ability to exercise, 6) lack of family/social support, 7) fear that participation would cause injury/harm, or 8) concern that treatment would not work. A five-item ordinal response scale was employed for each category ranging from 1 = ‘not a problem/concern at all’ to 5 = ‘a very significant problem/concern.’ A category of ‘other’ was also included to capture barriers not included in the eight-item list.

To identify facilitators to program participation, interviewers asked participants: “Are there things that you can think of that would help you to participate in the program?” Standard probes were used to help clarify participants’ responses and to gain additional details. Interviewers took detailed notes and transcribed all of the open-ended responses.

Other Variables

Information regarding participants’ demographic status was obtained, including age, level of education, living alone versus living with others and marital status. Presence of emotional support (“Is there any one special person you know that you feel very close and intimate with, someone you share confidences and feelings with, someone you feel you can depend on?” and instrumental support (“Is there anyone you can count on to help with daily tasks like grocery shopping, housecleaning, cooking, telephoning, giving you a ride?”) was assessed using a previously validated 2-item instrument [30]. Participants’ level of depressive symptomatology was determined using the previously validated 9-item Patient Health Questionnaire or PHQ-9 [31].

Medical comorbidity was determined by inquiring about the presence of 18 self-reported, physician-diagnosed chronic conditions [32]. Participants’ functional status was assessed by inquiring about their ability to perform three basic activities of daily living (BADL), specifically, bathing, dressing, and transferring; and four instrumental activities of daily living (IADL), specifically, getting to places out of walking distance, shopping for groceries, meal preparation, and doing housework. Response categories for each item included ‘can do task without help,’ ‘needs help,’ and ‘unable to do at all.’

Data Analyses

All variables were examined as a function of the 3-level race/ethnicity variable. Quantitative variables were analyzed in general linear models, whereas dichotomous variables were analyzed in logistic-linear models with binomial error. The primary outcome variable–willingness to participate in the newly developed pain program–was examined in a model that included, a priori, the following independent variables: living alone, age, depressive symptom (i.e., PHQ-9) score, pain intensity, pain interference score, and pain self-efficacy score in addition to race/ethnicity status. Given the modest sample size, we carried out a limited examination of interactions among variables in the a priori model, focusing on variables that were significant at P = 0.20 or better.

Using content analysis [33], two investigators (MP and CR) independently reviewed all responses to the open-ended question about program facilitators (described above). Meaningful phrases were identified by reviewing and then rereviewing the transcripts, and subsequently grouped into discrete categories, which were further condensed into specific themes [33]. The investigators met regularly to compare the groupings, create summary statements for each identified theme, and to reach consensus on any conceptual differences in the coding or category definitions [33].

RESULTS

Characteristics of Study Sample

Non-Hispanic White participants were older (78.33 years) than Hispanic (72.53) and African American (75.03) participants (Table 1). Most subjects were female and lived alone. A majority of non-Hispanic white (73%) and 50% of the African American respondents reported having more than twelve years of education vs. only 17% of Hispanic participants. A substantial majority (89%) reported having emotional support, irrespective of race/ethnicity status. While overall depressive symptom levels were low across all race/ethnicity groups, Hispanic subjects endorsed significantly more depressive symptoms relative to the other two groups.

Table 1.

Characteristics of study sample

	White n=30	Hispanic n=30	African American n=30	All Participants n=90
Demographic
Mean (sd) age in years	78.33 (8.01)	72.53 (7.97)	75.03 (6.75)	75.30 (7.95)
Female	73%	70%	77%	73%
Married	20%	20%	30%	23%
Lives alone	77%	70%	53%	67%
More than 12 yrs of education	73%	17%*	50%†	47%
Psychosocial
Reported having instrumental support	55%	70%	52%	57%
Reported having emotional support	87%	83%	97%	89%
Mean (sd) depressive symptom score (range, 0–27)‡	2.43 (1.77)	4.77 (5.65)†	1.83 (1.78)†	3.01 (3.73)
Medical
Mean (sd) comorbidity score (range, 0–18)‡	2.70 (1.24)	3.07 (1.87)	2.93 (.94)	2.90 (1.39)
Functional status
One or more basic activity of daily living deficits	7%	10%	7%	8%
One or more instrumental activity of daily living deficits	57%	40%	33%	43%
Pain-related characteristics
Mean (sd) duration of back pain in yrs	14.80 (17.28)	11.73 (13.74)	12.77 (16.32)	13.10 (15.63)
Mean (sd) pain intensity (range, 0–10)‡	6.90 (2.20)	7.77 (2.10)	6.40 (2.10)†	7.02 (2.14)
Mean (sd) pain self-efficacy score (range, 0–60)‡	48.43 (8.94)	47.73 (15.36)	53.07 (5.32)	49.74 (10.79)
Back pain-related interference with:
Bathing	50%	43%	50%	48%
Dressing/undressing	53%	47%	47%	49%
Transferring	47%	43%	40%	43%
Getting to places out of walking distance	63%	63%	63%	63%
Shopping	80%	70%	63%	71%
Preparing meals	63%	53%	60%	59%
Doing housework	80%	60%	73%	71%
Mean (sd) pain-related interference score (range, 0–14)‡	5.30 (3.61)	5.20 (4.66)	4.60 (3.70)	5.03 (3.96)

^*Hispanic versus non-Hispanic White pairwise comparison significant at P < 0.05 level.

^†African American versus Hispanic pairwise comparison significant at P < 0.05 level.

^‡Higher scores on the depressive symptom, pain intensity, pain-related interference and medical comorbidity measures reflect worse function or greater impairment/comorbidity. Higher scores on the pain self-efficacy measure reflect better functioning.

Participants reported experiencing back pain for many years (sample mean = 13.10 years, range, 0.5–67) and with high levels of pain intensity (mean = 7.02, range, 2–10). A majority of all participants, irrespective of race/ethnicity status, reported that back pain interfered with their ability to conduct instrumental ADLs (Table 1). Participants reported high pain self-efficacy scores, irrespective of race/ethnicity status (sample mean = 49.74, range, 11–60).

Self-Management Strategies for Pain

A substantial majority (83%) reported using one or more pharmacologic methods for managing pain (Table 2). Hispanic participants were significantly more likely to endorse use of acetaminophen (72%) vs. African Americans (25%) and non-Hispanic Whites (38%). No Hispanic participant and only 17% of African American participants endorsed opioid medication use as a means of managing pain compared to 38% of non-Hispanic Whites. Well over three quarters of each group used non-pharmacologic methods to manage pain, including exercise and prayer. Non-Hispanic Whites and Hispanics were more likely to endorse use of exercise as compared with African Americans. A significantly greater proportion of Hispanic (48%) and African American (46%) participants reported use of prayer as compared to non-Hispanic Whites (4%). Use of cognitive techniques such as relaxation and meditation was infrequently endorsed by all three groups. In terms of the total number of pain management strategies employed (including both pharmacologic and nonpharmacologic approaches), 20% of the sample reported using one or fewer, 38% employed two to three strategies, and the remaining 42% used four or more strategies [mean (standard deviation)= 3.18 (1.94)].

Table 2.

Percentage of sample reporting use of specific strategies to manage back pain

	White n=30	Hispanic n=30	African American n=30	All Participants n=90
Pharmacologic strategies
Any analgesic use	90%	83%	70%	83%
Non-steroidal anti-inflammatory drug	58%	40%	50%	49%
Acetaminophen	38%	72%*	25%†	45%
Other medications‡	43%	56%	33%	44%
Opioid	38%	0%	17%	15%
Non-pharmacologic strategies
Any non-pharmacologic method	87%	83%	80%	83%
Heat/cold	67%	72%	71%	70%
Exercise	83%	68%	42%	64%
Prayer	4%§	48%*	46%	32%
Physical therapy	33%	4%*	29%†	22%
Massage	8%	16%	12%	12%
Relaxation/meditation	8%	4%	8%	7%
Acupuncture	4%	0%	4%	3%

^*Hispanic versus non-Hispanic White pairwise comparison significant at P < 0.05 level.

^†African American versus Hispanic pairwise comparison significant at P < 0.05 level.

^‡Other category includes participants who reported taking one or more of the following medication classes: Anticonvulsants; antidepressants; over the counter medications (e.g., glucosamine); alternative/complementary agents (e.g., herbal preparations); epidural injections; and dermal preparations (e.g., lidoderm patch).

^§Non-Hispanic White versus African American pairwise comparison significant at P < 0.05 level.

Perceived Efficacy of Self-Management Strategies

Table 3 shows that opioid medications were perceived to be the most efficacious pharmacologic method overall. Non-Hispanic Whites were more likely than African Americans to rate NSAIDs as very or extremely helpful (73% vs. 36%, p = 0.05). With respect to nonpharmacologic methods, exercise was rated by 77% of those endorsing this strategy as very or extremely helpful. Use of prayer was rated as the most effective nonpharmacologic strategy followed by massage.

Table 3.

Percentage of sample reporting pain-reduction method to be ‘very’ or ‘extremely’ helpful

	White n=30	Hispanic n=30	African American n=30	All Participants n=90
Pharmacologic Methods
Opioid	90%	*	80%	87%
Acetaminophen	60%	61%	43%	57%
Other medications†	64%	69%	44%	57%
Non-steroidal anti-inflammatory drug	73%	60%	36%	56%
Non-Pharmacologic Methods
Prayer	100%	73%	100%	86%
Massage	100%	100%	50%	80%
Exercise	80%	82%	60%	77%
Relaxation/meditation	50%	100%	67%	67%
Heat/cold	57%	72%	50%	61%
Acupuncture	0%	*	50%	50%
Physical therapy	50%	0%	38%	39%

^*No person in this subgroup reported using this particular method.

^†Other category includes participants who reported taking one or more of the following medication classes: Anticonvulsants; antidepressants; over the counter medications (e.g., glucosamine); alternative/complementary agents (e.g., herbal preparations); epidural injections; and dermal preparations (e.g., lidoderm patch).

Primary Study Outcomes

Table 4 shows that 60% of non-Hispanic Whites reported prior participation in a self-management pain program versus significantly smaller proportions of Hispanic (23%) and African American (20%) participants. The type of program and number of participants reporting participation in each included exercise (n=17), yoga (n=10), tai chi (n=8), physical therapy (n=7), Arthritis Foundation’s Self Help Program (n=4), meditation (n=2), and a course in the use of distraction techniques (n=1).

Table 4.

Prior exposure to self-management pain programs, willingness to participate in new program, and level of perceived barriers to program participation

	White n=30	Hispanic n=30	African American n=30	All Participants n=90
Prior participation in self-management pain program	60%*	23%†	20%	34%
Strongly willing to participate in new pain program	83%	80%	77%	80%
Mean (sd) level of perceived barriers to program participation‡
Language concerns/cultural issues	1.00 (.00)	1.00 (.00)	1.27 (1.02)	1.09 (.59)
Transportation issues	1.30 (1.02)	1.03 (.18)	1.27 (.64)	1.20 (.70)
Time conflicts	1.63 (1.10)	1.70 (.95)	1.20 (.41)§	1.51 (.88)
Other health problems	1.20 (.55)	1.50 (1.11)	1.20 (.66)	1.30 (.81)
Lack of confidence to do exercise	1.10 (.40)	1.30 (1.02)	1.07 (.25)	1.15 (.65)
Little or no support from family	1.23 (.90)	1.50 (1.25)	1.07 (.37)	1.27 (.92)
Fear of pain/injury	1.23 (.82)	1.63 (1.30)	1.17 (.75)	1.34 (.99)
Concern that treatment will not work	1.27 (.78)	1.53 (1.22)	1.00 (.00)§	1.27 (.85)

^*Non-Hispanic White vs. African American pairwise comparison significant at P < 0.05 level.

^†Hispanic versus non-Hispanic White pairwise comparison significant at P < 0.05 level.

^‡Participants were asked to rate the extent to which each barrier may cause difficulty participating in the program. The magnitude of each barrier was assessed using a five-point scale ranging from 1 (not a problem/concern at all) to 5 (very significant problem/concern).

^§African American versus Hispanic pairwise comparison significant at P < 0.05 level.

Most participants (80%) were strongly willing to participate in the newly developed pain program irrespective of race/ethnicity status. We constructed a model (with 7 independent variables) to assess for associations between participant characteristics and willingness to participate in the new program. Only pain intensity had a trend toward significance (p = 0.07), with higher pain intensity scores associated with greater willingness to participate. An examination of the interaction of pain intensity with ethnicity showed this trend most strongly for whites (p=0.10). Race/ethnicity status was not associated with level of willingness to participate in the pain program.

Factors such as language problems, transportation issues, or concerns about the efficacy of the program were not perceived as significant barriers to program participation (Table 4). Several between-group differences were found. Both non-Hispanic White and Hispanic participants were more likely to endorse time conflicts as a barrier relative to African American participants (Table 4). In addition, Hispanic participants reported significantly higher levels of concern about whether the treatment would work relative to African Americans.

When asked about factors that would make it easier to participate in the new program, 41 said they could not think of any facilitators, whereas 49 (54%) provided one or more responses that were grouped into the following categories: 1) Flexibility in program scheduling. Thirty-one participants felt that offering the same class on different days of the week could help to address the problem of competing time demands such as conflicting doctors’ appointments. Offering the program multiple times throughout the year was also suggested as a way to maximize program reach by addressing the fact that many senior center clients spend part of each year in other geographic locations; 2) Need to tailor the program to accommodate individuals with differing functional abilities. Fifteen respondents reported that they would not be able to do certain exercises because of physical limitations/disabilities (e.g., participate in floor exercises on account of bilateral knee replacements) and suggested that the class offer a diverse range of exercises to accommodate those with physical disabilities/impairments; 3) Provide concrete information about program benefits. Eight respondents suggested that educating prospective participants about the value of the program by senior center clients who had previously taken the class or by the class instructor could help to maximize program participation; and 4) other suggestions. One participant felt that hosting the course at the same time and day each week could help to maximize program participation, while another suggested that paying participants for each class attended was the best way to ensure good attendance. No race/ethnicity differences were found with respect to the type and/or frequency of responses provided to the program facilitation question.

DISCUSSION

This study helps to expand our understanding of implementation issues regarding community-based self-management pain programs for older adults in the following ways. All three race/ethnicity groups reported a high willingness to participate in the newly developed pain program, suggestinig that lack of interest on the part of older adults is not the rate-limiting step in dissemination of self-management pain programs. This result is consistent with previous research which found that older persons are willing to adopt diverse self-care strategies to better manage pain [26,27,33,34].

Participants endorsed few barriers to participation in the newly developed pain program. The most frequently endorsed barrier was time conflicts and reflects the fact that many older adults have responsibilities/activities (e.g., caregiving roles, physicians’ appointments) that may limit participation. Transportation was less frequently endorsed as a barrier, likely reflecting the fact that most centers are located in neighborhoods proximate to where clients live and that many of the centers provide transportation to and from their respective facilities. Despite the fact that most participants infrequently endorsed barriers to participation, inquiring about possible barriers is recommended prior to instituting any community-based pain program. For example, for those endorsing “fear of pain or injury” as a barrier to participation, course instructors can inform prospective participants that exercise does not usually exacerbate pain or affect disease progression [15].

Previously identified barriers to engaging in self-management programs for chronic illnesses (not specifically those focusing on pain) include individual-level factors, e.g., depression, fatigue, pain as well as social/environmental-level factors, e.g., lack of emotional support from family and social networks, fear for personal safety, and neighborhood condition [36–39]. Barriers to engaging in self-management pain programs have been previously identified and include time conflicts, concern that the therapy will not work, fear of injury, and depression [26, 40–42].In one recent study [41], level of depressive symptoms was found to impact negatively on participants’ use of exercise as a means of self-managing pain. While overall depression scores were low in our sample, Hispanic participants reported the greatest level of depressive symptoms and also reported the highest number of perceived barriers to program participation. Barriers to program participation and uptake of self-help strategies might be best addressed by offering self-management classes within the community in a safe environment with a trained instructor who is skilled in delivering culturally appropriate learning materials. Program success is likely to be enhanced by incorporation of educationally-appropriate learning materials, tailoring interventions to fit the needs of older adults with diverse functional abilities, and assessing for symptoms of depression (with appropriate treatment referral).

Participants provided valuable suggestions about ways to facilitate program implementation. Tailoring the course to best meet the needs of those with physical limitations/disabilities, providing flexibility in the timing of the classes, and informing prospective participants about program benefits prior to enrollment were felt to be important methods for maximizing enrollment and regular class attendance. Inquiring about potential time conflicts or specific health problems prior to enrollment is also recommended given that a substantial minority of participants noted that class participation could be enhanced by paying attention to these issues.

Our study also found that African American and Hispanic participants reported significantly lower rates of prior exposure to self-management pain programs relative to non-Hispanic Caucasians. This finding suggests the need for further inquiry to understand whether this difference is simply a function of pain management courses not being offered as often in minority communities or minority older adults’ general lack of knowledge about self-management programs for pain. Regardless of the reason(s), this result–along with prior research [43–45] showing substantial disparities in the management of pain as a function of race/ethnicity–provides strong support for future intervention efforts that preferentially target minority communities.

Our study also documented high levels of pain intensity and moderate levels of pain-related interference in daily functioning, reinforcing the appropriateness of targeting community-dwelling older adults attending senior centers. Our investigation also documented high levels of pain self-efficacy across all three race/ethnicity groups. The finding that a substantial proportion of participants reported high levels of pain intensity and pain self-efficacy may indicate that these individuals have confidence that their strategies for dealing with pain can keep the pain from getting even worse. Alternatively, this finding may be tapping into age-related expectations regarding the presence of pain; i.e., many older adults simply expect to have substantial pain as they age [46], irrespective of pain self-efficacy level. A substantial proportion of the sample reported back pain-related interference with daily functioning and also reported high levels of pain self-efficacy. This apparent contradictory finding may be explained by the fact that participants reported pain-related interference (not inability) to conduct functional tasks and may reflect high levels of task persistence in the face of pain. Indeed, other research has shown that perseverance despite pain is a commonly endorsed coping strategy among older adults [47, 48].

Our findings regarding use of strategies to treat pain are consistent with prior research [16,27,34,49].While most participants reported use of analgesic medications to treat pain, ethnic differences were found. For example, approximately 40% of non-Hispanic Caucasian participants reported using an opioid medication to treat their pain versus only 17% of African American and 0% of Hispanic participants. While a majority of non-Hispanic White and Hispanic participants used exercise as a means of managing their back pain, only 42% of African American participants endorsed use of this technique. We did not inquire about the frequency or extent of exercise that participants engaged in, but the substantial use of exercise as a pain management strategy is encouraging given that exercise in this age group has many health dividends [15]. A substantial minority of Hispanic and African American participants reported use of prayer as a means of dealing with their pain. This finding is consistent with earlier studies [50–52] and reinforces the important role that spiritual/religious practices can play in the management of chronic illnesses such as chronic pain. Finally, cognitive techniques such as relaxation and meditation were infrequently endorsed by all three groups. Given their established efficacy as a means of managing pain [10], efforts to promote increased awareness and use of these approaches are warranted.

The data regarding perceived efficacy of the various self-management strategies for pain are based on small numbers, and larger studies are needed to confirm these preliminary findings. The results suggest, however, several possible areas of future inquiry. First, the class of medications that had the highest level of perceived efficacy (opioids) was the least frequently employed by all three race/ethnicity groups, suggesting possible barriers to use at the patient and/or provider level. Identification of these barriers will be important as a recently published pain management guideline recommends that clinicians strongly consider opioid therapy for older patients who continue to report substantial pain or experience pain-related impairment in function [53]. Second, approximately one in four participants reported exercising to manage pain but did not find this method helpful. In terms of implementing self-management pain programs in community settings, the data support having program instructors gather information about participants’ methods for managing pain and perceived efficacy of the methods at the time of enrollment. It would be particularly important to work with participants who do not find exercise to be helpful to understand the reasons why (e.g., not doing the exercises correctly, doing the wrong exercises, have concern that exercise will make pain worse). Similarly, for individuals who have already incorporated exercise into their daily routine and are receiving benefit, more effort could be directed towards helping them learn relaxation/cognitive techniques. Finally, encouraging participants who have found other helpful techniques to manage pain (e.g., use or prayer or massage) to share this information during the class could provide co-participants with new ideas about ways to manage pain.

Our study has several limitations. We assembled a convenience sample of older adults with chronic back pain who may have been intrinsically motivated to participate in this study. Our findings may therefore not be generalizable to the larger population of community-dwelling older adults who experience chronic back pain. Participants may not have understood the definitions used to describe certain pain-reduction strategies (e.g., relaxation), which could have led to an underestimation of the true prevalence of their use. Participants were also financially compensated for their time, which may have caused some to be reluctant to share negative concerns about the newly developed pain program.

In conclusion, our study found that older adults with chronic back pain were strongly willing to participate in a self-management pain program. While few barriers to participation were identified, many participants felt that tailoring the course to best meet the needs of those with physical limitations/disabilities, providing flexibility in the timing of the classes, and informing prospective participants about program benefits prior to enrollment could help to maximize enrollment and adherence during the eight-week program. These recommendations can help to guide implementation efforts of self-management pain programs and perhaps enhance both their reach and success.