Applied Nursing Research xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Applied Nursing Research journal homepage: www.elsevier.com/locate/apnr

Predictors of self-management for chronic low back pain Jennifer Kawi, PhD, MSN ⁎ Department of Physiological Nursing, School of Nursing, University of Nevada, Las Vegas

a r t i c l e

i n f o

Article history: Received 16 November 2013 Revised 23 January 2014 Accepted 3 February 2014 Available online xxxx Keywords: Self-management Chronic low back pain Predictors

a b s t r a c t Aims: (a) Identify variables that predict self-management (SM) of chronic low back pain (CLBP), and (b) evaluate differences in these variables between participants in specialty pain centers (SPCs) and primary care clinics (PCCs). Background: Chronic low back pain is highly prevalent in various healthcare settings. Self-management strategies are recommended in pain care guidelines to help address CLBP. However, the evidence of SM effectiveness in CLBP remains unclear. Self-management may be effective for only certain patients. Hence, identifying the predictors to SM of CLBP is essential to help recognize the best responders to SM programs. Method: Secondary analysis was conducted on data collected from two CLBP primary research studies in SPCs (N = 110) and PCCs (N = 120). General linear modeling was utilized for the combined sample of 230 participants and for each practice setting. Results: Overall, in SPCs and PCCs combined, five variables were found to be predictors of SM: age, SM support, education, overall health, and helpfulness of pain management. In SPCs, SM support, support received from other than healthcare providers, religion or spirituality, and overall health were identified as significant predictors to SM. In PCCs, both SM support and overall health were also significant predictors. In addition, those with higher income scored better in SM. Conclusions: Findings provide essential information to healthcare providers in intervening appropriately toward engaging CLBP patients in SM. Other strategies need to be identified for those who do not respond effectively to SM strategies. Published by Elsevier Inc.

1. Introduction

2. Background

Chronic low back pain (CLBP) is the most common chronic pain complaint in the United States, afflicting 28.4% of people 18 years of age or over (National Center for Health Statistics, 2012). With the prevalence of CLBP and associated high healthcare costs, the Institute of Medicine, 2011 strongly emphasized the importance of self-management (SM) strategies to help reduce pain suffering and costs. Self-management encourages active patient participation so that patients are engaged in managing and improving their own health. Self-management has been shown to significantly improve outcomes and reduce healthcare costs (National Council on Aging, 2012). However, evidence of SM effectiveness for CLBP remains unclear (Oliveira et al., 2012). Self-management may be therapeutic for only a subset of patients. Therefore, to help alleviate the physical, psychological, and financial burden of CLBP, evaluating patientrelated variables to predict those who are the best responders to SM is critical to the goals of the IOM and the healthcare system related to CLBP.

Chronic low back pain is a discomfort of varying degree and character that continuously or intermittently persists for at least 3 months (National Institute of Arthritis and Musculoskeletal and Skin Diseases, 2009). The pervasiveness of CLBP is responsible for total healthcare costs of $100–200 billion annually in the United States, of which two-thirds account for indirect costs such as decreased wages and lost productivity (Freburger et al., 2009). Direct costs to the healthcare system include primary pain care, specialty care, inpatient services, rehabilitation, and pharmaceuticals (Freburger et al., 2009). Specifically in relation to pharmaceuticals, the widespread use of opioids in the management of CLBP adds to the healthcare challenges. Opioid abuse, misuse, and its diversion are common problems (Centers for Disease Control, Prevention, 2011). In 2010, the number of prescriptions and resulting sale of opioids would have supplied every American adult with 5 mg of hydrocodone every 4 hours for 1 month (CDC, 2011). In addition, almost 15,000 deaths were attributed to prescription painkillers during that same year (CDC, 2011). While the appropriate use of opioids can decrease CLBP, adequate SM can potentially reduce misuse of opioids through patient engagement in health-directed behaviors, thereby encouraging appropriate use and reducing dependence on external sources (IOM, 2011).

⁎ School of Nursing, University of Nevada, Las Vegas, 4505 Maryland Parkway, Las Vegas, NV 89154–3018, USA. Tel.: +1 702 895 5930; fax: +1 702 895 4807. E-mail address: [email protected]. 0897-1897/$ – see front matter. Published by Elsevier Inc. http://dx.doi.org/10.1016/j.apnr.2014.02.003

Please cite this article as: Kawi, J., Predictors of self-management for chronic low back pain, Applied Nursing Research (2014), http://dx.doi. org/10.1016/j.apnr.2014.02.003

2

J. Kawi / Applied Nursing Research xxx (2014) xxx–xxx

The theoretical foundation for this research was based on Lorig and Holman's (2003) concept of SM. Self-management is described as the performance of tasks and skills with self-efficacy so that patients are activated to make appropriate decisions and engage in health-directed behaviors (Lorig & Holman, 2003). The emphasis on improved strategies and support for SM are identified as essential in several CLBP research studies (Cooper, Smith, & Hancock, 2009; Crowe, Whitehead, Gagan, Baxter, & Panckhurst, 2010; Liddle, Baxter, & Gracey, 2007). Other studies have shown the effectiveness of SM in reducing pain intensity and disability of CLBP and improving overall function (Coudeyre et al., 2006; May, 2010; Schulz, Rubinell, & Hartung, 2007; Sokunbi, Cross, Watt, & Moore, 2010). Consequently, clinical practice guidelines have endorsed the importance of SM in CLBP (Arnstein & Marie, 2010; Institute for Clinical Systems Improvement, 2011; IOM, 2011; Toward Optimized Practice, 2009). However, a CLBP systematic review conducted by Oliveira et al. (2012) reported only modest effects of SM on pain and disability, although with an impact that was similar to more costly and intensive interventions. Interventions like back surgery and injections are invasive and expensive while SM has been shown to be cost-effective (IOM, 2011; National Council on Aging, 2012). Nevertheless, it is likely that individuals vary in their response to SM. This suggests the need to identify variables that can best predict which patients will respond more to SM. Evaluating these variables can allow healthcare providers to identify patients who will likely respond more to SM strategies in order to maximize its effectiveness in both the primary care and specialty pain settings while tailoring to individual needs and investigating more appropriate strategies to non-responders. 3. Purpose The primary aim of this research was to identify variables that predict SM of CLBP (aim 1). This study also evaluated the differences in these variables between participants in specialty pain centers (SPCs) and those in primary care clinics (PCCs; aim 2). 4. Methods This research study was designed conducting secondary data analysis using general linear modeling to identify variables that could predict which patients would respond best to SM of CLBP. The variables analyzed were: perceived SM support (support received from healthcare providers), pain intensity, functional ability, mental health state, and various demographic variables. Sources of the data were from two previous CLBP primary studies (Kawi, 2012; Kawi, in press). The first study was conducted in two SPCs, the second in four PCCs, both in the western United States. Both primary studies used correlational and mediation analysis, plus qualitative content analysis to describe several pain- and patient-related variables. Identifying variables that could predict SM was not addressed in the original analyses of either primary study. Additionally, this current research compared the predictive variables of SM of CLBP between participants in SPCs and participants in PCCs.

4.2. Measures Data were collected using a demographic survey and four selfreport measures. All of the following measures were used in both primary studies analyzed for this current research. 4.2.1. Patient Activation Measure (PAM) The PAM evaluates the knowledge, skills, and behaviors in SM of chronic illnesses (Hibbard, Stockard, Mahoney, & Tusler, 2004). Using a 4-point Likert scale, the PAM is converted to scores indicating the patient's activation or engagement in SM and categorized according to four specified cut-off points. These points represent a progressive hierarchical order: level 1, individuals are starting to take a role in SM; level 2, individuals are gaining confidence and knowledge about selfcare; level 3, individuals are actively practicing SM; and level 4, individuals are working to maintain their health even under stressful conditions (Hibbard et al., 2004). The PAM was conceptually validated with an original 22-item measure, and reliability was stable (.9 to .91) across various chronic conditions (Hibbard et al., 2004). It was reduced to 13 items for ease of completion with infit values ranging from .92 to 1.05 and outfit values from .85 to 1.11 indicating conformity of all items in the measure (Hibbard, Mahoney, Stockard, & Tusler, 2005). 4.2.2. Patient Assessment of Chronic Illness Care This measure evaluates patient perspectives on their chronic illness care examining SM support or support received from healthcare providers using a 5-point Likert scale. Items include questions on patient engagement, decision making, goal-setting, problem-solving, and follow-up (Glasgow et al., 2005). It is a validated 20-item measure, correlating moderately with measures of primary care and patient activation, with an overall internal consistency of .93 (Glasgow et al., 2005). 4.2.3. Oswestry Disability Index This is a specific measure for low back pain evaluating pain-related function (Fairbank & Pynsent, 2000). Version 2.1a contains items on pain intensity, personal care, lifting, walking, sitting, standing, sleeping, sex life, social life, and traveling. This instrument was validated with the Short Form 36 and the visual analogue pain scale with correlations at r = .62, and concurrent validity with the RolandMorris Questionnaire, another common disability measure, at r = .77 (Fairbank & Pynsent, 2000). A test–retest reliability of r = .83 was reported with internal consistency ranging from .71 to .87 (Fairbank & Pynsent, 2000). 4.2.4. Mental Health Inventory This measure was derived from the RAND Medical Outcomes Study: 36-Item Short Form Survey evaluating positive and negative aspects of mental health state from psychological well-being to distress (Veit & Ware, 1983). It is a 5-item questionnaire with a 6point Likert scale. Berwick et al. (1991) validated this measure with three common mental health screening questionnaires with moderate to strong correlations and an overall reliability of .82 (Berwick et al., 1991). 4.3. Data Analysis

4.1. Sample The inclusion criteria for the two primary studies (Kawi, 2012; Kawi, in press) were adults over 18 years of age with at least 3 months of doctor-diagnosed, nonmalignant CLBP, recruited through convenience sampling. Since all collected data were deidentified, this research was deemed exempt from review by the site's institutional review board.

Data were entered into a secured computer using the Statistical Package for the Social Sciences, version 20. Prior to scoring, missing cells for the PAM and the Patient Assessment of Chronic Illness Care measures were replaced with the most frequent response category so that all patients could be scored to prevent problems in analyzing data especially during regression analysis. Missing values comprised less than 5% of the data.

Please cite this article as: Kawi, J., Predictors of self-management for chronic low back pain, Applied Nursing Research (2014), http://dx.doi. org/10.1016/j.apnr.2014.02.003

J. Kawi / Applied Nursing Research xxx (2014) xxx–xxx

Variables of interest were compared among participants from SPCs and PCCs based on unadjusted responses. Categorical variables were tested using a χ2 (or G2 for small expected counts). Continuous variables were tested between the SPC and PCC groups using Student's t-test. Distributions of continuous variables were examined visually using Q-Q plots and by inspection of skewness and kurtosis measures. The data conformed to normality, so parametric modeling was utilized. In identifying the predictive variables of SM of CLBP for the combined data (SPCs and PCCs), a general linear model was used to examine which variables may have an impact on SM using PAM scores as the dependent variable. A saturated model was built first, and insignificant variables and interaction effects were removed. The final model retained significant variables. In evaluating the differences in variables that predict SM between SPCs and PCCs, the procedures conducted for data analysis were similar. 5. Results Tables 1 and 2 show the demographic and pain-related variables in SPCs, PCCs, and the total combined data. Continuous variables, shown as mean (SD), and categorical variables, shown as frequency (%) are presented in the tables. In the combined demographic data (Table 1), majority were females (63.9%), non-Hispanic (84.7%), Caucasians (50.4%), and with college education (53.9%). However, 47.4% of all participants had an income under $15,000 with 43% reported as disabled. For the combined data on pain-related variables (Table 2), 42.6% of the participants felt that their current pain management was helpful (‘a good amount’), and 71.2% rated their overall health from ‘fair’ to ‘good.’ Overall, they had at least 4 medical conditions other than CLBP, with the average duration of their CLBP at 10.7 years, using more than 4 different modalities to manage pain.

3

Although there were significant differences in some variables between the two settings (marital status, race, income, current healthcare insurance, number of pain management modalities used, and perceived SM support), no significant differences were found in terms of SM scores. Further, when setting was added to the model as a covariate, there were no main effects based on setting type or significant interaction effects between setting type. Therefore, all data were combined to evaluate the overall variables that predicted SM of CLBP. 5.1. Aim 1: Identify Variables That Predict SM of CLBP All variables in Tables 1 and 2 were used as potential predictor variables/covariates in the model. The predictors of SM in CLBP are summarized in Table 3 (SPCs and PCCs, N = 230). Five variables were found to be predictive of patient engagement in SM based on the combined data. There were no significant interaction effects in the final model. For the continuous variables, age was negatively associated with SM (β = − 0.197, SE = .074) while SM support was positively associated with SM (β = 2.292, SE = .965). For the categorical variables, adjusted mean scores for educational attainment were significantly lower (p = .013) for those who did not finish high school (M = 53.206, SE = 2.571) than those with college years (M = 60.003, SE = 1.475). Those reporting ‘poor’ overall health (M = 49.491, SE = 2.172) differed significantly from all other groups: ‘fair’ (M = 55.505, SE = 1.751, p = .029), ‘good’ (M = 59.445, SE = 1.906, p b .001), and ‘very good/excellent’ (M = 60.766, SE = 3.261, p = .004). For helpfulness of current pain management, those reporting a ‘little’ helpful (M = 53.585, SE = 2.036) differed from those reporting a ‘good amount’ (M = 58.427, SE = 1.694, p = .042) and ‘greatly’ helpful (M = 60.421, SE = 2.127, p = .015).

Table 1 Demographic Variables in SPCs and PCCs (N = 230). Variable Age Gender Male Female Marital status Married/Live-in partner Separated/Divorced/Widowed Never married Hispanic origin Yes No Race White Black Other Highest grade level bHigh school High school College+ Income b15,000 15,000–34,999 35,000–74,999 75,000+ Current healthcare insurance Yes No Current employment status Working Not working Disabled

SPC and PCC (N = 230)

SPC (n = 110)

PCC (n = 120)

Comparison between SPC and PCC

47.1 (13.61)

46.3 (12.39)

T = .504, p = .615 χ2 = 2.126, p = .145

83 (36.1) 147 (63.9)

45 (40.9) 65 (59.1)

38 (31.7) 82 (68.3)

105 (45.6) 79 (34.3) 46 (20)

59 (53.6) 36 (32.7) 15 (13.6)

46 (38.3) 43 (35.8) 31 (25.8)

35 (15.2) 195 (84.7)

18 (16.4) 92 (83.6)

17 (12.0) 103 (88.0)

116 (50.4) 77 (33.5) 37 (16)

77 (70.0) 18 (16.4) 15 (13.6)

39 (32.8) 59 (49.6) 22 (16.1)

38 (16.5) 68 (29.6) 124 (53.9)

21 (19.1) 25 (22.7) 64 (58.2)

17 (14.4) 43 (36.4) 60 (50.0)

109 (47.4) 64 (27.8) 36 (15.6) 21 (9.1)

41 (37.3) 29 (26.4) 24 (21.8) 16 (14.5)

68 (56.7) 35 (29.2) 12 (10.0) 5 (4.2)

189 (82.2) 41 (17.8)

98 (89.1) 12 (10.9)

91 (75.8) 29 (24.2)

62 (27) 68 (29.6) 99 (43)

26 (23.6) 37 (33.6) 47 (42.7)

36 (30.3) 31 (26.1) 52 (43.7)

χ2 = 7.374, p = .025⁎

χ2 = .906, p = .341

χ2 = 35.236, p b .001⁎

χ2 = 4.889, p = .087

χ2 = 16.609, p = .001⁎

χ2 = 6.886, p = .009⁎

χ2 = 2.044, p = .360

Note. SPCs = specialty pain centers. PCCs = primary care clinics. ⁎ p b .05.

Please cite this article as: Kawi, J., Predictors of self-management for chronic low back pain, Applied Nursing Research (2014), http://dx.doi. org/10.1016/j.apnr.2014.02.003

4

J. Kawi / Applied Nursing Research xxx (2014) xxx–xxx

Table 2 Pain-Related Variables in SPCs and PCCs (N = 230). Variable

SPC and PCC (N = 230)

Helpfulness of current pain management Not much A little A good amount A great deal Support received from other than healthcare provider Not much at all A little A good amount A great deal Importance of religion/spirituality Not much at all A little A good amount A great deal Rating of overall health Poor Fair Good Very good/excellent Length of time with Chronic low back pain Number of pain management modalities used Number of medical conditions PAM (SM scores) PAM level Level 1: Starting to take a role in SM Level 2: Gaining confidence and knowledge Level 3: Taking action Level 4: Staying within course PACIC (SM support) Pain intensity ODI (function/disability) ODI level Minimal/Moderate disability Severe/Extreme disability MHI

SPC (n = 110)

PCC (n = 120)

Comparison between SPC and PCC G2 = 4.143, p = .441

20 58 98 50

(8.7) (25.2) (42.6) (21.7)

8 (7.3) 25 (22.7) 49 (44.5) 28 (25.5)

12 (10.3) 33 (28.4) 49 (42.2) 22 (19.1)

63 70 62 35

(27.3) (30.4) (27) (15.2)

30 (27.3) 37 (33.6) 26 (23.6) 17 (15.5)

33 (27.5) 33 (27.5) 36 (30.0) 18 (15.0)

34 38 67 91

(14.8) (16.5) (29.1) (39.6)

17 (15.5) 21 (19.1) 25 (22.7) 47 (42.7)

17 (14.2) 17 (14.2) 42 (35.0) 44 (36.7)

46 82 82 20 10.7 4.4 4.1 58.4

(20) (35.6) (35.6) (8.7) (10.26) (3.13) (3) (15.46)

19 (17.3) 36 (32.7) 41 (37.3) 14 (12.7) 10.9 (10.58) 5.0 (3.33) 3.8 (3.06) 60.1 (15.05)

27 (22.7) 46 (38.7) 41 (34.5) 6 (5.0) 10.6 (9.97) 3.8 (2.82) 4.4 (2.94) 56.9 (15.75)

54 48 58 70 2.8 2.57 45.3

(23.5) (20.9) (25.2) (30.4) (1.03) (1.25) (17.85)

21 (19.1) 22 (20.0) 30 (27.3) 37 (33.6) 2.6 (.99) 2.55 (1.15) 44.5 (17.50)

33 (27.5) 26 (21.7) 28 (23.3) 33 (27.5) 3.0 (1.02) 2.58 (1.35) 46.0 (18.22)

100 (43.5) 130 (56.5) 55.4 (14.25)

51 (46.4) 59 (53.6) 54.7 (14.31)

49 (40.8) 71 (59.2) 56.1 (14.23)

G2 = 4.251, p = .514

χ2 = 4.407, p = .221

χ2 = 5.386, p = .146

t = .235, p = .814 t = 3.088, p = .002⁎ t = 1.611, p = .109 t = 1.540, p = .125 χ2 = 2.868, p = .412

t = −3.164, p = .002⁎ t = −.152, p = .88 t = −.639, p = .524 χ2 = .714, p = .398

t = −.716, p = .475

Note. SPCs = specialty pain centers. PCCs = primary care clinics. PAM = patient activation measure. SM = self-management. PACIC = patient assessment of chronic illness care. ODI = Oswestry Disability Index. MHI = Mental Health Inventory. ⁎ p b .05.

5.2. Aim 2: Evaluate the Differences in Predictor Variables Between SPCs and PCCs

Table 3 Predictors of Self-Management for Chronic Low Back Pain. Variables

Type III sum of squares

SPCs and PCCs (N = 230) Corrected model 12205.508 Intercept 28091.434 Age 1381.851 PACIC (SM support) 1102.405 Education 1552.159 Overall health 3003.840 HCPM 1642.122 SPCs (N = 110) Corrected model 8231.474 Intercept 27513.281 PACIC (SM support) 1609.488 Support 2381.409 Religion/Spirituality 1939.822 Overall health 2507.570 PCCs (N = 120) Corrected model 4378.294 Intercept 25847.888 PACIC (SM support) 2251.646 Income 2751.772 Overall health 2143.476

df

Mean square

F

p

10 1 1 1 2 3 3

1220.551 28091.434 1381.851 1102.405 776.079 1001.280 547.374

6.245 143.734 7.070 5.641 3.971 5.123 2.801

.000 .000 .008⁎ .018⁎ .020⁎ .002⁎ .041⁎

10 1 1 3 3 3

823.147 27513.281 1609.488 793.803 646.607 835.857

4.956 165.665 9.691 4.780 3.893 5.033

.000 .000 .002⁎ .004⁎ .011⁎ .003⁎

4 1 1 3 3

1094.573 25847.888 2251.646 917.257 714.492

5.010 118.300 10.305 4.198 3.334

.001 .000 .002⁎ .007⁎ .022⁎

Note. SPCs = specialty pain centers. PCCs = primary care clinics. PACIC = patient assessment of chronic illness care. SM = self-management. HCPM = helpfulness of current pain management. Support = support received from other than healthcare provider. ⁎ p b .05.

In SPCs, four variables were found to be predictive of patient engagement in SM (Table 3; SPCs, N = 110). There were no significant interaction effects in the final model. SM support was positively associated with engagement in SM (β = 4.081, SE = 1.311). For support received from other than their healthcare provider, those reporting ‘not much’ support (M = 50.477, SE = 2.658) differed significantly from those reporting: ‘a little’ (M = 60.523, SE = 2.349, p = .004), ‘a good amount’ (M = 60.160, SE = 2.753, p = .011), and ‘a great deal’ (M = 64.385, SE = 3.205, p = .001). For the importance of religion/spirituality, SM scores differed significantly (p = .007) between those reporting ‘a little’ (M = 51.249, SE = 3.140) and those reporting ‘a great deal’ (M = 61.099, SE = 1.941). Overall health scores also differed significantly (p b .01) between those reporting ‘poor’ (M = 52.34, SE = 3.33) and those reporting ‘good’ to ‘excellent’ (M = 64.01, SE = 1.96). In PCCs, three variables were found to be predictive of patient engagement in SM (Table 3: PCCs, N = 120). There were no significant interaction effects in the final model. SM support was positively associated with engagement in SM (β = 4.444, SE = 1.384). For income level, those making b $15,000 (M = 55.196, SE = 1.793) differed significantly from those making $35,000–$74,999 (M = 65.259, SE = 4.392, p = .036) and from those making $75,000+ (M = 75.084, SE = 6.611, p = .004). Lastly, overall health scores differed significantly (p b .05) between those reporting ‘poor’

Please cite this article as: Kawi, J., Predictors of self-management for chronic low back pain, Applied Nursing Research (2014), http://dx.doi. org/10.1016/j.apnr.2014.02.003

J. Kawi / Applied Nursing Research xxx (2014) xxx–xxx

health (M = 50.36, SE = 2.95) and those reporting ‘good’ to ‘excellent’ health (M = 61.21, SE = 2.24). 6. Discussion The demographic characteristics of the participants were mostly consistent with the IOM chronic pain report (2011) and CLBP national health statistics (Schiller, Lucas, Ward, & Peregoy, 2012). Majority were females, Caucasian, and not of Hispanic origin. Many were not working or were disabled, income was predominantly under $15,000, almost half did not attend college, but most were insured. Nationally, those with lower educational attainment were more likely to report CLBP. Further, reports showed that those with Medicaid and Medicare were more likely to have CLBP (Schiller et al., 2012). Comparing the participants' demographic data in both settings, significant differences were noted. The PCCs had considerably more single participants, more African-Americans than other races, with majority of the participants making less than $15,000 but having healthcare insurance. The locations of the PCCs were mostly within the inner city. For pain-related variables, the PCC participants also used significantly fewer pain management modalities and perceived more healthcare provider support (SM support) than those in the SPCs. The average age of the participants from the SPCs was 47.1 years old and 46.3 years old from the PCCs. The average durations of CLBP for the participants were 10.6 (PCCs) and 10.9 (SPCs) years. They used 5 (significantly higher in SPCs) and 3.8 (PCCs) modalities to manage pain, and had 3.8 (SPCs) and 4.4 (PCCs) chronic medical conditions other than CLBP. Consequently, 30% are not working, and 43% are disabled. Chronic low back pain is the leading cause of number of years lost due to disability creating an enormous disease burden (Murray & Lopez, 2013). In this study, average disability scores were 44.5 (SPCs) and 46 (PCCs) out of a high of 100. A score of 43.3 was found to be the mean disability score for chronic back pain patients from a systematic literature review of spine patients (Fairbank & Pynsent, 2000). Based on the Oswestry Disability Index, these values fall within the severe disability range of 41–60 (Fairbank & Pynsent, 2000). Americans are more disabled now compared to 10 years ago (Murray & Lopez, 2013). The average SM scores were 60.1 (SPCs) and 56.9 (PCCs) out of a maximum of 100. A score of 59.1 was the mean SM score found for chronic pain patients (Hibbard et al., 2005). When average SM scores were converted to an SM level, these scores fell under level 3; participants were taking action in SM although needing to consistently maintain positive health-directed behaviors (Hibbard et al., 2004). The average healthcare provider support or SM support scores were 2.6 (SPCs) and 3 (significantly higher in PCCs) out of 5. The average perceived SM support score in chronic pain patients is 2.64 (Glasgow et al., 2005). Average pain intensity were 2.55 (SPCs) and 2.58 (PCCs) out of 5. These values are consistent in both settings. Lastly, mental health state scores were 54.7 (SPCs) and 56.1 (PCCs) out of 100. The optimal cut-off point for psychological well-being is thought to be at 60; a score below this point indicates psychological distress (Kelly, Dunstan, Lloyd, & Fone, 2008). In optimal management of chronic pain patients, addressing mental health was found to be a necessary component in SM programs toward improved outcomes (Kroenke et al., 2009). 6.1. Aim 1: Identify Variables That Predict SM of CLBP In the combined data (SPCs and PCCs), the overall significant predictors of SM were age, perceived SM support, education, overall health, and reported helpfulness of current pain management. Overall, younger participants scored higher in SM. However, generalizing this finding is a challenge because there was some spread of data in the scatter plot when age was plotted against SM scores. A previous study

5

did report that almost 43% of the younger age group was not receiving any treatment for their chronic pain (Rustoen et al., 2005). Hence, they are likely managing their own pain symptoms. It would be interesting to conduct a longitudinal study to evaluate how SM scores change the longer patients have CLBP. Those receiving more healthcare provider support scored higher in SM. This is consistent with studies on CLBP where support provided by healthcare providers was considered essential in engaging patients in SM toward health-directed behaviors (Jordan, Holden, Mason, & Foster, 2010; Liddle et al., 2007; MacKichan, Paterson, & Britten, 2013). It was also noted that those with a higher education, better overall health, and those who reported that their current pain management was helpful (‘good’ to ‘a great deal’) were better able to self-manage their CLBP. This begs to question which strategy for pain management is most appropriate for those with lower educational attainment, poor overall health, and with persistent unrelenting CLBP. Interestingly, variables like pain intensity, disability, and mental health state were not significant predictors of SM. Since the average duration of low back pain for the participants was at least 10 years and they were using 3 to 5 pain management modalities, they may have learned to cope with their chronic pain through time. The question is whether these adaptation or coping mechanisms were positive and health-directed. It is also important to acknowledge that there may be other variables not included in this research that can potentially predict SM (i.e., coping, adaptation, fear avoidance, pain catastrophizing, and quality of life). 6.2. Aim 2: Evaluate the Differences in Predictor Variables Between SPCs and PCCs In SPCs, SM support, support received from other than healthcare providers, religion or spirituality, and overall health were identified as significant predictors of SM. Those who reported more support from their providers and significant others tend to have higher SM scores. Those who felt that religion or spirituality was of ‘little’ importance had significantly lower SM scores than those who regarded religion/ spirituality of ‘great’ importance. Collectively, it appears that participants in SPCs count on external support for their CLBP (SM support from healthcare providers, significant others, religious or spiritual being). This may not be unusual since patients in SPCs are typically more complex and harder to manage (IOM, 2011). In PCCs, both SM support and overall health were also significant predictors of SM. In addition, those with higher incomes scored better in SM. There was a greater disparity in income level in the PCCs with a majority (57%) earning less than $15,000. This may be related to the location of the PCCs in the inner city where socio-economic status is lower. 7. Implications for Practice The results of this research are relevant to all stakeholders in the chronic pain population, including CLBP patients, their significant others, healthcare providers, healthcare organizations, insurers, and policy makers. This study provides a stronger foundation for applying SM in various practice settings, especially for best responders in the CLBP population. Several implications for nursing practice are highlighted based on the study findings. Nurses are in the forefront of chronic pain care. An adequate nursing assessment and evaluation of the patient's willingness to engage in SM is important. This includes identifying barriers and facilitators to their SM. Accordingly, prior levels of engagement in SM play a substantial role in future SM participation (Caiata & Schulz, 2009). It is necessary to allow open discussions on the role of SM in CLBP during clinic visits. With this knowledge, nurses can take a leading role in providing support to facilitate SM, especially for CLBP

Please cite this article as: Kawi, J., Predictors of self-management for chronic low back pain, Applied Nursing Research (2014), http://dx.doi. org/10.1016/j.apnr.2014.02.003

6

J. Kawi / Applied Nursing Research xxx (2014) xxx–xxx

patients in outpatient settings. Individuals with greater SM scores are more likely to consistently maintain health-directed behaviors toward improved pain outcomes (Hibbard et al., 2004). To facilitate SM, nurses and other healthcare professionals need to be educated and well-trained in SM support, a significant predictor of SM, in order to engage patients in adequately managing their health. Although SM is the patient's responsibility, providing support to enhance SM and enabling patient–provider partnership is important. Strategies in SM programs typically include goalsetting, action-planning, problem-solving, decision-making, and appropriate communication skills. These strategies encourage active patient participation in their treatment plans, promote healthdirected behaviors, improve coping, and minimize acute episodes of pain. Addressing psychosocial variables in patients with CLBP is also essential. Enhancing mental health is necessary in the context of improving overall health. Participants were below the cut-off point for psychological well-being in this research. Depression, anxiety, fear, and catastrophizing are common in patients with CLBP. Interprofessional collaboration with behavioral therapists may be required. Further, SM programs that are attuned to relevant cultural issues and literacy levels may result in better adherence to SM strategies. Lastly, nurses' role as advocates for changes in healthcare system policy cannot be understated. Facilitating SM strategies take time beyond a 15-minute clinic visit. Reimbursement for SM programs and adequate SM resources in the community are vital. These resources need to include vocational rehabilitation. Identification of more appropriate pain management strategies for those who do not respond well to SM programs due to poor overall health, low level of education, or low income are vital. These approaches can minimize reinforcing pain behavior, secondary gain, and pain care disparities. This study is limited by its use of self-report data analyzed from convenience samples in the original primary research studies. Findings cannot be generalized to other pain population. Further, longitudinal and experimental studies would be valuable in evaluating the impact of CLBP SM programs on healthcare costs as well as on the use, misuse, and abuse of opioids. The IOM (2011) emphasized that dependence on external sources could be minimized by effective SM. These sources include opioid use. 8. Conclusion There is uncertain evidence on the effectiveness of SM on CLBP outcomes. Hence, this secondary analysis evaluated variables that best predict SM of CLBP. Those who are younger, perceiving adequate SM support, with higher education, better overall health, and receiving sufficient pain management tend to have higher SM scores. These findings provide valuable information to healthcare providers and healthcare organizations to intervene more appropriately in engaging SM for individuals with CLBP. Alternatively, other strategies need to be identified and evaluated for those who are not the best candidates for SM programs. Acknowledgment This research study was supported by a grant from the Nurse Practitioner Healthcare Foundation (NPHF)/Purdue Pharma L. P. Pain Management Award program. References Arnstein, P., & Marie, B. S. (2010). Managing chronic pain with opioids: A call for change. : Nurse Practitioner Healthcare Foundation (Retrieved from http://www. nphealthcarefoundation.org/programs/downloads/white_paper_opioids.pdf).

Berwick, D. M., Murphy, J. M., Goldman, P. A., Ware, J. E., Barsky, A. J., & Weinstein, M. C. (1991). Performance of a five-item mental health screening test. Medical Care, 29, 169–176. Caiata, Z. M., & Schulz, P. J. (2009). Self-management of chronic low back pain: An exploration of the impact of a patient-centered website. Patient Education and Counseling, 77, 27–32, http://dx.doi.org/10.1016/j.pec.2009.01.016. Centers for Disease Control, Prevention (2011). Vital signs: Overdose of prescription opioid pain relievers-United States, 1999–2008. MMWR weekly. (Retrieved from http://www.cdc.gov/mmwr/preview/mmwrhtml/mm6043a4.htm? s_cid=mm6043a4_w). Coudeyre, E., Givron, P., Vanbiervliet, W., Benaim, C., Herisson, C., Pelissier, J., et al. (2006). The role of an information booklet or oral information about back pain in reducing disability and fear-avoidance beliefs among patients with subacute and chronic low back pain. A randomized controlled trial in a rehabilitation unit. Annales de Réadaptation et de Médecine Physique, 49, 600–608, http://dx.doi.org/10. 1016/j.annrmp.2006.05.003. Cooper, K., Smith, B. H., & Hancock, E. (2009). Patients' perceptions of self-management of chronic low back pain: Evidence for enhancing patient education and support. Physiotherapy, 95, 43–50. Crowe, M., Whitehead, L., Gagan, M. J., Baxter, D., & Panckhurst, A. (2010). Selfmanagement and chronic low back pain: A qualitative study. Journal of Advanced Nursing, 66, 1478–1486. Fairbank, J. C., & Pynsent, P. B. (2000). The Oswestry Disability Index. Spine, 25, 2940–2953. Freburger, J. K., Holmes, G. M., Agans, R. P., Jackman, A. M., Darter, J. D., Wallace, A. S., et al. (2009). The rising prevalence of chronic low back pain. Archives of Internal Medicine, 169, 251–258, http://dx.doi.org/10.1001/archinternmed.2008.543. Glasgow, R. E., Wagner, E. H., Schaefer, J., Mahoney, L. D., Reid, R. J., & Greene, S. M. (2005). Development and validation of the patient assessment of chronic illness care. Medical Care, 43, 436–444. Hibbard, J. H., Mahoney, E. R., Stockard, J., & Tusler, M. (2005). Development and testing of a short form of the Patient Activation Measure. Health Services Research, 40, 1918–1930, http://dx.doi.org/10.1111/j.1475-6773.2005.00438. Hibbard, J. H., Stockard, J., Mahoney, E. R., & Tusler, M. (2004). Development and patient activation measure (PAM): Conceptualizing and measuring activation in patients and consumers. Health Services Research, 39, 1005–1026, http://dx.doi.org/10.1111/ j.1475-6773.2004.00269.x. Institute for Clinical Systems Improvement (2011). Assessment and management of chronic pain. Agency for Healthcare Research and Quality. (Retrieved fromhttp:// guidelines.gov/content.aspx?id=36064). Institute of Medicine (2011). Relieving pain in America: A blueprint for transforming prevention, care, education, and research. (Retrieved from http://www.iom.edu/ Reports/2011/Relieving-Pain-in-America-A-Blueprint-for-TransformingPrevention-Care-Education-Research.aspx). Jordan, J. L., Holden, M. A., Mason, E. E., & Foster, N. E. (2010). Interventions to improve adherence to exercise for chronic musculoskeletal pain in adults. Cochrane Database of Systematic Reviews, http://dx.doi.org/10.1002/14651858.CD005956. Kawi, J. (2012). Self-management and self-management support on functional ablement in chronic low back pain. Pain Management Nursing, http://dx.doi.org/ 10.1016/j.pmn.2012.05.001 (Advance online publication). Kawi, J. (2014). Self-management and self-management support on chronic low back pain patients in primary care. Journal of the American Academy of Nurse Practitioners (in press). Kelly, M. J., Dunstan, F. D., Lloyd, K., & Fone, D. L. (2008). Evaluating cutpoints for the MHI- 5 and MCS using the GHQ-12: A comparison of five different methods. BMC Psychiatry, 8, http://dx.doi.org/10.1186/1471-244X-8-10. Kroenke, K., Bair, M. J., Damush, T. M., Wu, J., Hoke, S., Sutherland, J., et al. (2009). Optimized antidepressant therapy and pain self-management in primary care patients with depression and musculoskeletal pain: A randomized controlled trial. Journal of the American Medical Association, 301, 2099–2110. Liddle, S. D., Baxter, G. D., & Gracey, J. H. (2007). Chronic low back pain: Patients' experiences, opinions and expectations for clinical management. Disability & Rehabilitation, 29, 1899–1909. Lorig, K. R., & Holman, H. R. (2003). Self-management education: History, definition, outcomes, and mechanisms. Annals of Behavioral Medicine, 26, 1–7. May, S. (2010). Self-management of chronic low back pain and osteoarthritis. Nature Reviews. Rheumatology, 6, 199–209, http://dx.doi.org/10.1038/nrrheum.2010.26. MacKichan, F., Paterson, C., & Britten, N. (2013). GP support for self-care: The views of people experiencing long-term back pain. Family Practice, 30, 212–218, http://dx. doi.org/10.1093/fampra/cms062. Murray, C. J. L., & Lopez, A. D. (2013). Measuring the global burden of disease. New England Journal of Medicine, 369, 448–457. National Center for Health Statistics (2012). Health, United States, 2011: With special feature on socioeconomic status and health. Centers for Disease Control and Prevention. (Retrieved from http://www.cdc.gov/nchs/data/hus/hus11.pdf). National Council on Aging (2012). Chronic disease self-management. (Retrieved from http://www.ncoa.org/assets/files/pdf/NCOA-Chronic-Desease.pdf). National Institute of Arthritis and Musculoskeletal and Skin Diseases (2009). Handout on health: Back pain (NIH Publication No. 09–5282). (Retrieved from http://www. niams.nih.gov/Health_Info/Back_Pain/default.asp). Oliveira, V. C., Ferreira, P. H., Maher, C. G., Pinto, R. Z., Refshauge, K. M., & Ferreira, M. L. (2012). Effectiveness of self-management of low back pain: Systematic review with meta-analysis. Arthritis Care & Research, 64, 1739–1748. Rustoen, T., Wahl, A. K., Hanestad, B. R., Lerdal, A., Paul, S., & Miaskowski, C. (2005). Age and the experience of chronic pain: Differences in health and quality of life

Please cite this article as: Kawi, J., Predictors of self-management for chronic low back pain, Applied Nursing Research (2014), http://dx.doi. org/10.1016/j.apnr.2014.02.003

J. Kawi / Applied Nursing Research xxx (2014) xxx–xxx among younger, middle-aged, and older adults. The Clinical Journal of Pain, 21, 513–523. Schiller, J. S., Lucas, J. W., Ward, B. W., & Peregoy, J. A. (2012). Summary health statistics for U.S. adults: National Health Interview Survey, 2010. National Center for Health Statistics.Vital and Health Statistics, Series, 10(252). (pp. 1–207), 1–207. Schulz, P. J., Rubinell, S., & Hartung, U. (2007). An internet-based approach to enhance self-management of chronic low back pain in the Italian-speaking population of Switzerland: Results from a pilot study. International Journal of Public Health, 52, 286–294, http://dx.doi.org/10.1007/s00038-007-5127-9.

7

Sokunbi, O., Cross, V., Watt, P., & Moore, A. (2010). Experiences of individuals with chronic low back pain during and after their participation in a spinal stabilisation exercise programme - A pilot qualitative study. Manual Therapy, 15, 179–184, http://dx.doi.org/10.1016/j.math.2009.10.006. Toward Optimized Practice (2009). Guideline for the evidence-informed primary care management of low back pain. Agency for Healthcare Research and Quality (Retrieved from http://guideline.gov/content.aspx?id=15668). Veit, C. T., & Ware, J. E., Jr. (1983). The structure of psychological distress and wellbeing in general populations. Journal of Consulting and Clinical Psychology, 51, 730–742.

Please cite this article as: Kawi, J., Predictors of self-management for chronic low back pain, Applied Nursing Research (2014), http://dx.doi. org/10.1016/j.apnr.2014.02.003

Predictors of self-management for chronic low back pain.

(a) Identify variables that predict self-management (SM) of chronic low back pain (CLBP), and (b) evaluate differences in these variables between part...
265KB Sizes 0 Downloads 3 Views