British Journal of General Practice Patient Enablement Moderates the Effect of Pain on Health-related Quality of Life of Primary Care Patients with Chronic Back and Knee Problems-a cross-sectional study

Background: Chronic back and knee pain impairs health-related quality of life (HRQOL) and patient enablement can improve HRQOL. Aim: To determine whether enablement was a moderator of the effect of chronic back and knee pain on HRQOL. Design and Setting: A cross-sectional study on Chinese patients with chronic back and knee problems in public primary care clinics in Hong Kong. Method: Each participant completed the Chinese Patient Enablement Instrument-2 (PEI-2), the Chinese Western Ontario and McMaster University Osteoarthritis Index (WOMAC), and the Pain Rating Scale (PRS). Multivariable regression examined the effects of PRS score and PEI-2 score on WOMAC total score. Moderation regression model and simple slope analysis were used to evaluate whether the interaction between enablement (PEI-2) and pain (PRS) had a significant effect on HRQOL (WOMAC). Results: Valid PRO data from 1306 participants were analysed. PRS score was associated with WOMAC total score (β = 0.326, p<0.001), while PEI-2 score was associated inversely with WOMAC total score (β = -0.260, p<0.001) and PRS -2 score. The effect of the interaction between PRS and PEI-2 (PRS*PEI-2) scores on WOMAC total score was significant (β = - 0.191, p<0.001) suggesting PEI-2 was a moderator. Simple slope analyses showed the relationship between PRS and WOMAC was stronger for participants with a low level of PEI-2 (gradient=3.056) than for those with a high level of PEI-2 (gradient =1.746). Conclusion: Patient


INTRODUCTION
Musculoskeletal problems not only diminish functioning, increase distress, and worsen self-perceived health for individuals, but also constitute a burden on the healthcare system (1)(2)(3)(4).They account for 7% of primary care consultations in Hong Kong, with back and knee problems being the most prevalent (5).
Health-related quality of life (HRQOL) is a patient-reported outcome (PRO) that assesses a person's subjective judgment on how their health impacts their life (6).It can inform decisions on service needs and outcomes (7).Chronic musculoskeletal pain impairs HRQOL (8,9), disturbs sleep, and induces psychological distress (10,11).Patient enablement is defined as "the extent to which a patient is capable of understanding and coping with his or her issues" (12) and is an indicator of the quality of consultations (13)(14)(15).It starts with a better understanding of patients' needs and expectations and encompasses shared decision-making with acknowledgment of patients' strengths (16)(17)(18) by providing them with the ability to look after their illness (17,19).Patients with better enablement are more likely to have better chronic disease outcomes (20).Furthermore, patient enablement is modifiable and is a goal of patient-centred care (20).Tracking PROs, like patient enablement, can help to assess the effectiveness of patient-centred care programs (21).There are two major components of patient enablement -health literacy (22) and coping (23).Health literacy is the "ability of an individual to obtain and translate knowledge and information in order to maintain and improve health in a way that is appropriate to the individual and system contexts" (24,25).Coping is defined as "effortful behaviour undertaken in reaction to a stressor" (26), which moderates the association between pain and HRQOL (27).Thus, more enabled individuals may maintain better HROQL than those who are less enabled for the same level of pain (28).
There are other concepts intertwined with patient enablement.Illness perception focuses on how a patient experiences and mentally frames their life with illness (29), which affects coping strategies and self-management (i.e.patient enablement) (30) and HRQOL (31).The concepts of enablement and empowerment are sometimes used synonymously (32); however, patient enablement focuses on obtaining skills and knowledge and patient empowerment involves gaining of power after obtaining those skills and knowledge, and therefore, patient enablement must occur before patient empowerment (33).
This study aimed to determine whether enablement was a moderator of the effect of chronic back and knee pain on HRQOL.We hypothesized that pain and enablement would affect HRQOL, and that the effect of pain on HRQOL could be moderated by enablement.
The conceptual model is shown in Supplementary Figure 1.

Study setting
The study was carried out in Hong Kong where the health care system is serviced by both the public and private sectors and residents can freely choose where they receive care without means testing.Due to the heavy financial subsidies, the public primary care clinics, called General Outpatient Clinics (GOPC), are the main care providers for patients with chronic diseases (34,35).There are 73 GOPCs organized in seven clusters around Hong Kong.The care in GOPCs is standardized by evidence-based guidelines while care is more variable in the private sector.Participants of this study were recruited from six GOPCs in two clusters.

Study design and participants recruitment:
This cross-sectional study was part of a single-blinded cluster randomized controlled trial to evaluate the effectiveness of measuring and reporting HRQOL in routine clinical practice (36).The subject inclusion criteria were adults aged 18 years or above, who had a doctor-diagnosed symptomatic back and/or knee problem that was expected to last for one month or more.The exclusion criteria were patients who had a life expectancy <12 months, had current cancers, were too ill, or were unable to communicate in Chinese.All eligible patients were recruited by their doctors or trained research assistants in the clinics between June 1, 2020 and December 31, 2021.After completion of a written consent form, we collected participants' sociodemographic and clinical data by structured questionnaires and PROs by a telephone survey administered by an independent survey organization.The participant flow chart is shown in Supplementary Figure 2.More details on subject recruitment and data collection are described in a previous paper (36).The study was approved by the Institutional Review Boards (IRB) of the HKU/Hospital Authority Hong Kong West Cluster (reference number: UW 18-270) and the Hospital Authority Kowloon Central/ Kowloon East (reference number: KC/KE-20-0070/ER-1),

Sample size
We included 1221 participants with complete PRO data in this moderation study.Post-hoc power analysis, using G*Power version 3.1.9.7 (37), showed a power of 0.95 for moderation analysis at a significance criteria of α = 0.05, F 2 = 0.011.

Structured questionnaires
Participants' socio-demographics, lifestyle factors, and self-reported comorbidities were collected by a questionnaire administered by a trained research assistant in the clinic.
Another questionnaire was completed by the physician to collect data on the diagnosis and duration of the musculoskeletal problem and to give a global rating scale (GRS) on the severity of the participant's condition on a five-point Likert scale from 1 (no problem) to 5 (very severe).

The Chinese Patient Enablement Instrument-2 (PEI-2)
The Chinese PEI-2 is a PRO measure that was adapted from the PEI originally developed by Howie et al. (15) and consists of six items: (1) ability to cope with life, (2) ability to understand one's illness, (3) ability to cope with one's illness, (4) ability to keep oneself healthy, (5) confidence about one's health, and (6) ability to help oneself.Each item is rated on a five-point Likert scale, ranging from one (not at all) to five (extremely well).The item scores are summated to give a total PEI-2 score (range: 6-30), with higher scores indicating better enablement.Missing item values can be imputed with the average of the scores of answered items, up to three items.The Chinese PEI-2 is valid, reliable, and sensitive in a Chinese population (14).

The Chinese Western Ontario and McMaster University Osteoarthritis Index (WOMAC)
The WOMAC is a widely used HRQOL measure specific to musculoskeletal problems, which has been applied to patients with knee (38) and back (39) problems.It consists of 24 items in three domains: pain (5 items), stiffness (2 items), and physical function (17 items).

Page 6 of 16
Each item is rated on a 5-point Likert scale (0 to 4) with higher scores indicating more symptoms or greater impairment.The item scores in each domain are summated as the domain score.The total WOMAC score is the sum of the three domain scores, ranging from 0 to 96 (40).For participants with completion of at least four of the five pain items, one of the two stiffness items, and 14 of the 17 function items, missing item values can be imputed with the average of the completed items in the relevant subscale (41).The Chinese version of WOMAC is valid, reliable, and sensitive in Chinese patients (42).

Pain Rating Scale (PRS)
The PRS is a self-reported tool consisting of a scale with extreme anchors of "no pain (0)" to "extreme pain ( 10)" (43) to assess the severity of pain.The validity and reliability of the Chinese version is established (44).

Statistical analysis
Descriptive statistics were used to describe the patients' socio-demographics, lifestyle, disease characteristics, and PROs.One-way analysis of variance (ANOVA) was used to test the mean differences in the PROs between the groups of patients diagnosed with a back problem only, knee problem only, and both back and knee problems.Post-hoc test (by Tukey's honestly significant difference) was used to compare each group against the others.Pearson's correlation between different PROs was examined.Multivariable regression was used to test the effect of disease characteristics, PRS score, and PEI-2 score on WOMAC total score, adjusted by socio-demographic characteristics.
We used moderation analysis to test if enablement (PEI-2 score as the moderator) would interact with pain level (PRS score as the predictor) to influence HRQOL (WOMAC total score as the outcome) as shown in Figure 1.Moderation analysis hypothesizes that a predictor and the interaction between a predictor and a moderator can predict the outcome if there is a "moderation effect" (45).The PRS and PEI-2 scores were standardized (converted to a Z score) to avoid multicollinearity in the moderation analysis.Other variables that had significant effects on WOMAC total score were included as confounders in the moderation regression model.The moderation analysis was carried out by Hayes's PROCESS macro (45,46) for SPSS Model 1.In addition, a simple slope analysis was used to show the moderation effect (47).The mean of the estimated WOMAC total score ( ) was calculated according to the  following equation: A 5% statistical significance was used for all analyses.All the analyses were carried out by IBM SPSS Statistics ver.27 (46).

Socio-demographic, lifestyle, and disease characteristics
Table 1 shows the participants' socio-demographic, lifestyle, and disease characteristics.
Of the 1319 participants, 69.1% were female and the mean age was 68.80 (SD 10.16) years old.
The physician diagnoses were back problem only (22.4%), knee problem only (67.6%), and both back and knee problems (10.0%).The study population was similar in age and gender distribution to the back and knee problem patient population (mean age 67.28 and 67.5% female) presenting to primary care clinics found in a territory-wide morbidity survey in 2021/2022 in Hong Kong (48).The mean and median GRS score was 2.45 (SD 0.64) and 2, respectively.

Patient-reported outcomes
Valid WOMAC scores (the primary outcome) were available from 1306 participants whose characteristics were similar to those with incomplete data (n=13) (Supplementary Table 1).Valid PEI-2 scores and PRS scores were available for 1258 and 1307 participants, respectively.Table 2 shows the overall mean WOMAC total score (20.93,S.D. and PRS score among diagnostic groups.Post-hoc analysis (Supplementary Table 2) found that participants with both back and knee problems had significantly higher WOMAC scores than participants with back or knee problems only.Participants with back problems only had significantly higher PRS scores than those with knee problems only.

Moderation analysis
The result of the moderation analysis is shown in Table 5.The effect of the interaction between PRS and PEI-2 (PRS*PEI-2) scores on the WOMAC total score was significant (β= -0.191, [-0.271, -0.111]), suggesting that patient enablement weakened the effect of pain measured by PRS on HRQOL measured by WOMAC.A sensitivity analysis was carried out by clusters (three clinics in each), which showed similar results (Supplementary Table 3).
Simple slope analysis (Figure 1 Since the WOMAC total score includes the pain domain that may bias the association with PRS, we carried out a sensitivity analysis on the moderation effect using the WOMAC stiffness domain and physical function domain summation score (excluding the pain domain score) as the outcome.It showed similar results (Supplementary Table 4 and Supplementary Figure 3), further substantiating that PEI-2 was a moderator between PRS and WOMAC total scores.

Summary
This study examined the moderation of patient enablement on the negative impact of pain on HRQOL in the context of chronic back and knee problems.The results confirmed that patient self-reported pain and enablement were the most significant determinants of HRQOL.
As we had hypothesized, a higher level of pain was associated with more impairment, and a higher level of enablement was associated with less impairment of HRQOL.We have established that enablement was a significant moderator of the association between pain and Page 10 of 16 HRQOL.Specifically, a high level of enablement weakens the effect of pain on HRQOL impairment.

Strengths and limitations
The strength of the study was the large number of participants recruited from multiple primary care clinics.The age and gender distribution of the participants was similar to those of the back and knee problem patient population found in the Hong Kong territory-wide primary care morbidity survey in 2021/2022.We believe the results may be generalizable to the general Hong Kong population.We measured the PRO data with standard valid measures and established patient enablement as a moderator of the association between pain and HRQOL.
The sensitivity analyses showed consistent results, supporting reliability.There were a few limitations to this study.Firstly, the analysis of cross-sectional data could not ascertain a causal relationship, and the correlations between PEI and WOMAC (-0.39) and PRS (-0.20) scores were relatively small.Secondly, the results from patients of public primary care clinics may not be applicable to those managed in the private sector.Thirdly, self-report data are subject to bias from recall and a tendency to give socially acceptable answers.We believe these biases applied to all subjects and should not have affected the moderation analysis results.Lastly, the results from Chinese patients may not be generalizable to non-Chinese populations.

Comparison with existing literature
Our study confirmed the findings from previous studies on the negative association between pain severity and HRQOL (8-11, 27, 49-51) and the positive association between enablement and HRQOL (52).We found a significant, although weak, negative association between enablement and pain, which moderated the association between pain and HRQOL.
Theoretically, enablement promotes the patient's ability to control their health, life, and coping Page 11 of 16 strategies, which can lessen pain, depression, and limitation in daily functioning.Therefore, it can improve HRQOL (26,27,53).Another mechanism for which enablement helps is through improved health literacy (25).Studies in Australia have shown that patients with chronic diseases who were more enabled were more likely to have a higher level of health literacy and to actively manage their health (20).In addition, health literacy can influence an individual's illness perception resulting in better coping ability (54).

Implications for practice
The finding on the role of enablement as a moderator between pain and HRQOL has implications for a paradigm change in clinical practice.Most chronic musculoskeletal problems are progressive degenerative conditions with no cure.Primary care physicians should recognize patient enablement as a first-line intervention for chronic musculoskeletal problems.
The scope of patient-centred care should be broadened to include an assessment of HQOOL and enablement so that management can be customized to enhance coping, health literacy, and self-care to maximize functioning.A reorientation of the treatment goal may help reduce the use of analgesics that can cause serious side effects and are subject to abuse (55,56).Education on the nature of the musculoskeletal problem and self-care strategies should be part of routine care for these patients (57).Exercise and self-care management programmes can reduce pain and improve function in patients with musculoskeletal problems (58,59).Appropriate referrals to other healthcare professionals can further enable patients to overcome specific environmental limitations.For instance, occupational therapists can advise on home modification so that patients can continue to live comfortably at home.

Conclusion
Patient enablement moderated the negative impact of pain on HRQOL.Higher level of enablement can lessen the impairment of HRQOL related to chronic back and knee pain.
Education, exercise and self-care programmes that aim to enhance patients' health literacy, functioning, and ability to cope with their illness should become routine treatments for chronic musculoskeletal problems.Future studies on more diverse populations are required to establish the interaction between enablement and pain on HRQOL and how patient enablement can be effectively enhanced in patients with chronic musculoskeletal problems.
14.77), PEI-2 score (21.58,S.D. 3.49), and PRS score (5.44, S.D. 2.32), and the distribution of these PRO scores by musculoskeletal diagnosis groups.One-way ANOVA analysis showed significant differences in WOMAC total, pain domain and physical function domain scores, PEI-2 score, ) shows a different gradient effect of PRS score on WOMAC total score for participants with high (PEI-2 score=25.05)and low (PEI-2 score=18.14)levels of enablement.The relationship between WOMAC and PRS score was stronger for participants with a low level of PEI-2 (gradient=3.056[2.626, 3.486] than for those with a high level of PEI-2 (gradient=1.746[1.367, 2.126]).The difference in the gradient of the WOMAC total score plot from the low level of pain (PRS score =3.17) to a high level of pain (PRS=7.77)was greater (difference=1.31) at the low PEI-2 than the high PEI-2 level.
Ontario and McMaster Universities Osteoarthritis Index (higher score indicates more limitation); SD = standard deviation; PEI-2= Patient Enablement Instrument (higher score indicates more enabled); PRS= Pain Rating Scale (higher score indicates more pain); GRS = Global Rating Scale on disease severity.Notes.† Multi-variable linear regression was used for the analysis.‡ By multi-variable linear regression adjusted for socio-demographic and lifestyle characteristics variables (including gender, age, education, marital status, occupation, household monthly income, drinking habit, and smoking habit); §Reference category; all β-Coefficients are statically significant

WOMAC=
The Western Ontario and McMaster Universities Osteoarthritis Index (higher score indicates more limitation); SD = standard deviation; PEI-2= Patient Enablement Instrument (higher score indicates more enabled); PRS= Pain Rating Scale (higher score indicates more pain); SE = standard error; CI = confidence interval; t = t value Notes Only participants with valid data in all three patient-reported outcomes were included in the moderation analysis (N=1221).† Moderated regression analysis adjusted by gender, total number of comorbidities, physician global rating scale score on severity and drinking status.