Osteoarthritis management: Does the pharmacist play a role in bridging the gap between what patients actually know and what they ought to know? Insights from a national online survey

Abstract Background Guidelines encourage engagement in self‐care activities for osteoarthritis (OA), but there are gaps in consumers' knowledge about suitable choices for self‐care. Community pharmacists are in an ideal position to contribute to OA management through screening and supporting evidence‐based pain management choices. Prior research established an association between health literacy and advice‐seeking and appropriateness of analgesics choices (both lower in participants with limited health literacy) amongst people living with OA. This article explores the implications of these data for pharmacists in OA management. Methods A national online survey was conducted amongst 628 adults aged 45–74 years, currently residing in Australia, with self‐reported symptoms of OA. All data were collected using a customized online questionnaire, which was completed only once. ‘Self‐reported symptoms of OA’ was based on six validated screening questions to identify people with OA without a formal clinical diagnosis. Results Respondents matched the typical profile of people diagnosed with OA; more than half were female (56%), knees (59%) and hips (31%) were the primary affected joints and 74% were either overweight or obese. Self‐identification of OA was limited (41%). Overall, 38% self‐managed their pain, and limited health literacy was associated with less advice‐seeking. Efficacy and ease of use were the main reasons cited for prompting use across all classes of nonprescription analgesic, with less than 20% reporting recommendation from a pharmacist. Participants were managing their pain with an average of 1.74 (95% confidence interval: 1.60–1.88) analgesics, but 73% reported inadequate pain relief and 54% had disrupted sleep. Conclusion Our findings highlight three key themes: lack of self‐identification of OA, suboptimal pain relief and limited use of the community pharmacist as a source of management advice. Equipping community pharmacists with tools to identify OA could bridge this gap. More research is needed to determine if it will improve consumers' ability to appropriately manage OA pain. Patient or Public Contribution Consumers living with OA contributed to the study outcomes, reviewed the survey questionnaire for face validity and advised on plain language terminology.


| INTRODUCTION
Osteoarthritis (OA), a complex, chronic health problem, is associated with significant burden. 1 Australian National Health Survey data collected in 2017-2018 estimated that one-fifth of Australians over the age of 45 years had OA. 2 Global data predict OA to become one of the most prevalent diseases in populations from high-income countries, 1 with the suggestion that by 2032, almost 30% of the population aged 45 years are expected to have doctor-diagnosed OA. 3 Current OA management focuses on symptomatic relief and reducing disability progression, via a combination of lifestyle, nonpharmacological, pharmacological and surgical approaches tailored to the needs of the individual. [4][5][6] To further facilitate lifestyle and nonpharmacological approaches, established guidelines encourage engagement in self-care activities. 4,7 Proposed strategies to mitigate the future burden of OA suggest that there is a need for 'better awareness, especially of the risk factors, and early diagnosis and treatment of OA together with the improvement of healthcare infrastructure for managing the increasing number of patients with OA'. 1 Health literacy has been highlighted as an important component of patient engagement self-management strategies. [8][9][10][11] There are negative relationships between health literacy and diagnosis comprehension, 12,13 health outcomes 14 and use of healthcare services, 15 and limited health literacy is identified as an obstacle to accessing primary care services. 16 Adequate health literacy has previously been shown to be integral to patients' decisions to adopt and engage with OA selfmanagement strategies. 11 However, data exploring this issue from the perspective of the patient are limited. 11,17 A recent survey, conducted amongst adults with self-reported symptoms of OA, investigated the participants' perceptions of pain and their use of self-management strategies for pain relief. 18 The study reported a significant association between health literacy and the appropriate selection of recommended self-management strategies and highlighted participants suboptimal knowledge about suitable pain relief choices.
Recognizing that the pharmacist is well placed to provide counselling to help improve patients' knowledge about and correct use of medicines, 19,20 this paper explores this study, focusing on the findings of relevance to the community pharmacist.

| METHODS
The methods have been published in detail elsewhere. 18 Briefly, a national online survey was conducted amongst adults aged  years, currently residing in Australia, with self-reported symptoms of OA. The cohort was derived from an Australian accredited research-only online panel (Dynata). Participants were incentivized through the receipt of panel points and provided informed consent before commencing the survey. All data were self-reported and collected using a customized online deidentified questionnaire, which was completed only once. Ethical clearance was received (Bellberry Human Research Ethics Committee; Eastwood, SA, Australia: HREC2020-05-494-A1).
The key study eligibility criterion 'self-reported symptoms of OA' was based on six screening questions designed specifically to capture a representative population of people with OA without seeking formal evidence of a clinical diagnosis. Three of the screening questions (1) 'age 45 years or over', (2) 'has activity-related joint pain' and (3) 'has either no morning joint-related stiffness or morning stiffness that lasts no longer than 30 min' were based on the definition of clinically diagnosed OA supported by the National Institute for Health and Clinical Excellence Guidelines for OA 6  The remainder of the questionnaire comprised validated tools to evaluate health literacy, [23][24][25] health-related quality of life, 26,27 OA symptoms, impact and comorbidities, 28-36 medication beliefs and adherence [37][38][39] and other questions developed empirically through review of the literature. 22,[40][41][42] A group of consumers living with OA contributed to selecting the study outcomes, reviewed the survey questionnaire for face validity and provided advice on plain language terminology. Literature-based quality indicators 4,5,43 were predefined in the protocol based on a traffic light system (green, amber, red) and used to categorize the appropriateness of the respondents' choices for OA self-care.
Within this system, green represented core strategies and/or those with a strong recommendation or conditional recommendation (Levels 1A, 1B, 2), amber strategies with a conditional recommendation neutral or against (Levels 3, and 4B) and red strategies, which are not recommended (Levels 4A and 5).
The primary analysis population comprised all participants who fulfilled the criteria for self-reported symptoms of OA. Four prespecified, but not mutually exclusive, subpopulations were defined to account for differences in management recommendations amongst participants with comorbidities: OA and no comorbidity, OA plus gastrointestinal comorbidities, OA plus cardiovascular comorbidities and OA plus widespread pain and/or depression.
The average proportion of green, amber and red management strategies that a respondent was using provided a numeric measure of the appropriateness of their self-care choices, and health literacy was measured using a categorical variable that classified each person's health literacy into low (score ≤ 32), moderate (score: [33][34][35][36][37][38] or high (score > 38) using validated cut-off criteria. 25,44 For each participant, the number of appropriate strategies (based on the traffic light system, green/amber/red) they reported using were first summed and then divided by the total number of strategies recommended for their level of comorbidity.
A score of 0 was used to indicate that they are not using any green/amber/red strategies, a score of 0.5 was used to indicate that they are using half of the green/amber/red strategies and a score of 1 was used to indicate that they were using all green/ amber/red strategies. A higher score was, therefore, indicative of more appropriate choices for green strategies, but less appropriate choices for red strategies. A one-way multivariate analysis of variance test, with the three management appropriateness metrics as the dependent variables and health literacy as the independent variable, was used to test for statistical significance.
Baseline variables and outcome measures were analysed among the primary and subpopulations. Continuous variables were summarized descriptively, and differences were analysed using two-sample t-tests and categorical variables were summarized as contingency tables using sample sizes and percentages with two-sided 95% confidence intervals (CIs) and analysed using two-sample z-tests. A p-value less than .05 was considered significant. All statistical analyses were generated using Q-research software (Ver 5.9.7.0; Displayr).

| RESULTS
Out of the total pool of respondents (N = 6800), 6348 were in the target age range (45-74 years) and answered all screening questions.
Of these, 697 (10%) fulfilled all of the screening criteria for selfreported OA, and 69 withdrew consent, leaving a population of 628 eligible participants (Table 1). Detailed demographics for the primary and secondary populations are provided in Table S1.

| Health literacy and appropriateness of management choices
The proportions of participants classified as having low, moderate or high health literacy were similar in the overall population and the secondary analysis populations. A significantly higher proportion of females were categorized as having high health literacy than were males ( Table 1). The primary endpoint analysis showed evidence of a statistically significant interaction between health literacy and overall appropriateness of management strategies (p < .001). 18 3.2 | Baseline characteristics, pain symptoms and self-recognition of OA The majority of participants (82.5%) were nonsmokers; almost threequarters were overweight or obese and health-related quality of life was generally good ( Table 2). Participants reported having current pain in 4.1 joints; the most frequent location of joint pain was in the lower body, and the majority (82.1%) had been experiencing joint pain for more than a year ( Table 2).
All participants fulfilled the study eligibility criterion for 'selfreported symptoms of OA', but when presented with a list and asked to indicate if they had any current health conditions or were taking medications/receiving treatment for current conditions, 40.8% selfidentified OA and, amongst those, only half (52.0%) reported currently taking medication and/or receiving treatment for OA. Women, participants aged 55-74 years and those who had experienced pain for longer (5+ years) were more likely to self-identify as having OA

| Resources used for pain management
The general practitioner/family doctor was the main healthcare provider from whom advice about current pain management was received (Table 3). Almost 4 out of 10 (37.9%) participants reported that they were not receiving help from anyone to manage their pain  (Table 3), and a similarly high proportion (44.9%) reported that they used the internet search as a source for advice or information about how to manage their joint pain. There was an overall association between health literacy and advice seeking, but the differences were not statistically significant. When compared to participants classified as having moderate (64%) and high (63%) health literacy, fewer (54%) participants with low health literacy reported having sought any advice for pain management (Table 3).

| Awareness, choice and use of nonprescription analgesics
There were high levels of awareness of different classes of analgesics that could be purchased without a prescription, and paracetamol was reported to be the most frequently used of these analgesics ( Table 4).
The two main sources of recommendation for use of nonprescription analgesics were the family doctor and the pharmacist, and these healthcare professionals prompted current use in around half of the respondents. Around 30% of participants reported that no one had recommended the use of their current analgesic; efficacy and ease of use were the main reasons cited for prompting use across all classes of non-prescription analgesics ( Table 4). Rates of adherence were suboptimal; less than two in five participants were found to have a high level of adherence irrespective of the nonprescription analgesic being used. Around three-quarters of participants reported being satisfied, or highly satisfied, with the pain relief achieved from their analgesic ( Table 4). The pharmacy was the primary place to purchase nonprescription analgesics, accounting for 75%-80% of purchases depending on the analgesic, with 25%-35% of purchases being made in the supermarket; some people reported purchasing from both places.

| DISCUSSION
A recent online survey amongst adults with self-reported symptoms of OA has identified limited self-recognition of OA, suboptimal knowledge about suitable choices for OA self-care and a statistically significant interaction between health literacy and overall appropriateness of management strategies. 18 Further exploration of this data has identified a high level of self-reliance for pain management advice-38% of participants were not receiving help from anyone and 45% were using the internet as a source of advice-an association between not seeking advice and inadequate health literacy, and ease of use as the key driver prompting the trial of a nonprescription analgesic.
The study results are strengthened by the use of a wide selection of validated assessment tools and the large prospective sample of more than 600 respondents but limited by the potential for bias from the online survey methodology. 18  Others have suggested that patients with early knee OA may lack sufficient knowledge to recognize their condition. 45 Our results corroborate this, and suggest that a lack of self-recognition of OA may also lead to poor self-management. Despite meeting guideline-  Resource utilisation (who is currently helping you to manage this pain?) up to 35% of users of nonprescription analgesics reported purchasing from a supermarket. The main reasons for choice of analgesic were efficacy and ease of use, and many respondents did not seek any advice on recommendation. This might imply that a proportion of respondents had been using whatever analgesic they had available rather than seeking out advice for an analgesic specifically for their joint pain. Further research would be needed to better define these parameters and to determine if there is any correlation between analgesic choice based on convenience and suboptimal pain control.
Adequacy of health literacy is associated with a patient's ability to access, understand, appraise and interpret health information. 49 Patients lacking these core competencies in health literacy are limited in their ability to apply information to their current medical or clinical issues, their risk factors for health and to determinants of health in their social and physical environments. All of these can impact decision-making in relation to healthcare needs, disease prevention and health promotion. Having adequate health literacy has been highlighted as a key factor in patients' decisions to adopt and engage with OA self-management strategies. 11 Health literacy has the potential to influence the appropriateness of self-care strategies and could therefore impact outcomes achievable via patient-directed OA management. 8,9,50 The primary endpoint analysis from the study highlights this to be of particular relevance. 18 Knowledge about sui-