My joint pain, a web-based resource, effects on education and quality of care at 24 months

Osteoarthritis (OA) is a leading cause of chronic pain and one of the top contributors to global disability [1]. As the most common form of arthritis, OA currently affects 2.1 million (9%) Australians of all ages. Particularly, in people over the age of 45, the prevalence increases to 21% [2]. In addition to the individual burden of OA, the direct and indirect costs also contribute to a substantial socioeconomic cost [3]. By 2020, it is estimated that the prevalence of OA will increase by 50% due to an ageing and increasingly obese population [4].

Evidence-based clinical guidelines recommended non-surgical, non-pharmacological treatments, such as exercise, weight loss and self-management education, as the cornerstone for the care of OA [5‐8]. However, current clinical practice is still largely limited to analgesic treatment followed by eventual joint replacement surgery [9]. The underuse of recommended management strategies and low levels of referrals to allied health professionals who can provide lifestyle interventions have been frequently reported within Australia and globally [10, 11]. In fact, 81% of patients indicated they would not accept surgery if alternative treatments were offered, and physiotherapy was the most popular option [12]. Additionally, people with OA often lack enough knowledge of their condition and have limited involvement in clinical decision making [13]. The provision of evidence-based information to patients outside the clinical encounter has long been carried out aiming to increase their self-efficacy and health outcomes. These included providing structured education and additional personalised support, allowing patients to play an active role in improving their condition [14, 15]. However, current OA self-management education programs have small or no clinical benefits and do not improve patient’s self-management skills [15]. It is suggested that investigations of alternative models of delivery for those programs may be warranted [15].

In recent years, the use of the Internet as a source of health information to aid the management of chronic diseases has become increasingly popular across diverse socioeconomic and age groups [16]. With the rising disease burden of an ageing population on the overstretched healthcare system, more people accept Internet-based health solutions as part of their management, especially if it facilitates their personal care [17]. Compared to in-person interventions, digital health interventions are more cost-efficient and can reach a broader population in remote areas [18]. It allows people with chronic conditions to administer their treatment at a suitable time point which could increase treatment adherence and compliance, ultimately, improving their health and quality of life. However, Internet-based health resources still have many obstacles which limited their medical usage. For example, much online information often failed to incorporate with up-to-date research evidence [19]. An investigation of 37 unique websites found the readability and quality of online health information for OA was more difficult than the recommended level defined by the Journal of the American Medical Association benchmark criteria [20]. Despite the increasing utilisations of Internet-based self-management among people with chronic diseases, only a few studies have targeted people with OA [21, 22]. These studies provide preliminary evidence of the effectiveness, acceptability and feasibility of different Internet-delivered self-management interventions, but long-term data is scarce.

To this end, we have previously investigated the effect of a publicly available resource, the ‘My Joint Pain’ website, which contains evidence-based OA management resources and self-management tools aiming to empower users in the clinical encounter and informed decision-making [23]. In a 12-month evaluation study, we found patients who used the website had significant improvements in several health education domains including health-directed activity, positive and active engagement in life, self-monitoring and insights, skill and technique acquisition, and social integration. Improvements in important aspects of quality of care, such as self-management, lifestyle and weight reduction, were also observed in people used the website [23]. Although the website provides extensive information on evidence-based treatments and facts about the disease, appropriate psychological approaches and technological strategies are still insufficient to assist the delivery of that information and shift participants’ viewpoint about the impact of OA [23]. In regard to health service navigation, the website does provide a tool to locate nearby healthcare resources. However, this tool is not easily accessible, well-integrated, and not located on the main toolbar, which might account for the observed outcome [23]. In response to our findings, the website has been redesigned to address the gaps highlighted by the results. The current study aims to investigate further the effects of the updated website on improving consumers’ health education and quality of care over 12 months.

Among the 300 invited participants, 289 participants completed the baseline questionnaires of which 12 participants were excluded due to missing/invalid data entry or not eligible (2 had no knee/hip pain and no previous diagnosed OA). A total of 277 people was included at baseline and 195 participants (70%) completed outcome measures at 12 months. At 24 months, we observed further attrition and received 122 completed responses (non-users: n = 87, users: n = 35) (Fig. 2). The response rate of 44% was much lower at this final 24-month follow up than at 12-months. Of the responders at 24 months (n = 122), 28% (n = 35) indicated they had used the updated website which included 12 new users (i.e. non-users at 12 months and users at 24 months) and 23 continuous users (i.e. users at both 12 and 24 months). Similar baseline demographics (e.g. age, gender, BMI and joint affected) were observed among non-users and non-completers at 24 months, but users at 24 months were slightly older, had lower BMI and more females than other groups (Table 1).

Within the user group, no significant changes were found in all the other domains of the HEIQ after adjustments for baseline age, sex and BMI. There were slight improvements in skill and techniques and constructive attitudes and approaches from 12 to 24 months. Non-users at 24 months showed slightly lower scores in all domains when subjected to the same analysis. There were improvements in skill and technique acquisition and less emotional distress in non-users but these changes were not statistically or clinically significant (Table 2). The comparisons between users and non-users at 24-months also showed no statistical and clinical significance in both univariable and multivariate models (Table 3). Only “emotional distress” score at 24 months was significantly higher in users (RR: 1.26, 95%CI: 0.11, 0.43), but became non-significant after adjusting for age, sex and BMI (Table 3). Excluding those crossed over from users to non-users over the 24-month period, a further sensitivity analysis was performed in people who were continuous users and non-users at 12 and 24 months showed similar results but even smaller differences (Additional file 1: Table S1).

The results of OAQI pass rates showed higher pass rates (i.e. better quality of care) in “Treatment Alternatives” (12%), “Self-Management” (17%), “NSAIDs” (13%) and “Referral to Orthopaedic Surgeons” (23%) domains and lower pass rates in “Functional Assessment” (11%) and “Walking Aid Assessment” (7%) in users of the website comparing to non-users at 24 months (Table 4). A higher proportion of users experienced an improvement of OAQI from baseline to 24 months in “Acetaminophen (Paracetamol)” compared to non-users (23% versus 8%, P <  0.05), but due to very limited number of participants, the differences were not considered as clinically meaningful.

This study evaluated the effects of health impact and quality of care after using an updated version of ‘My Joint Pain’, a web-based platform for OA self-management education at 24 months compared to non-users. There were no significant changes in health education measuring by the HEIQ from 12 to 24 months in users of the website. Between-group (users versus non-users) differences failed to find statistically significant favourable changes in HEIQ, however, non-users seem to experience less emotional stress. Users showed higher improvements on several items in the OAQI, including appropriate information about self-management, treatment alternatives and the use of NSAIDs (effects and side-effects). However, outcomes from this study were of little clinical relevance given the small differences observed between groups and the increasing loss of volunteers throughout the study.

The findings of the 24-month evaluation of the ‘My Joint Pain’ website were similar to the 12-month evaluation improving knowledge and healthy behavioural changes especially self-management and pharmacological treatments [23]. These findings reinforce the concept that the website aids in promoting self-management and non-surgical, non-pharmacological treatments which are the first-line recommendations [30].

An unexpected finding is the deterioration of emotional distress at 24 months in the users compared to non-users. The reason for the worsening emotional wellbeing could be attributed to the natural history of the disease that continuously progresses over time. Evidence also suggests people with fewer symptoms are more likely to drop out in both clinical and research settings [31], and those who needed more support or had more severe symptom tend to complete follow-ups.

The ‘My Joint Pain’ website addressed several barriers to self-management, such as financial limitations and lack of personalised recommendations by providing free resources with comprehensive assessment instruments to tailor strategies [32]. From the substantial reduction in the number of continuous users over time, it suggests that there is still a lack of a broader endorsement that allows healthcare professionals to encourage their patients to use the website. This has been recognised as a key barrier for the uptake of self-management and education resources [33]. However, this approach has challenges given that the attitudes and beliefs of primary care providers, especially general practitioners (GPs), are not necessarily supportive of recommended evidence-based treatments such as exercise [34, 35]. This might be due to the fact that the role of GPs in initiating exercise for OA management was not outlined in guideline recommendations [36, 37] and there is no explicit expectation about whether GPs should refer patients for exercise therapies, advise general or specific exercises or prescribe exercises. Therefore, future work is also needed to identify the optimal means of supporting and educating GPs at the clinical, educational and service level, to improve certainty and confidence about the value of recommended care and to use them in practice. Importantly, to increase its adoption and uptake by consumers and health professionals, it is critical to embed an optimised platform into routine clinical practice and develop more interactive functions to facilitate shared clinical decision-making between patients and doctors.

As an emerging area, there is still room for improvement in the design of current Internet-based health interventions. Take the example of ‘My Joint Pain’ website, it provided a personalised treatment plan for the users, but has not adequately provided social support and contact with health professionals through the platform [38]. When these pillars have been addressed in other rheumatic online interventions, improvements in quality of life and global health scores were observed especially in studies that have interactive online forums and web-conferences [39‐42]. Currently, the website is English-focused which is unable to match the increasing language and literacy needs in a growingly multicultural society with over 300 languages spoken in Australian homes [43, 44]. Similar resources in the future should include content that addresses health literacy and language barrier with the inclusion of consumer assisted content reviews and multilingual resources [45]. Additionally, to maintain relevance, the online platform should be continuously updated with the latest recommendations modified to meet the needs and regulations of the local osteoarthritic community.

Several limitations of the evaluation study were identified. A critical problem was the substantial dropout of participants over the 24 months. With over 56% of the baseline population not responding at 24 months, group sizes were vastly different and below the required sample size. The attrition of volunteers also prevented further evaluation by stratified analyses into subgroups based on gender or age. The low response rate was also hampered by the limitation of the contact information collected at baseline being restricted to electronic email addresses. The lack of other contact information prevented the possibility of following other communication methods to address the lack of response. In addition, the study did not collect information about population defining characteristics such as depression and anxiety status or evaluations of the quality of life, which would have been valuable in contrast to other studies. A caveat to the use of quasi-experimental study design was the inability to randomise participants, and more crucially in a study with several time points, it was not possible to maintain grouping for the whole study. This limitation resulted in several different groups with varying degrees of usage of both the old and updated websites which could not be assessed given the small sample size. In addition, the study survey was designed as a self-reported indicator of usage limited to “Yes” or “No”. As a result, we were unable to ascertain a dose-effect of the website. Quantitative web statistics such as the number of secure logins and time spent should be collected which will be helpful for future evaluation and optimisation. Findings from this latest evaluation indicate a need for better-conducted and well-designed studies such as RCTs to efficiently evaluate the use of an online intervention.

Web-based healthcare resources are increasingly utilised as a response to the overburdened health system and an ageing population. The 24-month evaluation of the ‘My Joint Pain’ website didn’t find significant improvement in terms of health education, but it may help to deliver useful information on self-management and appropriate use of pharmacological treatments. A large number of drop-outs at 24 months suggested that there is a need to increase its adoption and uptake by consumers and health professionals. Additional research with a larger sample size would be helpful to evaluate the cost-effectiveness of this web-based intervention further. More strategies are also needed to facilitate the process of implementing evidence-based OA self-management and education resources into current practice.