Chronic conditions, at accelerating rates globally, are the leading causes of death and disability [1]. Due to an increase in average life expectancy and a higher prevalence of obesity and sports-related joint injuries, the estimated disease burden of one of the most common and costly disabling chronic diseases, osteoarthritis (OA), will markedly increase over time [2]. Thus, health care systems need to prepare for the large increase in demand for OA treatment [3]. According to evidence and international guidelines, the primary treatment of knee and hip OA is nonsurgical and based on exercise, information, and, when necessary, weight loss [4-6]. Despite these evidence-based guidelines, only a minority of all people with OA receive adequate treatment; for instance, more than 50% of patients with OA in Sweden are referred to surgery as a primary option [7].

An alternative to nonsurgical treatment delivered face-to-face to individuals with OA is eHealth. Web-based treatment can increase the accessibility to care, due to its inherent flexibility in comparison with face-to-face treatment. In addition, there are numerous examples of Web-based programs in the literature, reporting beneficial effects on key health-related outcomes [8-11]. To this end, we developed the Web-based OA treatment platform Joint Academy [12] based on national and international guidelines and on a successful Swedish face-to-face concept (Better Management of Patients With Osteoarthritis [BOA] program), in which, to date, more than 90,000 patients have participated [7]. Joint Academy entails a 6-week Web-based program that adheres to evidence-based OA treatment. According to a pilot study, the results of the Joint Academy program on joint pain in individuals with knee or hip OA were promising [12]. Yet the level of physical function and health-related quality of life (HRQoL) after an eHealth intervention are unknown in a larger population. Hence, the primary aim of this study was to evaluate joint pain, physical function, and HRQoL in users of the Joint Academy. The secondary aims were to investigate whether use of the 6-week program is associated with decrease fear of physical activity and desire for surgery, and improve self-reported difficulties walking. We compared pain results with those in the structured face-to-face BOA program.

This was an observational and quasi-experimental study. We recruited patients via advertisements and campaigns on the Web to participate in Joint Academy, a 6-week Web-based OA treatment program that provides information, exercises, an online physiotherapist, and education regarding factors of relevance to OA, including lifestyle. The advertisements were placed on search engines and social networks. The program is accessed by using a smartphone, tablet, or computer and encourages user activity by sending email prompts to participants on a regular basis, as described in detail previously [12]. An orthopedic surgeon and a physiotherapist, using synchronous and asynchronous chat conversations, supervised the clinical progress and were responsible for making a correct patient diagnosis and for identifying eligible patients. The program contained 2 telephone consultations with a physiotherapist that were compulsory, 1 at the start and 1 after 6 weeks. Participants’ costs were covered through either a private or a public health plan.

We asked patients to complete an e-questionnaire regarding their overall health and OA symptoms. The e-questionnaire also assessed pain level using the sensitive and reliable numerical rating scale [13], HRQoL using the 3-level EQ-5D (EQ-5D-3L) [14], and physical function measured by the 30-second chair-stand test [15], where participants sit and stand from a chair for 30 seconds, and the number of repetitions is recorded. Participants were also asked whether they experienced difficulty walking and were afraid of physical activity due to their OA (dichotomous replies), as well as whether they desired surgery. All outcomes were self-assessed and chosen according to the International Consortium for Health Outcomes Measurement Standard Set for Hip & Knee Osteoarthritis [16]. To enable a comparison with face-to-face treatment, we matched data from the Joint Academy database on reported joint pain with results at 12 weeks from the BOA initiative, found in the BOA yearly report of 2015 [17].

We defined as completers those participants who reported one of the main outcomes (joint pain, HRQoL, or physical function) at baseline and postintervention in the hip or knee. We excluded outliers of adherence—that is, participants with an activity level in the program below 10%.

Statistical analysis of pre- versus postintervention measurements was by 2-tailed t test for pain level, physical function, and HRQoL. For dichotomous and polytomous variables, we used McNemar and Madansky test of marginal homogeneity, respectively. The calculations were performed using the Statsmodels package in Python v3.6.1 (Python Software Foundation) and the coin library in R v3.4.1 (R Foundation).

We collected consent to participate at registration and obtained ethical approval from the Regional Board of Ethics in Lund, Sweden (Dnr 2017/651).

The datasets for this study are available from the corresponding author upon reasonable request.

The study cohort comprised 350 individuals with a clinical diagnosis of OA in accordance with American College of Rheumatology criteria and international guidelines as judged by a physiotherapist or an orthopedic surgeon after scrutinizing the questionnaires [18]. Table 1 presents patient demographics.

The majority reported “working” as their occupational status, while 4.0% (14/350) were on sick leave. Most participants reported level of physical activity of 90 to 150 (85/350, 24.3%) and structured exercise of less than 30 minutes per week (97/350, 27.7%).

On the basis of the requirements of available data and level of adherence, we included 71.4% (n=250) of the study cohort in the study. We used the questionnaires within the program to secure a clinical diagnosis of OA and to establish baseline study values. Table 2 shows the changes in the outcomes studied. Data were missing for pain measurement, HRQoL, and physical function; however, none exceeded 10% (10/250, 4.0%; 15/250, 6.0%; and 16/250, 6.4%, respectively).

When investigating the distribution of pain improvement over self-reported joint pain at baseline, we observed improvement in patients with both low and high initial pain. However, most of those reporting improvement had more severe pain at baseline (Figure 1).

Investigation of each dimension of EQ-5D-3L separately showed that mobility and pain or discomfort were significantly improved from baseline to follow-up, while changes in self-care, usual activities, and anxiety or depression were not significant. Of participants who completed the 6-week program, 22.0% (55/250) reported that they had reduced the amount of OA-related medication, while 24.1% (13/54) no longer desired surgery (13 of 54 individuals altered their opinion).

The comparison between Joint Academy and the BOA initiative showed similar levels of pain at 3 months but suggested a greater reduction in Joint Academy due to a higher level of pain at baseline (Table 3).