After patients are diagnosed with knee osteoarthritis, what do they do?


  • Presented in part at the 71st Annual Scientific Meeting of the American College of Rheumatology, Boston, MA, November 2007, and the Canadian Pharmacists Association Annual General Meeting, Victoria, British Columbia, Canada, June 2008.



To learn more about the health services and products that patients use after receiving a diagnosis of knee osteoarthritis (OA), as well as the trajectory of their health-related quality of life (HRQOL).


Using a simple screening survey, community pharmacists identified 194 participants with previously undiagnosed knee OA. Of these participants, 190 were confirmed to have OA on further investigation. At baseline and 1, 3, and 6 months after diagnosis, a survey was administered to assess health services, product use, and HRQOL, including the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), the Medical Outcomes Study Short Form 36 (SF-36) health survey, the Paper Adaptive Test (PAT-5D-QOL), and the Health Utilities Index Mark 3.


With a mean age of 63 years, participants were mostly women, white, and overweight. By 6 months, more than 90% of the participants had visited their family physician to discuss their OA, and more than 50% of participants took either prescription or nonprescription analgesics. In addition, three-quarters of the participants started exercising, one-third initiated activity aids, and one-third had started natural medicine products. At 6 months compared with baseline, significant improvements were seen in the SF-36 physical component summary (P = 0.001) and bodily pain domain scores (P = 0.02), the PAT-5D-QOL overall, pain, and usual daily activities scores (P < 0.001 for all), and the WOMAC total, pain, and function scores (P < 0.001 for all).


Within 6 months of receiving a diagnosis of knee OA, participants made several lifestyle interventions, often without the advice of a health professional, and saw improvements in their pain and function.


Osteoarthritis (OA) is the most common form of arthritis and contributes to half of all disability among older persons in North America (1). Guidelines for the management of OA recommend that treatments be tailored to patient needs and include nonpharmacologic therapies such as education, exercise, appliances, and weight reduction in addition to the use of analgesic medications (2, 3). Unfortunately at present, many patients with OA do not receive the care needed to improve symptoms or slow disease progression (4, 5).

In today's health care system, chronic disease management in the primary care setting has become a focus of many advocates and critics alike. In the case of arthritis, the burden of the disease on health care systems and on the patients' quality of life underscores the need for programs that effectively deliver ongoing care. In most developed countries, pharmacists are the most accessible health professional. Located in multiple community-based settings, they are generally available around the clock and on demand. In addition, pharmacists that actively participate in chronic disease management improve patient outcomes, the use of health services, and health care costs (6, 7).

Previously, we described the results of a pilot study demonstrating that community-based pharmacists, using a simple screening questionnaire, can reliably identify patients with undiagnosed knee OA (8). The majority of the participants with knee OA were experiencing the early course of disease, were overweight or obese, and were likely to benefit from nonpharmacologic and, in some cases, pharmacologic interventions. In this study, we describe the followup assessment of the pilot study participants to explore the use of health services and products following the diagnosis of knee OA.

The objectives of this longitudinal study were to describe the participants' lifestyle interventions as well as use of health services and products following diagnosis, and to assess the health-related quality of life (HRQOL) of community pharmacy clients with recently diagnosed knee OA. The methodology used to identify participants with knee OA is described in detail elsewhere (8).


Briefly, participating community pharmacists were located in 27 pharmacies in the Greater Vancouver Regional District of British Columbia and in the city of Edmonton, Alberta, Canada. Through display posters in the pharmacy department and newspaper flyers, we recruited participants who had pain, aching, or discomfort in or around the knee during the previous year lasting >4 weeks and who were ≥50 years of age. Participants were excluded if they reported having a previous diagnosis of OA, rheumatoid arthritis, gout, psoriatic arthritis, ankylosing spondylitis, or fibromyalgia. Pharmacists assessed participant eligibility using a screening questionnaire (8). Informed consent was provided by each participant, and the Clinical Research Ethics Board at the University of British Columbia approved the study protocol. The study was initiated in 2005, with the 6-month followup completed in early 2006.

Prior to the study, pharmacists were provided with an information session on the management of OA and oriented to the questionnaire. During the study, participants thought to have knee OA were identified by pharmacists. Following this identification, primary care practitioners were contacted to confirm that participants did not, in fact, have a prior diagnosis of OA and to review the medical chart for any knee-related symptoms. Eligible participants were then referred to a local arthritis center for physical examination and radiographs to confirm the diagnosis of OA.

All participants with confirmed OA were offered enrollment in the Arthritis Society's Arthritis Self-Management Program and a brief information pamphlet on OA. In addition, a baseline questionnaire was used to obtain the participants' demographic information. At 1, 3, and 6 months, a followup questionnaire was administered to determine the participants' self-reported activities for managing their OA. Specifically, the questionnaires inquired about exercise, analgesic medications, treatment/aids, and natural medicine supplements, and they also inquired about who provided advice on these treatment interventions (e.g., physician, family member, physiotherapist). Because we did not ask about the identity of the practitioner, we could not determine the number of unique practitioners that provided advice.

HRQOL was also measured at baseline and at each followup period using the following 4 questionnaires: the complete Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC); the Paper Adaptive Test (PAT-5D-QOL); the Health Utilities Index Mark 3 (HUI-3); and the Medical Outcomes Study Short Form 36 (SF-36) health survey. The SF-36 and HUI-3 are generic instruments used extensively in both disease-specific and general health assessments including OA (9, 10). The WOMAC is a widely used measure of symptoms and physical disability that originate from knee and hip OA (11). The PAT-5D-QOL is a new adaptive generic quality of life instrument developed to specifically assess musculoskeletal conditions (12).

The participants' baseline characteristics were described using summary statistics including means, SDs, frequencies, and percentages. The differences between demographics of those with partial followup and those with complete followup were investigated using chi-square and 2 sample t-tests. The differences between HRQOL at baseline and at 6 months approximated a normal distribution and were assessed using paired t-tests. Statistical significance was set at 0.05 (2-tailed). The minimum important difference (MID) was used to find the percentage of participants who met the minimum level of change in the domains of the HRQOL scores.


Of the clients that expressed an interest in the study, only 7% were picking up a prescription or an over-the-counter medication for knee pain. As described elsewhere (8), more than 400 pharmacy clients were screened, and 194 clients were thought to have previously undiagnosed knee OA. Using physical and radiographic examination, the study rheumatologist (JC) and physiotherapist confirmed that 190 participants had knee OA (Figure 1). Partial followup data were available for all 190 participants, and complete followup data were available for 124 participants (65%). The participants' mean age was 63 years, and most participants were women, white, and either overweight or obese (Table 1). There were no significant differences in the demographics of participants between those with partial followup data and those with complete followup data.

Figure 1.

Trial flow of recruited participants. OA = osteoarthritis; F/U = followup.

Table 1. Baseline characteristics of participants with partial and complete followup at 6 months
CharacteristicComplete followup (n = 124)Partial followup (n = 66)
Women, no. (%)79 (64)40 (61)
Ethnicity, no. (%)  
 White110 (89)53 (80)
 East Asian (Asian)3 (2)4 (6)
 Southeast Asian (East Indian)3 (2)4 (6)
 Other8 (7)5 (8)
Age, mean ± SD years63.1 ± 9.060.6 ± 7.3
Height, mean ± SD cm167.9 ± 9.0168.4 ± 10.4
Weight, mean ± SD kg77.2 ± 17.579.3 ± 17.0
Body mass index, no. (%)  
 Underweight (<18.5 kg/m2)1 (0.8)0 (0)
 Normal (18.5–24.9 kg/m2)39 (31)16 (24)
 Overweight (25.0–29.9 kg/m2)53 (43)32 (49)
 Obese (>30.0 kg/m2)30 (24)18 (27)

In the first month after diagnosis, <50% of the participants initiated aerobic, strength, or endurance exercises (e.g., walking, biking), whereas nearly 75% claimed to participate in routine exercise by month 6 (Figure 2). Of those participants initiating exercise, 80% believed that exercise helped. Activity aids such as acupuncture, knee tape, knee braces, and shoe inserts were used by 33% of the participants and 75% found this to be helpful. While all participants were provided with information on the 6-week Arthritis Self-Management Program (12 hours total), only 8% completed the program.

Figure 2.

Treatment interventions started by participants within 6 months of receiving a knee osteoarthritis diagnosis. Exercise includes aerobic, strength, endurance, and walking exercise. Treatment/aids include acupuncture, knee taping, knee brace, and shoe inserts. Supplements include glucosamine and chondroitin natural medicine supplementation.

In total, 52% of participants started analgesic therapy (Figure 3), 36% started natural medicine supplementation, and 66% started therapy from one or both categories. Of those taking analgesic medications, 50% of the participants initiated nonsteroidal antiinflammatory drugs (NSAIDs) and ∼25% tried acetaminophen. Of those participants initiating glucosamine and/or chondroitin, 60% started glucosamine alone and 40% started a combination of glucosamine with chondroitin. By 6 months, more than 50% of the participants believed these natural medicine supplements were beneficial in the management of their knee OA.

Figure 3.

Pharmacologic therapy started by participants within 6 months of receiving a knee osteoarthritis diagnosis. NSAID = nonsteroidal antiinflammatory drug.

By the end of the 6-month followup period, 93% of the participants had visited their primary care practitioner to discuss their recent diagnosis of knee OA. While a variety of different individuals were credited with recommending interventions for the management of OA, more than 50% of the participants stated that they decided to start exercising, using activity aids, or taking analgesic medication on their own (Figure 4). Family physicians were the second most common contact, followed by family members and friends. Pharmacists, despite identifying the undiagnosed knee OA, were credited with providing advice in only 2 instances: once for analgesic medications and once for glucosamine. The decision to start natural medicine supplementation with glucosamine and/or chondroitin was made by 44% of the participants alone and on the advice of friends and family in 20% of cases.

Figure 4.

Source of the recommendation to start treatment interventions within 6 months of a knee osteoarthritis diagnosis. Exercise includes aerobic, strength, endurance, and walking exercise. Treatment/aids include acupuncture, knee taping, knee brace, and shoe inserts. Supplements include glucosamine and chondroitin natural medicine supplementation. GP = general practitioner.

As seen in Table 2, at 6 months the mean changes in SF-36 physical component summary and pain domain scores were significantly different from baseline. Significant changes were also seen in the WOMAC global, pain, and function domain scores, and the PAT-5D-QOL global, pain, and discomfort and usual daily activities domain scores. No significant changes were observed in the SF-36 physical function domain or the HUI-3 global scores. Sixty-three percent of participants met or exceeded the MID in the pain domains of the WOMAC and PAT-5D-QOL, whereas ∼50% met the MID in the SF-36 pain domain. In the function domains, 64% met or exceeded the MID for the WOMAC, 54% for the SF-36, and 58% for the PAT-5D-QOL. Just under 50% of the participants met or exceeded the global HUI-3 MID.

Table 2. Participants' change in HRQOL 6 months after knee OA diagnosis*
HRQOLNo.Mean ± SD baseline scoreMean change from baseline (95% CI)PImprovement, no. (%)/MID
  • *

    HRQOL = health-related quality of life; OA = osteoarthritis; 95% CI = 95% confidence interval; MID = minimum important difference; WOMAC = Western Ontario and McMaster Universities Osteoarthritis Index; SF-36 = Short Form 36 health survey; PAT-5D-QOL = Paper Adaptive Test; HUI-3 = Health Utilities Index Mark 3.

  • WOMAC Likert scales, version 3.0.

  • Normalized score with population mean of 50.

WOMAC score, normalized     
 Global (0–30)1139.03 ± 0.42−1.91 (−2.72, −1.10)< 0.001
 Pain subscale (0–10)1133.17 ± 0.17−0.91 (−1.23, −0.59)< 0.00171 (63)/0.5
 Function subscale (0–10)1132.29 ± 0.16−0.54 (−0.83, −0.26)< 0.00172 (64)/0.147
SF-36 score     
 Physical component11141.78 ± 0.862.12 (0.64, 3.60)0.001
 Physical function11442.67 ± 0.891.47 (−0.05, 2.99)0.05962 (54)/5.0
 Bodily pain11544.55 ± 0.892.05 (0.27, 3.83)0.02454 (47)/6.0
 Global1160.84 ± 0.010.04 (0.02, 0.05)< 0.001
 Pain and discomfort11744.52 ± 0.733.09 (1.75, 4.42)< 0.00174 (63)/1E-7
 Usual daily activities11642.68 ± 0.702.96 (1.56, 4.36)< 0.00167 (58)/8.16E-7
HUI-3 global score830.71 ± 0.030.03 (−0.01, 0.08)0.15438 (46)/0.03


Within 6 months of a diagnosis of knee OA, the participants' pain and function significantly improved compared with baseline. In addition, many participants initiated lifestyle interventions that were consistent with current guidelines, including exercise, activity aids, and the use of analgesic medications (2, 3). One interesting finding was that most participants stated that they made these changes independent of the advice of a health care professional, despite all participants being initially identified by a pharmacist and most having followed up with their primary care practitioner.

Considering that few study participants credited a health professional with the decision to become more active, it is encouraging that many participants started aerobic and strength training within 6 months of their diagnosis. Although it remains unclear whether primary care practitioners or pharmacists actually did recommend lifestyle modifications, a recent population-based survey in the US found that fewer than 50% of adults with arthritis recalled being advised to become more physically active by a health professional (13). This suggests that many participants in this study were not advised to increase their activity level or did not consider the advice-giving role of health professionals. Whether perception or reality, this observation is an important consideration for the development of arthritis management programs that target early disease management.

Regardless of where the decision to initiate lifestyle interventions originated, the majority of study participants were overweight and may have already been contemplating lifestyle changes (14). Receiving the diagnosis of a chronic condition may have provided motivation to increase their activity level, helping to gain some control of their condition and improve symptoms (15). Although compliance with exercise-based interventions is often low and of insufficient intensity (16, 17), and although we did not ask about the frequency, appropriateness, or consistency of their exercise regimens, these findings suggest that many individuals are open to trying exercise-based lifestyle interventions, especially early after diagnosis. In addition, individuals contemplating or preparing for changes may require a health professional's advice on significant lifestyle changes later in the course of the disease.

The outcome that was most concerning was the high number of participants starting analgesic therapy with NSAIDs prior to an appropriate trial of acetaminophen. Although patients generally prefer NSAIDs (18), the initial treatment of choice for pain relief in mild, moderate, and severe knee OA is acetaminophen because of its efficacy, safety, and low cost (2, 3, 19). Despite clear treatment recommendations and participant access to health professionals well versed in medication therapy, only 25% of participants tried acetaminophen therapy during the followup period, while 50% tried NSAIDs. Furthermore, the decision to try analgesic therapy was often made on their own or on the advice of family or friends, suggesting that many may choose to disregard the advice of health professionals in favor of their own preferences or well-being (20). This is further evidenced by the large number of participants who started glucosamine, despite evidence of limited efficacy (21). However, a limitation to our approach was that precautions (hepatic dysfunction, alcohol use) and previous history of previous treatment failures with acetaminophen were not assessed to determine the influence of these factors on therapeutic selection by participants.

In preparation for a randomized controlled trial of a multidisciplinary approach to the management of knee OA, this pilot study assessed whether pharmacists could reliably identify participants with undiagnosed knee OA living in the community. The study is subject to limitations. The first major limitation is that the study design did not include a control group. Second, we relied on participants to self-report their interventions, which is subject to recall bias and likely overestimated their actual involvement in activities such as exercise. Third, we did not obtain baseline information on the prior use of interventions such as lifestyle modifications or analgesic therapy. Rather, participants were asked to describe whether they initiated or modified their use of exercise, activity aids, natural medicine supplements, and analgesic therapy, thereby limiting our ability to comment fully on the effect of a diagnosis of knee OA.

In keeping with the observation that pharmacologic therapy was not optimal in the 6 months following a diagnosis, pharmacists did not appear to be providing ongoing care to participants. The limited involvement of pharmacists in patient care was unexpected since this study introduced the pharmacist as the first point of contact for participants and demonstrated that a pharmacist can reliably identify a chronic medical condition that was previously undiagnosed. Furthermore, this is all the more relevant with the publication of a recent study demonstrating that medication reviews by a pharmacist can reduce potentially dangerous NSAID use in those with knee OA (22).

In conclusion, this study demonstrated that following the identification of previously undiagnosed knee OA by pharmacists, many participants initiated nonpharmacologic interventions, including exercise, knee taping, knee braces, and insoles. In addition, half of the participants started analgesic medications, half of which started NSAIDs. Also, most participants made the decision to initiate interventions on their own. While nearly all participants visited their family physician within 6 months, health professionals were rarely credited with advising participants on interventions. This study provides valuable insight into decisions made by patients following a diagnosis of knee OA and their perceptions of the role of health professionals. In addition, it provides some areas for health professionals to improve care with a focus on the initiation of appropriate analgesic therapy.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Marra had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Marra, Cibere, Tsuyuki, Esdaile, Gastonguay, Kopec.

Acquisition of data. Tsuyuki, Esdaile, Gastonguay.

Analysis and interpretation of data. Grindrod, Marra, Colley, Cibere, Tsuyuki, Gastonguay.


Merck Frosst Canada had no role in the study design, analysis, or interpretation of the data. Merck Frosst Canada was not involved in the writing of the manuscript and did not approve the final version for publication.


The authors wish to thank Dr. Judith Soon, acting director for the Community Pharmacy Research Network, for assistance with the recruitment for the study and the study participants.