This is an open-access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR mhealth and uhealth, is properly cited. The complete bibliographic information, a link to the original publication on http://mhealth.jmir.org/, as well as this copyright and license information must be included.
Osteoarthritis (OA) is a leading cause of disability in the United States. Although no disease-modifying therapies exist, patients with knee OA who increase walking may reduce risk of functional limitations.
The objective of the study is to evaluate the impact of a mobile app (OA GO) plus wearable activity monitor/pedometer (Jawbone UP 24) used for 90 days on the mobility of patients with knee OA treated with hylan G-F 20.
Patients with knee OA aged 30 to 80 years who were eligible to receive hylan G-F 20 and were familiar with smartphone technology were enrolled in this randomized, multicenter, open-label study. Patients who had a body mass index above 35 kg/m2 were excluded. All patients received a single 6-mL injection of hylan G-F 20 and wore the Jawbone monitor. The patients were then randomized 1:1 to Jawbone and OA GO (Group A; n=107) with visible feedback (unblinded) or Jawbone only (Group B; n=104) with no visible feedback (blinded). The primary endpoint was mean change from baseline in steps per day at day 90 between Groups A and B.
Baseline characteristics were similar between groups. There were significant differences between the increases in least squares (LS) mean number of steps per day (1199 vs 467,
Use of a novel smartphone app in conjunction with a wearable activity monitor provided additional improvement on mobility parameters such as steps per day and pain with walking in the 6-minute walk test in patients with knee OA who were treated with hylan G-F 20. Results also highlight the amenability of patients and physicians to using mobile health technology in the treatment of OA and suggest further study is warranted.
Osteoarthritis (OA) is a leading cause of disability in the United States [
To date, there is no disease-modifying therapy available for OA [
Walking has been shown to significantly reduce symptoms and the risk of functional limitations due to knee OA [
Adoption of mobile health apps has the potential to improve patient outcomes. Mobile health apps have been shown to be a useful tool in weight loss programs [
The use of mobile apps in health care may be one such strategy to increase physical activity and enhance OA management. However, there are currently no published data on the use of mobile apps for the management of knee OA.
The objective of this study was to examine the impact on mobility in knee OA patients who were treated with hylan G-F 20 and standard of care follow-up plus a mobile smartphone app (OA GO) and a wearable activity monitor (Jawbone UP 24) versus hylan G-F 20 and standard of care follow-up only over 90 days. We hypothesized that patients using the OA GO app and a wearable activity monitor would experience a positive impact on their mobility compared with those not using the app.
Consecutive patients with knee OA whom the physician investigator decided to treat with one 6-mL injection of hylan G-F 20 (in accordance with the US label) who consented and qualified were enrolled in this open-label, multicenter, randomized, parallel-group study. Study sites and physicians were recruited using a detailed questionnaire and site qualification visit. Patients were recruited from the selected private community-based practices and research-only practice sites. All eligible patients were able to read and understand English and provided informed consent prior to starting the study. The protocol complied with the 18th World Health Congress (Helsinki, 1964) recommendations and applicable amendments and with any applicable country-specific laws, regulations, and guidelines. The protocol was approved by relevant ethics committees and/or institutional review boards. After working with the US Food and Drug Administration (FDA) to initiate registration of the trial, the sponsor deemed the study not eligible to be posted on ClinicalTrials.gov considering that clinical studies using devices whereby the primary outcome measure relates solely to feasibility and not to health outcomes are excluded from registration. The FDA concurred there was not a requirement to register the trial at that time.
To be included in the study, all patients must have had unilateral knee OA and have been suitable for treatment with hylan G-F 20 based on the decision of the physician investigator. Study sites were requested to report whether or not a knee examination was performed or x-ray was taken but the results of these examinations were not captured.
Patients were excluded if they were aged younger than 30 years or older than 80 years, were unfamiliar with smartphones, or had baseline pain greater than 9 on the 11-point numeric pain rating scale (NPRS; pain ratings could range from 0, no pain, to 5, moderate pain, to 10, worst possible pain) in the target-for-treatment knee while walking on a flat surface. Patients with bilateral disease were excluded with the exception of patients who were treated in only one knee and had contralateral knee pain less than 4 on NPRS while walking on a flat surface. Patients whose baseline daily step average was less than 500 or more than 8000 as assessed during the screening and run-in phases were not eligible. Also excluded were patients who had a body mass index (BMI) greater than 35 or life expectancy less than 12 months, were currently using a wearable activity monitor or analogous device, had planned surgery on any lower extremity joint or any significant medical condition that would interfere with study participation, were chronic narcotic users, or were pregnant or breastfeeding or likely to become pregnant. The protocol did not specifically exclude patients based on use of previous intra-articular injections (cortisone or hyaluronic acid).
All eligible patients received hylan G-F 20. Participants wore a commercially available activity monitor (Jawbone UP 24) on their wrists and were instructed to remove the monitor only during the weekly charging times and in situations where the device would be submerged in water. Patients were randomized 1:1 (stratified by site) to Group A or Group B. The randomization scheme was generated by the study sponsor and stratified by site. Sealed envelopes, numbered in an ascending order for use, were provided to each site. The envelopes were opened according to ascending sequence to ensure proper randomization.
Group A patients were provided with regular follow-up as per standard-of-care (information on the benefits of walking in a brochure available from the Arthritis Foundation) plus an unblinded wearable activity monitor and a mobile app (OA GO) (
The primary endpoint was mean change from baseline to day 90 in mobility as measured by steps per day. Baseline steps per day was the average over at least 3 days, and last assessment was the average steps per day over the 7 days immediately preceding the last visit. Change was calculated as the difference between the 2 averages. Use of average steps per day avoided bias due to any one particularly good or bad day for the patient. The secondary endpoints were mean percentage change from baseline in steps per day at each assessment visit (average of a 7-day period), and at day 90, mean percentage change from baseline in the 6-minute walk test (distance and pain assessed by the NPRS), patient and physician satisfaction with treatment, percentage change in Patient Activation Measure (PAM)-13 questionnaire score [
For Group A, the satisfaction of patients and physicians was captured using survey questionnaires. The patient survey consisted of 7 questions: the first 6 were answered on a 7-point Likert scale related to the extent to which the patients experienced changes (−3 to +3, where 0=no change), and the seventh was a 4-point scale asking about their likelihood of using the device in the future (not at all likely to very likely). Valid questionnaires were those in which at least 4 of the first 6 questions were answered. A physician survey was similarly constructed with a final question, using the 4-point scale, which asked whether they would recommend the device in the future (not likely to very likely). Physicians answered the survey at day 90, before viewing the data, and must have answered at least 3 of the first 4 questions. All responses to survey questions were based on patient and physician observations and opinions. The survey questionnaires were developed for this study and were not externally validated.
OA GO mobile app.
The PAM-13 [
The VAMS was used to assess mood as captured in 8 domains with raw scores transformed into T-scores with a mean of 50 and standard deviation of 10. The VAMS has been validated in both normal and neurologically impaired individuals as a brief measure of internal mood state [
The Consolidated Standards of Reporting Trials guidelines were followed [
All efficacy endpoints were analyzed in the modified intent-to-treat population, which included all randomized patients who had both a baseline value and an on-treatment value for the primary endpoint on or after day 30. All efficacy endpoint data comparing change from baseline between groups were obtained from an analysis of covariance model with baseline mean steps per day as the covariate and treatment assignment and pooled site as class variables. They are presented as least squares (LS) means with 95% confidence intervals (CIs). Statistical analyses were based on initial verification of parametric model assumptions, and a rank analysis of covariance was then considered if conditions for a parametric model were not met. To allay concerns by the readers, nonparametric analyses are presented here. Patient and physician satisfaction surveys at last visit were summarized.
Safety endpoints were analyzed for all patients provided with a wearable activity monitor. Adverse events (AEs) were presented as the number and percentage of patients experiencing an AE. Multiple occurrences of the same event in the same patient were counted only once. The denominator for computation of percentages is the safety population within each randomized group.
All patients had a knee exam at baseline, and 119 of 211 (56.4%) had an x-ray of the knee. A total of 211 knee OA patients treated with hylan G-F 20 were randomized (Group A=107; Group B=104), and nearly all patients completed the 90-day observation period (Group A, 104/107 [97.2%]; Group B, 103/104 [99.0%]) (
Patient disposition (a: 90-day observation period, b: 91 to 180 days).
Patient demographics and baseline characteristics.
Group A |
Group B |
Total |
||
Age (years), mean (SD) | 61.6 (9.5) | 63.6 (9.3) | 62.6 (9.4) | |
Male, n (%) | 48 (44.9) | 57 (54.8) | 105 (49.8) | |
Caucasian, n (%) | 87 (81.3) | 98 (94.2) | 185 (87.7) | |
BMIa, kg/m2, mean (SD) | 29.4 (3.9) | 29.3 (3.4) | 29.3 (3.7) | |
Steps per day, mean (SD) | 4279.7 (1787.3) | 4271.5 (1837.0) | 4275.7 (1807.2) | |
Pain NPRSb, mean (SD) | 4.6 (2.3) | 5.1 (2.0) | 4.8 (2.2) | |
Distance, meters, mean (SD) | 402.8 (120.5) | 395.6 (104.2) | 399.3 (112.6) | |
Activation score, mean (SD) | 71.7 (13.5) | 70.6 (13.0) | 71.2 (13.2) |
aBMI: body mass index.
bNPRS: numeric pain rating scale.
cPAM-13: Patient Activation Measure-13.
A significant increase from baseline to last assessment in mean number of steps per day was observed for all subjects (916 [SE 14.4];
An increase in mean steps per day was observed for both groups at each visit. The percentage increase in steps per day was significantly greater for all visits in Group A versus Group B. LS mean percentage change in steps per day for Group A versus Group B at day 7 was 16.8% versus 3.1% (mean difference: 13.7%, 95% CI 0.7-26.7;
In the 6-minute walk test, Group A experienced a significantly greater improvement from baseline to day 90 in LS mean percentage change in pain versus Group B, −55.3% versus −33.8% (mean difference: −21.5%, 95% CI −37.8 to −5.2;
Steps per day. Mean change from baseline at day 90 in number and mean percentage change from baseline. Data are presented as least squares means and standard error. P values obtained from rank analysis of covariance.
Six-minute walk test. Mean percentage change from baseline to day 90 in pain during the test and distance walked. Data are presented as least squares means and standard error. P values obtained from rank analysis of covariance.
A greater number of Group A patients (68/104, 65.4%) reported they would be likely or very likely to use the devices compared with patients (36/104, 34.6%) who reported that they would be somewhat likely or not at all likely to do so (
PAM-13 scores improved from baseline to day 90 in both groups. The LS mean change from baseline was 5.0% in Group A versus 6.9% in Group B (
There were no significant changes in sleep from baseline to day 90 for either group and no significant differences between groups. Changes in VAMS scores from baseline to day 90 were also not significantly different between groups.
The occurrence of TEAEs was similar in the 2 groups (
Satisfaction survey results in patients and physicians. Only patients in Group A (n=104) and their associated physicians participated in the satisfaction survey, which was completed at day 90.
Mean change from baseline in Patient Activation Measure-13. Data are presented as least squares means and standard error.
Treatment-emergent adverse events.
Group A |
Group B |
||
Any TEAEa,b | 42 (39.3) | 38 (36.5) | |
Arthralgia | 8 (7.5) | 12 (11.5) | |
Upper respiratory tract infection | 7 (6.5) | 2 (1.9) | |
Any serious TEAEsc | 5 (4.7) | 1 (1.0) | |
Any TEAE leading to death | 0 | 0 | |
Any TEAE leading to discontinuation of Jawboned | 1 (0.9) | 0 | |
Any TEAE leading to discontinuation of OA GOd,e | 1 (0.9) | — |
aTEAE: treatment-emergent adverse events.
bAll TEAEs were mild to moderate in severity.
cSerious AEs included upper respiratory tract infection, transient ischemic attack, large intestine perforation, arthralgia and worsening OA, each in one patient in Group A and cholecystitis in one patient in Group B.
dSame patient discontinued both devices; the TEAE leading to discontinuation was worsening OA.
eOA: osteoarthritis.
Walking has been shown to have a beneficial effect on symptoms and decreases risk of functional limitations in patients with knee OA [
Use of the OA GO app provides direct feedback on mobility function and pain (whereby steps and mobility become surrogates for pain and function, assuming there are no other reasons for decreased activity), which should give a more valid and reliable report of pain over time as it does not rely on patient recall of pain experienced. In contrast, visual analog scales [
In this study, the results of the distance portion of the 6-minute walk test were not statistically different between groups. However, this assessment represents only a very brief snapshot in time; distance monitored continuously may be a more reliable measure of improvement in patient mobility. Despite these results not achieving statistical significance, those using the OA GO app ended the study with an average of more than 5500 steps per day, which approaches 6000 steps per day, a level that has been shown to cut the risk of developing functional limitation by half within 24 months [
PAM-13 scores were also not statistically different between groups in this study. The observed lack of statistical difference between groups could be due to the high mean baseline activation state of patients in this study (71.2 [SD 13.2]), which may have resulted from the selection criteria (ie, selecting relatively high functioning patients with knee OA) and may have been higher than for those excluded from the study. Indeed, the baseline activation score in this study was higher than scores for the general population in the US (61.9 [SD 21]) and populations with other chronic conditions (a US population with diabetes and with/without other comorbidities, 57.1 [
This study does have some limitations. The patients studied may not be fully representative of the general OA population with respect to gender distribution (women are at higher risk for knee OA than men [
The practical aspects of incorporating a mobile health app into treatment paradigms to improve patient mobility should be further investigated. This strategy requires access to a smartphone, which may be a financial barrier for some patients [
Overall, the results of this study provide evidence that in patients suffering from knee OA who received hylan G-F 20, additional improvement was achieved in this study, based on mobility parameters of steps per day and pain reduction, with use of a mobile app and wearable activity monitor.
Reduced mobility in patients with knee OA can be a significant issue negatively affecting quality of life and experience of pain. Increasing patients’ motivation to walk and thereby increasing their mobility may reduce these negative effects. Use of a novel smartphone app in conjunction with a wearable activity monitor provided additional improvement on mobility parameters such as steps per day and pain with walking in the 6-minute walk test in patients with knee OA who were treated with hylan G-F 20 (see
Patient experience of smartphone app use during the study.
CONSORT-EHEALTH-v1-6.
adverse events
Food and Drug Administration
least square
numeric pain rating scale
osteoarthritis
Patient Activation Measure-13
treatment-emergent adverse events
visual analog mood assessment
Sanofi Biosurgery LLC funded the study. Assistance with the writing and development of the manuscript was provided by Nicholas C Stilwell, PhD, and Evidence Scientific Solutions and was funded by Sanofi Biosurgery LLC. The authors would like to acknowledge Bruno Leroy, MD, of Sanofi who contributed to the trial concept and study design; Lynn Mcroy, MD, who contributed to the trial concept and design and OA GO app development; and Scott D Kelly, who was an employee of Sanofi when the study was conducted and contributed to data analysis. The authors would also like to thank the following investigators: Ladan Bakhtari, MD, Doctors of Internal Medicine Ltd/Clinical Research Advantage, Plano, TX; John Ervin, MD, The Center for Pharmaceutical Research, Kansas City, MO; Richard Glover II, MD, Heartland Research Associates, LLC, Newton, KS; B Matthew Hicks, MD, Fort Wayne Orthopedics, Fort Wayne IN; Jeffry Jacqmein, MD, Jacksonville Center for Clinical Research, Jacksonville, FL; Ted Jagielo, MD, Medical and Procedural Specialists of Illinois/Clinical Research Advantage, Inc, Chicago, IL; David Kirby, MD, Clinical Research Advantage, Birmingham, AL; Tracy Klein, MD, Heartland Research Associates, LLC, Wichita, KS; Louis Levy, MD, TriWest Research Associates, LLC, El Cajon, CA; Barry Lubin, MD, National Clinical Research Center, Norfolk, VA; Erich Schramm, MD, St. John’s Center for Clinical Research, Ponte Vedra, FL; Tammi Shlotzhauer, MD, Rochester Clinical Research, Rochester, NY; Michael Skyhar, MD, CORE Orthopaedic Medical Center, Encinitas, CA; Cynthia Strout, MD, Coastal Caroline Research Center, Mt Pleasant, SC; and L Tyler Wadsworth, MD, Sundance Clinical Research, Norfolk, VA.
Nebojsa Skrepnik has been a consultant for Sanofi and Ortofix. Andrew Spitzer has been a consultant for DePuy, Flexion Therapeutics, and Sanofi-Aventis; has received research support from DePuy; and has been a speaker for Sanofi-Aventis. Roy Altman has been a consultant for Cytori, DuPuy, Ferring, Flexion, Iroko, McNeil, Novartis, GlaxoSmithKline, Olatec, Pfizer, Q-Med, Rotta (MEDA), Strategic Science & Technologies, and Teva and has been a speaker for Iroko. John Hoekstra was a principal investigator in this study while employed at National Clinical Research Inc. John Stewart is an employee of Sanofi. Richard Toselli was an employee of Sanofi at the time the work was conducted and is currently an employee of Cochlear Ltd, Sydney, Australia.