Background: The prevalence of knee osteoarthritis is increasing in the aging U.S. population. The efficacy and cost-effectiveness of the use of hyaluronic acid (HA) injections for the treatment of knee osteoarthritis are debated. In this study, we assessed the utilization and costs of HA injections in the 12 months preceding total knee arthroplasty (TKA) and evaluated the usage of HA injections in end-stage knee osteoarthritis management in relation to other treatments.
Methods: MarketScan Commercial Claims and Encounters and Medicare Supplemental and Coordination of Benefits databases (Truven Health Analytics) were reviewed to identify patients who underwent TKA from 2005 to 2012. The utilization of patient-specific osteoarthritis-related health care (including medications, corticosteroid injections, HA injections, imaging, and office visits) and payment information were analyzed for the 12 months preceding TKA.
Results: A total of 244,059 patients met the inclusion criteria. Of those, 35,935 (14.7%) had ≥1 HA injection in the 12 months preceding TKA. HA injections were responsible for 16.4% of all knee osteoarthritis-related payments, trailing only imaging studies (18.2%), and HA injections accounted for 25.2% of treatment-specific payments, a rate that was higher than that of any other treatment. Patients receiving HA injections were significantly more likely to receive additional knee osteoarthritis-related treatments compared with patients who did not receive HA injections.
Conclusions: Despite numerous studies questioning the efficacy and cost-effectiveness of HA injections for osteoarthritis of the knee, HA injections are still utilized for a substantial percentage of patients. Given the paucity of data supporting the effectiveness of HA injections and the current cost-conscious health-care climate, decreasing their use among patients with end-stage knee osteoarthritis may represent a substantial cost reduction that likely does not adversely impact the quality of care.
1Pritzker School of Medicine (J.W.W.) and Department of Orthopaedic Surgery and Rehabilitation Medicine (H.S.B. and D.R.D.), Division of Biological Sciences, University of Chicago, Chicago, Illinois
The prevalence of osteoarthritis is increasing in the aging U.S. population; recent estimates indicate that approximately 46 million people in the U.S. have symptomatic osteoarthritis1. Osteoarthritis of the knee impairs quality of life and is the primary clinical indication for total knee arthroplasty (TKA)2. In 2013, more than 732,000 primary TKAs were performed in the U.S., and the annual number continues to rise3,4. The increasing number of TKAs can be attributed to the safety and cost-effectiveness of the procedure in the setting of a growing elderly population and the higher prevalence of obesity5,6.
Treatment guidelines for knee osteoarthritis suggest at least 6 months of nonoperative treatment prior to considering a TKA7. The specific nonsurgical treatments recommended by various sources have changed over time7-9. The goal of intra-articular injection of hyaluronic acid (HA) is to improve the lubrication of the knee joint by enhancing the viscoelastic properties of the synovial fluid10. It also has been suggested that HA may have anti-inflammatory and analgesic properties11. Despite these theoretical benefits, the efficacy of treatment with HA injections has been called into question. A number of meta-analyses, the earliest from more than a decade ago, have suggested that HA injections may not be effective and may actually be associated with a greater risk of adverse events12-14. The 2008 guidelines of the American Academy of Orthopaedic Surgeons (AAOS) for the nonoperative treatment of osteoarthritis of the knee stated that the evidence for the use of HA injections was “inconclusive”8. Updated AAOS guidelines in 2013 included a strong recommendation against the use of HA8. In contrast, a position statement by the American College of Rheumatology supported the use of HA injections in the management of knee osteoarthritis in patients who cannot tolerate nonsteroidal anti-inflammatory drugs (NSAIDs)15. Despite these contradictory recommendations, the costly injections continue to be administered in large numbers. In 2012, Medicare paid $207 million for HA preparations alone16. Furthermore, each HA injection generates an additional charge for its administration, thereby raising the cost and making administration lucrative for providers. The high costs to the health-care system of HA injections despite the paucity of supporting evidence for their clinical effectiveness warrants further investigation into their use.
Population-level data would be helpful to better characterize the magnitude of HA use and its role in end-stage knee osteoarthritis; to our knowledge, such data have not been previously published. The purpose of the current study was to examine a national sample of patients to assess the utilization of, and payments for, the use of HA injections in the treatment of end-stage knee osteoarthritis, which was defined as the 12 months prior to TKA. Additionally, we sought to compare the use of HA injections with that of other osteoarthritis treatments, such as corticosteroid injections, prescription narcotic and non-narcotic analgesic medications, and physical therapy, in this patient population.
Materials and Methods
This was a retrospective cohort study conducted using the MarketScan Commercial Claims and Encounters (commercial insurance) and Medicare Supplemental and Coordination of Benefits (Medicare with supplemental commercial insurance) databases (Truven Health Analytics). The databases provide de-identified, integrated, person-specific claims data for approximately 17 to 51 million individuals per year. The commercial insurance database includes health-care claims information for individuals with insurance through a commercial provider or a self-insuring employer under fee-for-service, fully capitated, or partially capitated health plans. The database for Medicare with supplemental commercial insurance includes claims information for individuals with both Medicare and commercial employer-sponsored coverage. Patients are included in the Medicare database only if they have supplemental commercial insurance in addition to Medicare. All claims reflect the coordination of benefits between the commercial insurer and Medicare such that all payments made by either entity are captured within the database. The age distribution in the database is representative of the overall Medicare population. International Classification of Diseases, Ninth Revision (ICD-9) diagnosis codes and Current Procedural Terminology (CPT)-4 codes can be identified in individual claims. The data include claims made from both inpatient and outpatient clinical encounters as well as claims for prescription medications. Information regarding durable medical equipment and certain injections can also be identified using Healthcare Common Procedure Coding System (HCPCS) codes. National Drug Codes (NDCs) are utilized to organize prescription medication claims. The NDCs specify both the type and dosage of the prescribed medication.
The databases were reviewed for subjects with a CPT-4 code for TKA (CPT-4 = 27447) from 2005 to 2012. Because TKA is the final common pathway for patients with end-stage osteoarthritis of the knee, identifying TKA patients and assessing their treatment history is a valid method of examining the management of end-stage knee osteoarthritis. To be included in the study cohort, these TKA subjects were required to have prescription medication information included in the database, continuous enrollment in the database for at least 12 months prior to TKA, and an associated diagnosis of osteoarthritis of the lower leg. Patient-specific knee osteoarthritis-related health-care information was gathered for the 12 months prior to TKA. Data collected included the utilization of, and payment information for, HA injections, analgesic medications, corticosteroid injections, imaging of the knee, non-TKA knee procedures (arthroscopies and unicondylar knee arthroplasties), durable medical equipment relating to the knee (including canes, crutches, wheelchairs, walkers, knee braces, cold therapy, and continuous passive motion devices), outpatient office evaluations, and physical and occupational therapy. Prescription analgesic medications were divided into narcotic analgesics or non-narcotic analgesics on the basis of NDC. Specific NDC, CPT-4, ICD-9, and HCPCS codes are provided in the Appendix.
Charlson comorbidity index (CCI) scores were calculated for each subject using ICD-9 codes for comorbidities, as previously described17. Mean age, percentage of female patients, CCI scores, and geographic region were compared between patients who received an HA injection in the 12 months prior to TKA (the HA group) and those who did not receive an HA injection (the non-HA group). Chi-square tests were performed for categorical variables and Student t tests were performed for continuous variables. All statistical analyses were performed with SAS software (version 9.3; SAS Institute).
All HA injection claims were recovered. We then performed descriptive statistical analysis of the HA group and calculated the mean number of injections per patient. Variations exist in how HA injections are administered. Some guidelines recommend 1 injection per week for 3 weeks, whereas others recommend once-weekly injections for 5 weeks18. The number of HA injection claims per year in the study period was calculated; these data were expressed as claims per 100,000 TKAs per year to control for variations in the numbers of TKAs performed each year.
Payment analysis was performed using payment information for each claim supplied by the databases. For individuals with Medicare with supplemental commercial coverage, all payments made for each individual claim by both the supplemental commercial insurance and Medicare were captured in this analysis. All payments were adjusted to 2012 U.S. dollars using the Medical Consumer Price Index.
Payments included were specific to knee osteoarthritis-related health care. For example, if an osteoarthritis-related procedure occurred while the patient was admitted as an inpatient, only the osteoarthritis-related claims payments were included; other hospital fees with ICD-9 and CPT-4 codes unrelated to osteoarthritis were excluded. Payment information related to HA and corticosteroid injections included payments for the administration of the injection as well as for the medication. Payments associated with HA injections were compared with those associated with other forms of knee osteoarthritis-related health care to assess the role the injections played in the overall cost of care received in the 12 months prior to TKA.
Association of HA Utilization with Other Knee Osteoarthritis-Related Health Care
Further analysis was conducted to assess the use of osteoarthritis-related treatment modalities in the HA and non-HA groups. The groups were analyzed separately to evaluate utilization of other forms of knee osteoarthritis treatment, specifically, corticosteroid injections, prescription narcotic and non-narcotic analgesic medications, and physical therapy. Multivariate logistic regression analysis was performed to compare the prevalence of these interventions in the study period in the HA group compared with the non-HA group. The model controlled for the effects of age, sex, insurance status, CCI score, and region.
A total of 244,059 subjects met the inclusion criteria; 55.7% were covered under commercial insurance and 44.3%, under Medicare with a commercial supplement (Fig. 1). Overall, 61.2% of the study population was female, and 38.8% was male. A total of 35,935 patients (14.7% of the study population) had ≥1 claim for HA injection in the 12 months preceding TKA. A significantly greater percentage of the HA group (64.5%) was female compared with the non-HA group (60.7%) (p < 0.001). Significant differences between the groups were also demonstrated for age, insurance status, and CCI score (Table I); however, these differences did not appear to be clinically important.
HA Injection Utilization
Among the 14.7% of the study population who received ≥1 HA injection in the 12 months preceding TKA, the mean number of injections received was 3.6. Between 2004 and 2012, the number of HA injections per 100,000 patients in the study population ranged from a low of 24,030 in 2004 to a high of 30,914 in 2008 (Fig. 2).
HA Payment Analysis
Total payments associated with HA injections in the study cohort amounted to $40,547,881 over the study period. The mean payment per individual HA injection was $310, and the mean total payment for HA injections per patient receiving an HA injection (all HA injections for a patient) was $1,128. HA injections accounted for 16.4% of all knee osteoarthritis-related health-care payments in the study population. HA injections ranked as the second highest knee osteoarthritis-related health-care expenditure, behind only imaging (knee radiography and magnetic resonance imaging [MRI]), which accounted for 18.2% of payments (Fig. 3-A). Outpatient office evaluations and prescription non-narcotic analgesic medications accounted for 16.0% and 14.9% of payments, respectively. Corticosteroid injections and non-TKA knee procedures represented 11.9% and 11.6% of payments. Next were prescription narcotic analgesic medications, durable medical equipment, and physical and occupational therapy, at 6.3%, 2.9%, and 2.0% of payments, respectively.
When focusing on payments specific to treating knee osteoarthritis (excluding office visits and imaging, which are evaluative expenses), HA injections emerged as the largest expenditure, accounting for 25.2% of payments (Fig. 3-B). Prescription non-narcotic analgesic medications (22.9%) and corticosteroid injections (18.2%) ranked as second and third, respectively.
Multivariate Analysis to Characterize the Role of HA in Patient Care
Multivariate logistic regression analysis controlling for the effects of age, sex, insurance status, CCI score, and region showed that patients in the HA group demonstrated an increased likelihood of also having received other knee osteoarthritis-related care (Table II). Patients in the HA group were 4.2 times more likely to have also received ≥1 corticosteroid injection in the study period than those in the non-HA group (odds ratio [OR] = 4.172; 95% confidence interval [CI] = 4.073 to 4.274). Patients in the HA group were also 2.2 times more likely to have received ≥1 prescription for a narcotic analgesic medication (OR = 2.178; 95% CI = 2.128 to 2.230), 2.5 times more likely to have received ≥1 prescription for a non-narcotic analgesic medication (OR = 2.545; 95% CI = 2.467 to 2.582), and 2.2 times more likely to have had physical therapy (OR = 2.231; 95% CI = 2.172 to 2.292).
Despite the paucity of data supporting the efficacy of HA injections, they are still widely administered, as demonstrated by the rate in the present study, in which 14.7% of patients received HA injections within the 12 months preceding TKA. Patients treated with HA injection received an average of 3.6 injections during this time period. This may be due to existing guidelines, which recommend 3 to 5 weekly injections in a single treatment course18. We found that HA injections were more likely to be administered in the 12 months preceding TKA to patients who are female (64.5% of the HA group compared with 60.7% of the non-HA group; p < 0.001), who are younger (mean age of 64.7 years compared 64.9 years; p = 0.002), and who have more comorbid conditions (mean CCI score of 0.060 compared with 0.043; p < 0.001) (Table I). Although the differences noted in age and CCI score were significant, the small difference in absolute values observed is likely of little clinical importance.
Patients who received an HA injection were significantly more likely to receive additional knee osteoarthritis-related treatments, being over 4 times more likely to receive a corticosteroid injection (OR = 4.172), and approximately twice as likely to be prescribed narcotic analgesic medication (OR = 2.178), non-narcotic analgesic medication (OR = 2.545), or physical therapy (OR = 2.231) compared with the non-HA group (Table II). These results indicate that patients receiving HA injections prior to TKA are, generally speaking, high utilizers of treatments directed at osteoarthritis of the knee compared with patients not receiving HA injections. It also indicates that HA treatments, rather than replacing other knee osteoarthritis-related treatments, appear to be used in addition to other forms of treatment. Whether this is because of these patients having more advanced disease, because of a lack of effectiveness of HA injections, or because these patients have better access to care cannot be determined from this study and warrants investigation.
Although this study does not permit an assessment of the effectiveness of HA injections, the study data indicate that HA injections are still highly utilized, despite the findings of prior studies questioning their effectiveness as well as clinical guidelines published by the AAOS discouraging their use. As shown in Figure 2, the rate of utilization of HA injection remained relatively stable during the study period. The high utilization of HA injection also results in substantial health-care costs. HA injections represented 16.4% of all knee osteoarthritis-related health-care payments in this patient population, second only to imaging studies (18.2%). When evaluative expenses were excluded and the analysis was focused specifically on treatment expenses, HA injections emerged as the single largest expense in the study population, accounting for 25% of payments. Furthermore, intra-articular corticosteroid injections, which are an established treatment modality for symptomatic knee osteoarthritis, only represented 11.9% of total payments and 18.2% of treatment-specific payments. Additionally, physical therapy, an effective, relatively low-cost option, appeared to be underutilized (only 15.1% of the patients had a claim for physical therapy in the 12 months prior to TKA) and accounted for only a small fraction of the osteoarthritis-related health-care cost (1.8% of treatment-specific payments). These results highlight that, despite the controversy surrounding the efficacy of HA injections, they continue to constitute a substantial portion of payments in the treatment of end-stage knee osteoarthritis and are more costly than all other currently utilized treatment modalities. While it is possible that HA injections could be beneficial for a certain subset of patients with specific phenotypes, we are unaware of any studies that have been able to successfully identify such an individual profile. A decrease in the use of HA injections in end-stage knee osteoarthritis could reduce the overall health-care cost burden, potentially without causing an adverse effect on outcomes.
The study utilized Truven Health Analytics MarketScan Commercial Claims and Encounters and Medicare Supplemental and Coordination of Benefits databases, which allow for large sample sizes but also have inherent limitations. Although the databases are likely a representative sample of the overall population, they do not include patients without insurance or on Medicaid, or claims that may have been filed under a different system, such as Veterans Affairs. Additionally, claims information in the databases does not designate the laterality of the injections or of the eventual TKA. It is possible that treatments under study were designated for the knee contralateral to the knee of interest, or, in the case of analgesic prescriptions, for diseases other than osteoarthritis of the knee. However, this is unlikely to have had a major impact on our findings given that the study included 244,059 patients and only 3% of the study population had a subsequent TKA claim in the 6 months following the index TKA. Therefore, it is reasonable to assume that the osteoarthritis-related expenses evaluated pertained mostly to the joint that was eventually replaced. The databases only contain claims information and do not include any knee function or pain scores; as such, we cannot comment on the patient-related outcomes of HA injection but merely on the utilization of other modalities in addition to HA injection. The study design did not permit the determination of who prescribed the HA injections or other treatments (e.g., the primary care physician or an orthopaedic surgeon). It is possible that some of the narcotic and non-narcotic analgesic medications were prescribed for conditions other than knee osteoarthritis. However, in such a large population of patients sampled for 12 months prior to undergoing TKA, it is unlikely that a substantial percentage of analgesic medications were for another source of pain needing treatment. Additionally, if some of the analgesic medications captured in this study were prescribed for conditions other than knee osteoarthritis, then the study findings would underestimate the true percentage of payments represented by HA injections, thereby strengthening the assertion that HA injections represent a substantial expense in this patient population. Lastly, our study examined patients with severe knee osteoarthritis symptoms (all patients underwent TKA within 12 months); therefore, the findings of this study may not be generalizable to the overall population.
Our study highlights the prevalence and expense associated with HA injections in patients on the path to TKA. Previous studies have questioned the clinical benefit of HA injections8,12-14. Despite this, the results of this study indicate that use of HA injections represents a substantial direct cost to the health-care system—second only to imaging studies—and the largest treatment-specific expense. Furthermore, increased utilization of overall resources was found among the patients who received HA injections compared with patients who did not, indicating that HA may have added to, rather than replaced, other treatment modalities and expenses. In the current health-care climate, with greater emphasis on cost, reducing the use of HA injections in patients with end-stage knee osteoarthritis may represent an easily implementable means to substantial cost savings. Additional investigation will better determine if this can be done without adverse effects on patient outcomes.
A table showing the codes and drug information used in the identification of knee osteoarthritis-related health care is available with the online version of this article as a data supplement at jbjs.org.
Investigation performed at the Department of Orthopaedic Surgery and Rehabilitation Medicine, Division of Biological Sciences, University of Chicago, Chicago, Illinois
Disclosure: No external funding was received for this investigation. The Disclosure of Potential Conflicts of Interest forms are provided with the online version of the article.
1. Helmick CG, Felson DT, Lawrence RC, Gabriel S, Hirsch R, Kwoh CK, Liang MH, Kremers HM, Mayes MD, Merkel PA, Pillemer SR, Reveille JD, Stone JH; National Arthritis Data Workgroup. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part I. Arthritis Rheum. 2008 ;58(1):15–25.
2. Nguyen USDT, Zhang Y, Zhu Y, Niu J, Zhang B, Felson DT. Increasing prevalence of knee pain and symptomatic knee osteoarthritis: survey and cohort data. Ann Intern Med. 2011 ;155(11):725–32.
4. Losina E, Thornhill TS, Rome BN, Wright J, Katz JN. The dramatic increase in total knee replacement utilization rates in the United States cannot be fully explained by growth in population size and the obesity epidemic. J Bone Joint Surg Am. 2012 ;94(3):201–7.
5. NIH Consensus Panel. NIH Consensus Statement on total knee replacement December 8-10, 2003. J Bone Joint Surg Am. 2004 ;86(6):1328–35.
6. Daigle ME, Weinstein AM, Katz JN, Losina E. The cost-effectiveness of total joint arthroplasty: a systematic review of published literature. Best Pract Res Clin Rheumatol. 2012 ;26(5):649–58.
7. Zhang W, Moskowitz RW, Nuki G, Abramson S, Altman RD, Arden N, Bierma-Zeinstra S, Brandt KD, Croft P, Doherty M, Dougados M, Hochberg M, Hunter DJ, Kwoh K, Lohmander LS, Tugwell P. OARSI recommendations for the management of hip and knee osteoarthritis, part II: OARSI evidence-based, expert consensus guidelines. Osteoarthritis Cartilage. 2008 ;16(2):137–62.
9. Van Manen MD, Nace J, Mont MA. Management of primary knee osteoarthritis and indications for total knee arthroplasty for general practitioners. J Am Osteopath Assoc. 2012 ;112(11):709–15.
10. Ghosh P, Guidolin D. Potential mechanism of action of intra-articular hyaluronan therapy in osteoarthritis: are the effects molecular weight dependent? Semin Arthritis Rheum. 2002 ;32(1):10–37.
11. Strauss EJ, Hart JA, Miller MD, Altman RD, Rosen JE. Hyaluronic acid viscosupplementation and osteoarthritis: current uses and future directions. Am J Sports Med. 2009 ;37(8):1636–44. Epub 2009 Jan 23.
12. Lo GH, LaValley M, McAlindon T, Felson DT. Intra-articular hyaluronic acid in treatment of knee osteoarthritis: a meta-analysis. JAMA. 2003 ;290(23):3115–21.
13. Arrich J, Piribauer F, Mad P, Schmid D, Klaushofer K, Müllner M. Intra-articular hyaluronic acid for the treatment of osteoarthritis of the knee: systematic review and meta-analysis. CMAJ. 2005 ;172(8):1039–43.
14. Rutjes AWS, Jüni P, da Costa BR, Trelle S, Nüesch E, Reichenbach S. Viscosupplementation for osteoarthritis of the knee: a systematic review and meta-analysis. Ann Intern Med. 2012 ;157(3):180–91.
15. Hochberg MC, Altman RD, April KT, Benkhalti M, Guyatt G, McGowan J, Towheed T, Welch V, Wells G, Tugwell P; American College of Rheumatology. American College of Rheumatology 2012 recommendations for the use of nonpharmacologic and pharmacologic therapies in osteoarthritis of the hand, hip, and knee. Arthritis Care Res (Hoboken). 2012 ; 64(4):465–74.
16. Schmajuk G, Bozic KJ, Yazdany J. Using Medicare data to understand low-value health care: the case of intra-articular hyaluronic acid injections. JAMA Intern Med. 2014 ;174(10):1702–4.
17. Quan H, Sundararajan V, Halfon P, Fong A, Burnand B, Luthi JC, Saunders LD, Beck CA, Feasby TE, Ghali WA. Coding algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Med Care. 2005 ;43(11):1130–9.
Copyright 2016 by The Journal of Bone and Joint Surgery, Incorporated
18. Wen DY. Intra-articular hyaluronic acid injections for knee osteoarthritis. Am Fam Physician. 2000 ;62(3):565–70: 572.