Keywords
negative pressure wound therapy
Issue: Volume 64 - Issue 1 - January 2018 ISSN 1943-2720
Index: Ostomy Wound Manage. 2018;64(1):26-33. doi: 10.25270/owm.2017.64.2633
Login or Register to download PDF
Abstract
Traditional negative pressure wound therapy (NPWT) systems are considered durable. The pump is designed for use by numerous patients over a period of several years. Recently developed smaller, disposable devices are designed for single-patient use. A retrospective analysis of 2012–2014 national Medicare claims data was used to examine payments associated with the use of traditional and disposable NPWT systems.Data extracted included NPWT episodes from the Limited Data Set Standard Analytic Files including the 5% sample for traditional NPWT and 100% sample for disposable NPWT. NPWT episodes were identified using claim service dates and billing codes. Mean costs per episode were compared and analyzed using chi-squared tests for comparisons between patients who received traditional and those who used disposable NPWT. For continuous variables, statistical significance was assessed using Mann-Whitney U tests. The data included traditional (n = 2938; mean age 66.6 years) and disposable (n = 3522; mean age 67.6 years) episodes for the 2 NPWT groups. Wound types differed for NPWT groups (P <.0001) and included surgical (1134 [39%] versus 764 [22%]), generic open (850 [29%] versus 342 [10%]), skin ulcers (561 [19%] versus 1301 [37%]), diabetic ulcers (240 [8%] versus 342 [10%]), and circulatory system wounds (105 [4%] versus 563 [16%]). Average payment amounts were $4650 ± $2782 for traditional and $1532 ± $1767 per disposable NPWT episode (P <.0001). Payment differences were not affected by wound or comorbidity characteristics. Using the 2016 rates, average payments were $3501 for traditional and $1564 for disposable NPWT. Considering the rate of NPWT use in the United States and the results of this study suggesting substantial potential cost savings, additional analyses and cost-effectiveness studies are warranted.
Introduction
According to estimates, at least $33 billion is spent annually in the United States (US) on direct medical costs of chronic wounds, including pressure ulcers, diabetic foot ulcers, and venous leg ulcers.1-3 This estimate understates the total cost of wound care because it does not include the cost of surgical, traumatic, or other nonhealing open wounds; indirect costs; and lost productivity. Approximately 6.5 million Americans or 2% of the general population suffer from chronic wounds4; Medicare-covered older adults account for the majority. The cost to treat wounds will increase as the US population ages and chronic diseases such as diabetes become more prevalent. As a result, wound care imposes a significant cost and morbidity burden on patients, their families, and the health care system.
Negative pressure wound therapy (NPWT) is commonly used to treat a variety of chronic and acute wounds.5-7 Retrospective analyses of data obtained from clinical trials8,9 as well an outpatient wound care setting10 indicate NPWT is more cost-effective than non-NPWT care in treating chronic wounds of various etiologies, including diabetic foot ulcers, venous leg ulcers, pressure ulcers, arterial ulcers, and surgical wounds.8-10 In these studies, the cost savings for those receiving NPWT versus standard wound care was attributed to improved clinical outcomes (more patients achieving complete wound closure and shorter time to wound closure), which in turn resulted in shorter duration of treatment and fewer medical/surgical procedures and reduced the number of hospitalizations, days hospitalized, and clinic visits.
Traditional NPWT systems are considered durable medical equipment (DME) and consist of an electric suction pump to generate negative pressure, canisters to collect wound exudate, and wound dressing kits. Traditional systems are “durable” because the pump is designed for several years of use with numerous individual patients. A more recent innovation in NPWT is smaller, disposable devices designed for single-patient use. These single-use or disposable systems are predicated on the methods of action of traditional NPWT systems11-14 and designed to deliver equivalent outcomes. Furthermore, a case series14 (N = 326) conducted among patients in a community care setting showed disposable NPWT simplifies application and management for health care professionals and makes the therapy more accessible to patients, including more active and home-bound individuals. According to that study, more than 90% of patients who were treated with single-use, disposable devices were able to perform their daily activities unimpeded, and 97% were pleased or satisfied with the therapy.
The cost of NPWT use is a significant concern to the Medicare program. As a result, payment and coverage rules for NPWT continue to evolve. Historically, when a home health agency (HHA) provided a beneficiary traditional NPWT, Medicare made a payment to the HHA for the visit and a monthly payment to a DME supplier for the NPWT system. However, if a HHA used disposable NPWT (not provided during another facility visit such as a hospital-based outpatient department [HOPD]) or inpatient discharge) Medicare only made a visit payment to the HHA, which then absorbed the cost of the disposable NPWT.
One way to control the cost of NPWT was through the competitive bidding program implemented by the Centers for Medicare and Medicaid Services (CMS). The competitive bidding program is a process where DME providers submit bids for providing certain DME in bidding areas for Medicare beneficiaries. From these bids, Medicare sets the payment rates and secures savings for the program. In a recent development, Congress acted to encourage access to disposable NPWT and required Medicare to make a new, separate payment to HHAs for disposable NPWT systems in addition to the normal HHA visit payment starting January 1, 2017.15,16
With this in mind, the authors hypothesized that disposable NPWT systems might present a new option to control Medicare expenditures. To date, limited research has been published identifying the potential cost savings of disposable NPWT devices to payers.17,18 The objective of this retrospective cost-minimization analysis was to assess Medicare payments associated with traditional and disposable NPWT for Medicare beneficiaries from national Medicare claims data.
Methods
Data source. Medicare is the federal health insurance program for people 65 or older, certain younger people with disabilities, and people with end-stage renal disease. Medicare has 4 distinct parts (A, B, C, and D) to help beneficiaries cover certain health care services; Parts A and B are of consequence to this analysis. Medicare Part A covers inpatient hospital stays, skilled care in a skilled nursing facility up to 100 days, medically necessary hospice care, and some HHA care. Part B covers certain Qualified Health Plan services and procedures, HOPD care, some medical supplies, and preventive services.
This analysis used the 2012–2014 national Medicare Limited Data Set Standard Analytic Files, including Denominator (enrollment), hospital inpatient and outpatient, DME, carrier (physician), and HHA claims. The 5% sample data (nationally representative sample of beneficiaries) was used to study traditional NPWT episodes, while the 100% sample (all Medicare beneficiaries) was used for disposable NPWT episodes to increase the number of disposable episodes. These files are claims data sets that contain information on care provided to Medicare beneficiaries and on patient characteristics such as age, gender, enrollment history, medical diagnoses, dates and places of service, and payment amounts. Claims data are based on the bills submitted by providers to payers representing services provided, provider setting of care, and patient comorbidities. Although claims data often are utilized to approximate cost of service, the cost identified is not direct medical cost.
Construction of episodes of care. NPWT episodes were identified from the data using service dates on Medicare claims. Traditional NPWT patients were identified using the Current Procedural Terminology/Healthcare Common Procedure Coding System (CPT/HCPCS) codes A6550, A7000, E2402, 97605, and 97606; disposable NPWT patients were identified with the CPT/HCPCS codes A9272, G0456, G0457, 97607, and 97608 (see Table 1). Medicare-denied NPWT claims and home health claims not directly related to nursing care were excluded from analysis.
For traditional NPWT, episodes began with the first observed DME monthly pump rental claim (HCPCS E2402) and continued until a break of >31 days occurred between rentals. To ensure the episode of care was followed to its completion, a break of 31 days was used based on DME monthly pump rental policies. The episode end date was estimated beyond the last pump rental by assuming all supplies were used at the standard Medicare allowed use rates of 15 dressings (HCPCS A6550) and 10 canisters (HCPCS A7000) per month. For disposable NPWT, episodes began with the first observed HOPD NPWT claim (HCPCS G0456/G0457 during this period) and continued until a break of >31 days occurred between claims. The same 31 day cutoff as based on traditional NPWT monthly pump rental policies was used to maintain consistency in evaluating the end of disposable NPWT episodes. For this analysis, neither group episode payment analysis started in the acute inpatient setting.
The episode end date for paid claims (ie, nondenied) was estimated using 1 of 2 methods to ensure the full length of discrete episodes was captured. First, for episodes with more than 1 HOPD NPWT bill, the average interval between device changes (for that episode) was added to the last HOPD NPWT bill. Second, for episodes with just 1 HOPD NPWT bill, the average device change interval for all patients in the data file was added. Patients with disposable or traditional NPWT episodes were mutually exclusive.
Episode costs (total payments) were constructed from relevant wound diagnosis and treatment claims on the billing forms within the episode and included DME, physician, or HOPD paid claims lines with any relevant CPT/HCPCS codes (see Table 1). Cost was defined as total payment including both Medicare program payments and beneficiary coinsurance and deductible liabilities. To ensure a proportion of the HHA bill associated with NPWT was captured and arrived at a more accurate episode cost, the portion of HHA bills that overlapped the NPWT episode was prorated. Because HHA bills do not have line-item payment amounts, HHA claim total payments were prorated based on charges, excluded line items other than nursing visits (eg, physical therapy), and then selected all nursing visits falling within the NPWT episode. Using this method, some HHA cost not directly related to NPWT may have been captured, but the principal diagnoses on HHA claims overwhelmingly indicated wound care as the main reason for the HHA care.
Risk adjustment and sensitivity analyses. Several analyses were performed to test the robustness of the total mean payments results. First, the influence of wound type and comorbidities on episode cost was examined. Episodes of care were stratified by episode type (traditional versus disposable) and wound type (surgical wounds, generic open wounds, skin ulcers, diabetic ulcers, and circulatory disease wounds). Mean cost within strata were calculated, and the means were weighted by the all-episodes proportion of cases by wound type. Episode costs also were assessed for different wound types.
Second, ordinary-least-squares (OLS) regression was used to predict episode costs as a function of episode type, patient demographics, and diagnoses from the claims (diagnosis-based risk adjustment). Comorbidities were identified using diagnoses reported on 12 months of physician and hospital claims, aggregated into Clinical Classification System disease categories.19
Third, aspects of Medicare payment policy were handled. The presence of traditional NPWT qualifies any homebound patient for home health care; all HHA costs and recomputed episode costs were removed. Medicare updates its payment rates on an annual basis through a rulemaking process to reflect updated costs and evolving payment policies. Similarly, Medicare adjusts DME fee schedule rates through its competitive bidding process among DME providers. Therefore, because Medicare payment rates for DME items changed in response to its recent competitive bidding initiatives, all claims were priced using 2016 Medicare rates20,21 and recomputed episode costs.
Statistical analyses. For categorical variables, statistical significance was assessed using chi-squared for comparisons between patients who used traditional and patients who used disposable NPWT. For continuous variables, statistical significance was assessed using Mann-Whitney U tests. The claims data analyses were conducted using Statistical Analysis System (SAS) software, version 9.4 (SAS Institute, Cary, NC).
Results
The Medicare claims database included 2938 traditional NPWT episodes and 3522 disposable NPWT episodes (see Table 2). Differences in demographics between the cohorts were identified, including a difference in mean age of patients (66.6 versus 67.6, P = 0.001) and race (2350 [80%] versus 2923 [83%] Caucasian, P = 0.0021) for traditional and disposable NPWT, respectively. Differences in the mix of wounds were identified between the groups (P <.0001) and included surgical wounds (1134 [39%] versus 764 [22%]), generic open wounds (850 [29%] versus 342 [10%]), skin ulcers (561 [19%] versus 1301 [37%]), diabetic ulcers (240 [8%] versus 342 [10%]), circulatory system disease wounds (105 [4%] versus 210 [6%]), and all others (48 [2%] versus 563 [16%)], for traditional and disposable NPWT, respectively (see Table 2).
Total direct costs associated with each patient’s traditional NPWT episode were 3 times greater than total costs associated with each patient’s disposable NPWT episode ($4650 ± $2782 versus $1532 ± $1767, respectively, P = .0001; see Table 3). Device costs dominate both estimates; traditional NPWT pump rental, wound dressings, and exudate canisters represented mean costs of $2390 ± $1406, $499 ± $301, and $104 ± $61, respectively (see Table 3). Pump rental was the single highest cost item for traditional NPWT ($2390 ± $1406), while device replacement (including labor) in the HOPD was the largest single item for disposable NPWT ($1076 ± $1436; see Table 3). Necessarily, disposable NPWT does not have DME costs as outlined previously for traditional NPWT, but costs were identified and compared in the HHA ($1553 ± $1422 versus $456 ± $865; P = .0001) and HOPD ($104 ± $284 versus $1076 ± $1436; P = .0001) for traditional and disposable NPWT, respectively. Mean cost per day was estimated to be $142 ± $203 and $56 ± $33 for traditional NPWT and disposable NPWT, respectively (see Figure). A difference in the mean episode lengths of 43.3 days and 28.3 days was observed for traditional and disposable NPWT, respectively. Finally, as a result of dissimilar product treatment paths, additional differences were identified in days of service for HOPD and HHA services (1.5 versus 4.9 days, P = .0001) and (13.5 versus 3.9 days, P= .0001; see Table 4) for traditional and disposable NPWT, respectively. For traditional NPWT, dressing changes are frequently paid as HHA services while HOPD dressing changes are less common. By contrast, essentially all disposable device replacements occur in the HOPD.
Table 5 provides the results of the sensitivity analyses that assessed whether patient characteristics, Medicare HHA payment rules, or periodic Medicare payment rates changes can explain or eliminate the observed difference in cost. As estimated, the cost of traditional NPWT was 3 times that of disposable NPWT. The first 3 analyses show patient characteristics (2 on wound type and 1 on comorbidities plus demographics) do little to explain the cost difference between traditional and disposable NPWT. The weighted analysis holding the proportion of wound types constant showed the total mean payment per episode of care for traditional NPWT was 3.1 times more than disposable NPWT ($4837 versus $1567 for wound types; see Table 5). Similarly, when costs were assessed by wound type, the cost difference between traditional NPWT and disposable NPWT remained close to the ratio of 3 observed in the primary analysis, ranging from 2.4 for open wounds to 3.7 for skin ulcers (see Table 5). Even though OLS regression indicated 18% of episode cost variation (adjusted R-squared) and only 1 comorbidity (paralysis) had a statistically significant effect in excess of $1000 per episode, the cost of traditional NPWT remained approximately 3 times greater than disposable NPWT after adjusting for comorbidities plus demographics ($4670 versus $1511; see Table 5). The fourth sensitivity test removed HHA costs for homebound patients to determine if it affected the payment delta ($3097 versus $1076; see Table 5). Finally, Medicare’s substantial reductions in DME rates (mainly NPWT pump rental) and modest increases in HOPD rates reduced the initial 3 to 1 ratio to 2.2 to 1 (from $4837 versus $1567 to $3501 versus $1564; see Table 5).
Traditional NPWT episodes (N = 2938 based on the 5% LDS; see Table 2) were significantly more common than disposable NPWT episodes (N = 3522 based on the 100% sample; see Table 2). Given the current level of traditional NPWT use (N = 2938 multiplied by 20 equals approximately 60 000 episodes; see Table 2), and the cost savings identified after adjustment to 2016 rates ($1937; see Table 5), a potential estimated Medicare cost savings of more than $1 billion would be realized over the next decade if traditional NPWT was replaced by disposable NPWT.
Discussion
The purpose of this study was to examine the cost of NPWT use from a Medicare payment policy perspective. Although the authors are not aware of other research on cost from the Medicare payment perspective, previously published studies8-10 from the provider perspective report NPWT can reduce the cost of wound care. Two (2) of these studies8,9 assessed the cost of traditional NPWT versus standard moist wound therapy (not otherwise specified) using results obtained in randomized clinical trials. Apelqvist et al8 analyzed data from 162 patients with diabetes with postamputation wounds and found the average direct cost per patient was $27 270 versus $36 096 for the NPWT and standard therapy groups, respectively. Driver and Blume9 analyzed data from 324 patients with grade 2 or 3 diabetic foot ulcers and found mean cost per patient was $11 984 and $13 558 for NPWT and advanced moist wound therapy groups, respectively. A third study10 examined outpatient data from 304 patients with either diabetes or peripheral artery disease and multiple comorbidities likely to interfere with wound healing. The cost-effectiveness analysis showed an incremental net health benefit of $1371 per ulcer-free month and an incremental cost-effectiveness ratio of $366 682 per quality adjusted life year gained in the NPWT group.
A limited number of studies17,18 has examined costs associated with the use of disposable NPWT. One (1) was a case study17 of 21 patients with various postoperative or posttraumatic wounds of which 8 patients would have been hospitalized based on hospital surgery department treatment practice if disposable NPWT was not available. The cost savings for treating these patients with disposable NPWT was estimated to be £7800 per patient based on hospital costs averted. A decision analytic model study18evaluated the cost-effectiveness of disposable NPWT in patients undergoing primary hip and knee replacement surgery. Data from 209 patients were analyzed and showed an estimated cost savings for health care payers in the United Kingdom of £1132 per patient for those treated with disposable NPWT compared with standard care.
Unlike the previously published studies described that focus on cost savings from the providers’ perspective, the current study examines costs from the payers’ perspective. Specifically, Medicare payments for traditional NPWT and disposable NPWT episodes of care were compared using data extracted from the national Medicare database. The analysis demonstrated a considerable opportunity for Medicare (and by extension, private payers) to substantially lower costs for NPWT episodes while embracing the use of more patient-friendly disposable technology. The analysis demonstrated the average risk-adjusted cost for Medicare to treat a post-acute patient with traditional NPWT was $3118 more than disposable NPWT ($4650 versus $1532); the cost of pump rental for traditional NPWT and the cost of the device for disposable NPWT dominate each group’s costs. It is important to note the large difference in cost is a result of Medicare payment policy for traditional versus disposable NPWT episodes. Medicare’s payment policy is a significant contributor to the cost difference identified in this research. Patients treated with traditional NPWT in the HOPD, clinic, or home by a HHA use electrical pumps and supplies and receive services from trained nurses, physical and occupational therapists, and physicians. For Medicare patients treated with traditional NPWT, Medicare reimburses the DME supplier for the pump, the HOPD for any services of qualified health care professionals (MD, DO, DPM, NP, PA, CNS) for care, and a HHA for their assessment of the wound, application of the NPWT, and education of the patient or caregiver. In addition, large differences in cost remain after accounting for patient characteristics, possible bias from HHA costs, and evolving DME and HOPD payment changes (since 2013). These results seem reasonable given that the episodes, by definition, are limited to ≤3 months and the disposable systems may be suitable for almost all of the wounds that can be treated with traditional NPWT systems. This analysis, based on similar ages and across various wound types, demonstrates that payers can capture significant savings by encouraging the use of disposable NPWT.
Because the claims data showed surgical wounds were more prevalent among traditional NPWT users and skin ulcers were more prevalent among disposable NPWT users, the authors performed a sensitivity analysis based on wound type. Results of the analysis indicated wound type had a negligible impact on the estimated payment difference between traditional and disposable NPWT therapy. To further test the impact of wound type on payment, the authors performed an analysis holding the proportion of wound types constant. Results of the weighted analysis also showed the observed differences in the types of wounds were unlikely to affect the payment difference between traditional and disposable NPWT technologies. However, it should be noted that the differences in wound types observed may be an artifact of coding practices because DME claims are more likely to report nonspecific (eg, not otherwise specified) diagnosis codes with traditional NPWT cases compared to specific codes for other wounds identified more frequently with disposable NPWT cases. Although the current analysis showed a difference in mean episode lengths for traditional and disposable NPWT, the authors are not aware of any literature demonstrating shorter treatment episodes for disposable cases, a finding that may be an artifact of their analysis.
Despite the substantial cost advantage of disposable NPWT devices, these devices are infrequently provided to Medicare beneficiaries (approximately 6% of all NPWT claims in this analysis were for disposable NPWT devices). A possible explanation for underutilization may be the silo nature of traditional Medicare payments. As such, traditional Medicare separate per-service payment to the prescribing physician, HOPD, and DME supplier may insulate each player from the other regarding the cost implications of their choice of NPWT modality. Another possible explanation for the limited use of these devices on Medicare beneficiaries is the relative newness of disposable NPWT as a treatment alternative to traditional NPWT. As awareness of this new option develops, evolving Medicare payment policy may have a tremendous impact on the adoption of disposable NPWT. Medicare’s new initiatives implementing value- and episode-based payments should destroy silos and motivate providers through payment tied to quality reporting, costs, outcomes, and patient satisfaction.
In 2015, the CMS set aggressive targets of having 90% of all Medicare fee-for-service payments tied to quality reporting; by 2018, 50% of all payments will be part of an “alternative” model (ie, bundled payments).22 The authors speculate that value- and episode-based payments will hasten the adoption of disposable NPWT over traditional NPWT. Additionally, Congress passed a consolidated appropriations act in 2015 that included a provision supporting the adoption of disposable NPWT technologies.15 Starting January 1, 2017, Medicare began making a new, separate payment to HHAs that furnish Medicare beneficiaries with disposable NPWT. Given the current level of traditional NPWT (DME) use (approximately 60 000 episodes per year as identified in the Medicare 5% sample) and the per episode cost savings of $1937 identified here after adjustment to 2016 rates, the estimated potential cost savings could be more than $1 billion over the next decade if traditional NPWT is replaced by disposable NPWT where clinically appropriate.
However, realizing those savings will likely require action by Medicare. Disposable NPWT systems offer policymakers a new option to reduce expenditures, limit waste (both bulk supply waste and unnecessary cost), and improve the quality of life for patients. In isolation, competitive bidding has generated some reductions in payment for traditional NPWT to date but not enough to bend the NPWT cost curve. The authors expect Medicare’s initiatives in value- and episode-based payment will spur adoption of the more cost-effective technology within those episodes. However, for the bulk of Medicare claims, current payment policies offer limited incentive for adoption of the lower cost technology.
Limitations
The current analysis was based on claims data and therefore subject to traditional limitations of claims data sets (ie, coding errors, absence of detailed wound etiology, and limited diagnostic information). Claims data do not reflect the direct cost of products and services and may not reflect provider cost; instead, the data reflect payments made to the provider(s) based on Medicare payment policy. All episodes start with a home health or hospital outpatient claim that may understate episode lengths for persons starting NPWT treatment before discharge from an acute inpatient facility. A further important limitation of claims data is accurately knowing when a patient completes his/her course of NPWT and whether NPWT was stopped because the goal of care was met. Clinical outcomes are not included in the claims data.
The overall purpose of the research was to compare Medicare payments and to do this the authors ended treatment episodes when patients had a 31-day break with no NPWT bills. The findings of this research are specific to Medicare patients, Medicare payment policy, and patients in postacute settings of care. The reader should decide the applicability of the findings to other settings of care or other payer groups. The statistical significance values presented may be affected by the sample sizes in the study and should be interpreted with caution. The authors acknowledge that disposable NPWT systems may not be appropriate for all wounds and that a clinician’s assessment of a particular wound should determine treatment choice. Although efforts were made to compare similar patients, the analysis did not attempt to match cohorts of patients. The analysis did not control for patients that may have had more than 1 treatment episode separated by 31 days of no bills. Instead, the authors relied on varying sensitivity analyses to determine conditions that may or may not impact the effect of Medicare payment policy. Confounding bias is a risk of this analysis, although the authors attempted to control for extraneous reasons for the ultimate result of significant cost difference. Despite the limitations and assumptions made, the strength of this analysis is the simplicity of using Medicare’s own data to approximate the average payment difference based on Medicare’s payment policies for patients prescribed traditional and disposable NPWT systems.
Conclusion
This retrospective cost-minimization analysis of 2938 traditional NPWT episodes and 3522 disposable NPWT episodes of use in the 2012–2014 national Medicare paid claims database demonstrated that using disposable NPWT resulted in much lower total episode payments than using traditional NPWT. On average, disposable NPWT costs were $1937 lower per episode of care. All else being equal (eg, if patient outcomes are similar), the lower payments represent a significant savings opportunity for Medicare and private payers. Considering the cost savings potential, additional cost and cost-effectiveness studies to validate these findings are warranted.
Acknowledgement
The authors thank Joann Hettasch, Arbor Communications, Inc, a member of the Fishawack Group of Companies, Conshohocken, PA, who provided medical writing assistance on behalf of Smith & Nephew
Disclosure
This study was funded by Smith & Nephew, Inc (Fort Worth, TX), which was involved in developing the study design, data collection, data analysis and interpretation, and the decision to publish these results. Dr. Hogan and Dr. Nair are paid consultants for Smith & Nephew. Mr. Delhougne and Ms. Tarka are employees of Smith & Nephew and may own shares of Smith & Nephew stock.
Affiliations
Mr. Delhougne is Senior Director of Health Economics, Smith & Nephew, Inc, Fort Worth, TX. Dr. Hogan is President, Direct Research LLC, Vienna, VA. Ms. Tarka is Vice President of Post-Acute Care, Smith & Nephew. Dr. Nair is Medical Director, Wound Healing Center at Presence St. Francis Hospital, Evanston, IL.
Correspondence
Please address correspondence to: Gary Delhougne, Smith & Nephew Biotherapeutics, 5600 Clearfork Main Street, Suite 600, Fort Worth, TX 76109; email: 该Email地址已收到反垃圾邮件插件保护。要显示它您需要在浏览器中启用JavaScript。.
