Ralph P. Insinga, PhD

: Limited real-world data are available on the administration costs associated with use of denosumab and zoledronic acid for the treatment of postmenopausal osteoporosis. The objective of the present study was to describe these costs within a large US healthcare claims database.

: Retrospective administrative claims database analysis.

: Data were obtained for US women ≥50 years with a healthcare claim for denosumab or zoledronic acid in 2013 within the Truven MarketScan Medicare and Commercial Claims database. After excluding women with cancer or Paget’s disease codes in 2012 or 2013, there were 9803 denosumab and 9141 zoledronic acid administrations. Administration costs potentially reflected payments to providers for healthcare professional encounters, injection-/infusion-related services, specimen-collection/lab-testing services to determine treatment eligibility (zoledronic acid only), and the drugs themselves. Specific criteria were used to determine whether to attribute observed resource use to denosumab/zoledronic acid administration versus other concomitant health issues managed during the encounter.

: Average nondrug administration costs were $55 per denosumab injection ($110/year when administered twice annually), composed of healthcare professional encounter ($11) and injection and related supply costs ($44). For zoledronic acid, costs were $187 per infusion, composed of healthcare professional encounter ($9), infusion and related supply costs ($168), and blood draw/lab costs ($10). Including annual costs for the drugs themselves ($1728 for denosumab and $1269 for zoledronic acid), total

annual costs were $1838 for and $1456 for zoledronic acid.

: This appears to be the first study to report, in detail, patterns of resource use and costs associated with real-world administration of denosumab and zoledronic acid. Total annual costs were higher than reported in previous analyses based on modeled resource use.

Osteoporosis is characterized by a systematic impairment of bone mass, strength, and micro-architecture, which increases the risk of fragility fracture.

 The percentage of US women estimated to meet current National Osteoporosis Foundation guidelines for the treatment of osteoporosis ranges from 12% among women aged 50-59 years to 79% among women ≥80 years.

To evaluate the cost-effectiveness and budget impact of available treatments for osteoporosis, an accurate understanding of the costs of therapy is needed.

 For oral osteoporosismedications, this may be chiefly estimated from a published drug price such as a wholesale acquisition cost (WAC) or from data on observed payments. In the case of the medications denosumab and zoledronic acid, costs also include those associated with treatment administration by a healthcare professional. Denosumab is administered via subcutaneous injection once every 6 months, and zoledronic acid via intravenous infusion once every 12 months.

Prior economic analyses of denosumab have based estimates of administration costs on assumption, with variation in the presumed type of resource use incurred. These presumptions have ranged from a nurse visit in the United States,

; to a general practitioner visit in Belgium

; to speculation that treatment might only be administered at specialist visits in the United Kingdom.

For zoledronic acid, prior analyses have assumed either nursing time costs for the United States,

 a clinic visit cost (type unspecified) for Finland and Norway,

 Payments for injection services and supplies have not been typically included in costs for denosumab, and assumptions for lab testing, infusion service, and supply-related costs for zoledronic acid have varied.

To better understand the setting, resource use, and costs associated with denosumab and zoledronic acid administration in real-world clinical practice within the United States, the present study evaluates data from a large healthcare claims database.

Data were obtained from the Truven MarketScan Medicare and Commercial Claims database. To be eligible for inclusion in the analysis, subjects were required to be female; aged ≥50 years; continuously enrolled in the database from January 1, 2012, through the date of their denosumab/zoledronic acid administration in 2013; to have a Healthcare Common Procedure Coding System (HCPCS) code for receipt of denosumab (J0897) or zoledronic acid (J3488 or Q2051) in 2013; and to lack care for cancer or Paget’s disease in 2012 or 2013.

The last criterion was applied to ensure the selection of a population receiving treatment for postmenopausal osteoporosis, as opposed to for bone loss related to cancer or Paget’s disease. Criteria used to identify cancer cases were any inpatient or outpatient 

International Classification of Diseases, Ninth Revision, Clinical Modification 

) diagnosis code (140.0-208.9, V58.0-V58.12 or V67.2) or Current Procedure Terminology (CPT) code (96401-96549) for cancer or its treatment, or an encounter where denosumab or zoledronic acid was administered in an oncologist’s office. Cases of Paget’s disease were identified by the presence of 

 code 731.0. As this was an anonymized database study, institutional review board approval was not sought.

Based on a review of all types of services for which payment was received at visits during which denosumab and zoledronic acid were administered, resource use and costs associated with these medications were classified into 1 of 4 groups: (a) drugs; (b) healthcare professional encounters; (c) injection or infusion services, and related supplies; (d) specimen collection and lab testing services (for zoledronic acid only). Costs for these services reflect payments from all sources (insurance, patient, etc). Methods for attributing healthcare resource use and costs of denosumab and zoledronic acid within each of these groups will now be described.

Drug costs were calculated from all payments at each encounter for billed HCPCS codes J0897 (for denosumab) and J3488/Q2051 (for zoledronic acid).

. Billing for healthcare professional encounters was identified through CPT procedure codes (99201-99215, 99241-99245, 99396-99397) for office visits/outpatient consultations and preventive medicine services. Healthcare professional encounters, and associated costs, were only attributed to denosumab or zoledronic acid administration if the healthcare encounter was deemed to have exclusively occurred for the purpose of administering one of these treatments. This criterion was met if the billed encounter exclusively included 

 1 or more osteoporosis evaluation, treatment, or symptom-related 

diagnosis codes (eAppendix Table A, available at

), and 1 or more procedure codes related to denosumab/zoledronic acid administration (eAppendix Table B), including a drug code for these treatments, as previously described.

These codes were identified based on a review of individual claims records in which they were utilized. In the presence of any alternate diagnosis code (eg, diabetes, anemia) or procedure code (eg, oral biopsy, ultrasound) for that provider on that date of service, it was assumed that the healthcare encounter would have taken place even in the absence of denosumab/zoledronic acid administration, and thus the healthcare encounter costs were not included in the administration costs for denosumab/zoledronic acid. In practice, this reflected instances where injectable osteoporosis medications were administered within a general wellness visit, alongside evaluation/treatment of a comorbid medical condition, or alongside other osteoporosis-related care (eg, fracture care).

* Injection or Infusion Services, and Related Supplies.

 These services and supplies reflect procedure codes (eAppendix Table B) related to injection or infusion billed separately from the drug cost for denosumab or zoledronic acid, including for provision of an injection or infusion, syringes, needles, and flushing of intravenous lines. Payments associated with these procedure codes were attributed to denosumab/zoledronic acid if the primary diagnosis code assigned for that service/supply reflected 1 of the 

codes listed in eAppendix Table A. Otherwise, the costs of these procedures were not attributed to denosumab/zoledronic acid. An example of nonattribution would be a separate injection of a major joint/bursa (CPT 20610) to treat osteoarthrosis (a degenerative joint disease with erosion of cartilage), with a primary 

 diagnosis code for that service of 715.00, at the same visit that a denosumab injection was given for osteoporosis.

* Specimen Collection and Lab Testing Services.

 Prior to receiving each dose of zoledronic acid, patients are to be evaluated to ensure they do not have impaired kidney function (serum creatinine testing), and may also be checked for hypocalcemia, through corresponding blood tests. Tests and results might be obtained on the same day as the treatment administration, or on a preceding day. Lab services potentially related to zoledronic acid administration were therefore examined on the date of infusion as well as within the preceding 30 days. For lab services performed on the same date as the infusion to be considered attributable to the zoledronic acid infusion, the primary diagnosis code assigned for the lab service must have been an osteoporosis evaluation-, treatment-, or symptom-related 

For attribution of lab services performed 1 to 30 days prior to the infusion, a similar set of criteria were applied, but with the exclusion of the diagnosis codes related to drug allergies/adverse events (995.20, 995.27, and 995.29), and fitting of a vascular catheter (V58.81), as the zoledronic acid infusion was not performed on that date. Upon meeting these criteria, services for blood calcium testing, kidney function testing, or both (CPT codes 80048, 80053, 82310, 82565) were attributed to the zoledronic acid infusion.

Costs for services and supplies beyond those listed in eAppendix Table B were not attributed to denosumab or zoledronic acid administration.

All analyses were conducted using SAS version 9.3 (Cary, NC). Means and percentages were calculated for sample descriptive data, and for the type of healthcare visit at which treatment was administered. Means, standard deviations, and 95% confidence intervals were estimated for drug and drug administration cost components.

Among 44,815 women aged 50+ years who were administered denosumab or zoledronic acid in 2013, 35,383 had continuous enrollment within the database back to January 1, 2012, and of these, 17,201 lacked evidence of a diagnosis or care for cancer or Paget’s disease. Within this group, there were 9803 administrations of denosumab and 9141 administrations of zoledronic acid. There were 1702 instances reflecting more than one administration of denosumab over the course of the year and 32 such instances for zoledronic acid administration (9 women had both a denosumab and a zoledronic acid administration on different dates within the year). An 

 code for osteoporosis (733.00-733.09) was listed in a primary, secondary, tertiary, or quaternary position on the claim for 91% of denosumab encounters and 93% of zoledronic acid encounters, with the remainder listing exclusively other diagnosis codes (eg, fracture, kyphosis).

At patient encounters where denosumab was administered, the mean patient age was 70 years, similar to the mean age of 68 years at encounters where zoledronic acid was administered (Table 1). The majority of denosumab (62%) and zoledronic acid (57%) patient encounters reflected patients with both Medicare and commercial insurance, with the remainder for patients with commercial insurance only.

The type of healthcare visit (provider and/or institution type) at which denosumab and zoledronic acid injections were administered is summarized in Table 2. For denosumab, approximately 40% of healthcare encounters were clearly identified as being with an office-based primary care provider (internal medicine or family practice physician). Approximately 30% of healthcare encounters were made to physicians specializing in (in descending order of frequency) rheumatology; endocrinology and metabolism; obstetrics and gynecology; hematology; and orthopedics. For 14% and 5% of patients, respectively, denosumab was administered at an outpatient hospital or pharmacy visit. For zoledronic acid, just over 40% of healthcare encounters occurred in an outpatient hospital setting.

Eighteen percent were clearly identified in relation to a primary care provider, while approximately 25% were in specialist settings of (in descending order of frequency) rheumatology; hematology; endocrinology and metabolism; and obstetrics and gynecology. The remaining patients were administered denosumab or zoledronic acid in other specialist, other primary care, or other institutional settings.

Table 3 reports healthcare costs associated with denosumab/zoledronic acid, per administration. In the case of costs for encounters with healthcare professionals, a procedure code for an encounter was billed for in only 36% of all encounters in which denosumab was administered and 19% of all encounters in which zoledronic acid was administered. Of those with a billing code for an encounter with a healthcare professional, 68% included other healthcare services, or management of other health conditions, unrelated to administration of denosumab, which was likewise observed within 52% of encounters in which zoledronic acid was administered.

Healthcare encounter costs were not attributed to denosumab or zoledronic administration in these instances. For the remaining instances with a billed code for a healthcare professional encounter (observed in 11% and 9% of all encounters in which denosumab and zoledronic acid were administered, respectively), the average paid amounts for the encounter itself were $99 and $93, respectively. Averaged across all encounters in which denosumab was administered, the average paid amount for the healthcare professional encounter was $11, with an analogous value of $9 for zoledronic acid.

 Services were billed for costs of injections and related supplies in 86% and 94%, respectively, of healthcare encounters in which denosumab and zoledronic acid were administered. Billed services for injection and related supplies could be specifically attributed to denosumab administration (versus another comorbid condition managed at the encounter) in 88% of these encounters (76% of all healthcare encounters in which denosumab was administered), and where billed, the average paid amount was $58. For zoledronic acid, the corresponding figures were 96%, 91%, and $185. Averaged across all encounters in which denosumab was administered, the average paid amount for attributable injections and related supplies was $44 as compared with $168 for zoledronic acid.

Laboratory services such as a blood draw and calcium and kidney function testing attributable to the zoledronic acid infusion were performed for 24% of infusions overall, and occurred on the same date as the infusion in 8% of infusions. Where billed and attributable, the average paid amount for specimen collection and lab services was $42. Averaged across all encounters where zoledronic acid was administered, the average paid amount was $10.

Excluding drug costs, total administration costs per patient averaged $55 (95% CI, $53-$59) for denosumab ($110 when administered twice annually) and $187 (95% CI, $182-$193) for zoledronic acid. Including the costs of the drugs themselves, total costs per patient administered denosumab averaged $920 (95% CI, $908-$932; $1838 when administered twice annually) with a corresponding figure of $1456 (95% CI, $1433-$1479) for zoledronic acid.

This study has described the real-world administration costs of denosumab and zoledronic acid among insured US women receiving treatment for osteoporosis, along with the setting of administration. It has aimed to fill a gap in the literature, because real-world resource use and costs associated with administration of these drugs have generally been based on assumed patterns of care, with uncertainty and variation across analyses in the presumed setting of administration and components of resource use.

Among previous US analyses, Parthan et al assumed a nurse visit cost for denosumab of $41 per administration, and a WAC-based drug cost of $825 (2011 data), for a total annual drug cost of $1732.

 In the present analysis, the corresponding per-administration cost for denosumab—encompassing healthcare encounter fees and costs of injections and related supplies&mdash;was $55, and the drug cost was $864, for a total annual cost of $1838. Were a 2014 WAC price for denosumab ($908)

 to be substituted for the paid amount reported within the present study, the corresponding annual cost would be $1926.

For zoledronic acid, Ferko et al assumed a serum calcium test, infusion administration cost, and infusion supply fee, representing a nondrug administration cost of $234 (2007 data), plus a drug WAC price of $1073 (2009 data), for a total cost of $1307.

 Parthan et al assumed an infusion administration cost of $144 (2012 data) and an identical drug WAC price to Ferko et al (2011 data) for a total cost of $1217.

 In the present analysis, the total nondrug administration cost was $187, with an average paid drug cost of $1269, for a total cost of $1456. Were a 2014 WAC price for zoledronic acid

 to be substituted for the aforementioned paid amount, the corresponding total cost would be $1271 for use of branded treatment ($1084 WAC), and $462-$782 ($375-$595 WAC range) for use of generic versions of zoledronic acid. Interestingly, although much-lower-cost generic versions of zoledronic acid were available for most of the analysis period in the present study (starting from early March 2013), paid drug costs generally reflected values at or above the branded WAC price, potentially suggesting either a lag in adjustments to reimbursement or continued high uptake of branded product.

Several findings emerged from the examination of individual components of healthcare administration costs. First, although denosumab administrations were largely split between primary care and specialist offices, and zoledronic acid administrations between outpatient hospital and specialist settings, in most instances there was no healthcare professional fee associated with the encounter. Thus, the injection/infusion may have been performed without significant time spent on physician consultation (eg, by a nurse or technician specializing in injectable medication administration or infusions).

Second, for more than three-quarters of patients, there was no evidence of associated billed services for kidney function and/or blood calcium testing on the date of zoledronic acid administrations, or within the 30 days prior, which may reflect a lack of routine testing of patients in clinical practice. Third, the total estimated annual costs for denosumab and zoledronic acid were $150-$250 higher than reported in prior analyses based on modeled healthcare resource use. The differential may in part reflect higher payments for drugs than the WAC price (eg, zoledronic acid), inclusion of additional components of resource use, and a broader set of codes used to bill for certain types of resource use in clinical practice (eg, costs of injections/infusions and related supplies) than has previously been modeled.

The present analysis has several caveats. First, reported costs may be conservative if costs of encounters in which additional diagnoses/procedures unrelated to denosumab/zoledronic acid administration were observed would have been lower had these osteoporosis treatments not been administered, or the strict attribution methods of the analysis overestimate instances in which visits would have otherwise occurred due to other diagnoses/procedures. Second, drug payments for zoledronic acid may decline in the future with increased entry and market penetration of generic competition.

Third, reported cost data are reflective of a privately insured population, with primary coverage through Medicare for patients aged 65 years and over. Based on Census Bureau data for 2013, 72% of the US population aged 45-64 years, and 54% of the population aged 65 years and over, had private health insurance,.

 Finally, costs associated with adverse effects resulting from the administration of denosumab/zoledronic acid were beyond the scope of the present study; however, such costs would not be expected to be substantive relative to the costs reported herein based on the safety profile of these products.

This appears to be the first study to report in detail patterns of resource use and costs associated with real-world administration of denosumab and zoledronic acid for treatment of postmenopausal osteoporosis. Estimated annual costs of administration were $110 for denosumab and $187 for zoledronic acid excluding drug costs, with total annual costs of $1838 and $1456, respectively, if including payments for each drug.

Thanks to Mary Anne Rutkowski for programming assistance in obtaining the data set used in this analysis.

RPI is employed by Merck & Co, Inc, and potentially owns stock and/or stock options in the company.

An earlier version of this manuscript underwent internal scientific and technical peer review at Merck & Co, Inc, prior to submission.

1. Rachner TD, Khosla S, Hofbauer LC. Osteoporosis: now and the future. 

 2011;377(9773):1276-1287. doi: 10.1016/S0140-6736(10)62349-5.

2. Dawson-Hughes B, Looker AC, Tosteson AN, Johansson H, Kanis JA, Melton LJ 3rd. The potential impact of the National Osteoporosis Foundation guidance on treatment eligibility in the USA: an update in NHANES 2005-2008. 

 2012;23(3):811-820. doi: 10.1007/s00198-011-1694-y.

3. Ferko NC, Borisova N, Airia P, Grima DT, Thompson MF. How rebates, copayments, and administration costs affect the cost-effectiveness of osteoporosis therapies. 

 2012;21(11):44-52. doi: 10.1007/s00198-011-1694-y.

4. Si L, Winzenberg TM, de Graaff B, Palmer AJ. A systematic review and meta-analysis of utility-based quality of life for osteoporosis-related conditions. 

 2014;25(8):1987-1997. doi: 10.1007/s00198-014-2636-2.

5. Parthan A, Kruse M, Yurgin N, Huang J, Viswanathan HN, Taylor D. Cost effectiveness of denosumab versus oral bisphosphonates for postmenopausal osteoporosis in the US. 

 2013;11(5):485-497. doi: 10.1007/s40258-013-0047-8.

6. Chau D, Becker DL, Coombes ME, Ioannidis G, Adachi JD, Goeree R. Cost-effectiveness of denosumab in the treatment of postmenopausal osteoporosis in Canada. 

 2012;15 Suppl 1:3-14. doi: 10.3111/13696998.2012.737393.

7. Ström O, Jönsson B, Kanis JA. Intervention thresholds for denosumab in the UK using a FRAX-based cost-effectiveness analysis. 

 2013;24(4):1491-1502. doi: 10.1007/s00198-012-2115-6.

Articles

This is the first study to report, in detail, patterns of resource use and costs associated with real-world administration of denosumab and zoledronic acid. 