SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. ADDITIONAL DISCLOSURES
  10. REFERENCES

Objective

To assess the cost effectiveness of duloxetine compared to other oral postacetaminophen treatments for osteoarthritis (OA) from a Quebec societal perspective.

Methods

A cost-utility analysis was performed enhancing the Markov model from the 2008 OA guidelines of the National Institute for Health and Clinical Excellence (NICE). The NICE model was extended to include opioid and antidepressant comparators, adding titration, discontinuation, and relevant adverse events (AEs). Comparators included duloxetine, celecoxib, diclofenac, naproxen, hydromorphone, and oxycodone extended release (oxycodone). AEs included gastrointestinal and cardiovascular events associated with nonsteroidal antiinflammatory drugs (NSAIDs), as well as fracture, opioid abuse, and constipation, among others. Costs and incremental cost-effectiveness ratios (ICERs) were estimated in 2011 Canadian dollars. The base case modeled a cohort of 55-year-old patients with OA for a 12-month period of treatment, followed by treatment from a basket of post-discontinuation oral therapies until death. Sensitivity analyses (one-way and probabilistic) were conducted.

Results

Overall, naproxen was the least expensive treatment, whereas oxycodone was the most expensive. Duloxetine accumulated the highest number of quality-adjusted life years (QALYs), with an ICER of $36,291 per QALY versus celecoxib. Duloxetine was dominant over opioids. In subgroup analyses, ICERs for duloxetine versus celecoxib were $15,619 and $20,463 for patients at high risk of NSAID-related AEs and patients ages ≥65 years, respectively.

Conclusion

Duloxetine was cost effective for a cohort of 55-year-old patients with OA, and more so in older patients and those with greater AE risks.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. ADDITIONAL DISCLOSURES
  10. REFERENCES

Osteoarthritis (OA) is a chronic degenerative condition characterized by the breakdown of joint cartilage. The breakdown of cartilage causes bones to rub against each other, resulting in stiffness, pain, and loss of movement in the joint. The most frequently affected joints are those of the hands, knees, hips, and spine ([1]). The risk of developing OA increases with age. OA is more common in women, patients born with malformed joints or defective cartilage, and patients affected by conditions such as gout, diabetes mellitus, or Paget's disease of bone ([2]). OA usually presents as pain of insidious onset that is aching and not well localized. Other symptoms include reduced function, stiffness of short duration, joint instability, reduced movement, and crepitus ([3]).

Management guidelines recommend acetaminophen as first-line oral pharmacologic treatment of OA because of its efficacy in management of mild to moderate pain and low level of adverse events (AEs) when administered at dosages of up to 4 gm/day ([4-6]). Nonsteroidal antiinflammatory drugs (NSAIDs) are considered when acetaminophen cannot control symptoms or when inflammation is present. They are regarded as more effective than acetaminophen, but are associated with a higher risk of gastrointestinal (GI) and cardiovascular (CV) events ([5, 6]). Therefore, NSAIDs should be used at the lowest effective dose for the shortest possible duration ([4]). Opioid analgesics may be considered when NSAIDs are ineffective or contraindicated. However, the risks and benefits must be carefully weighed ([4, 5]). At least one guideline recommends that strong opioids be used only in cases of severe pain ([6]).

Duloxetine is a selective serotonin and norepinephrine reuptake inhibitor with demonstrated analgesic effects in OA, chronic low back pain, diabetic peripheral neuropathy, and fibromyalgia ([7]). Its efficacy with respect to pain inhibition is believed to result from potentiation of descending inhibitory pain pathways within the central nervous system ([8]). Multiple phase III clinical trials have demonstrated the efficacy of duloxetine in management of OA ([9, 10]). Although a link between depression and physical symptoms such as pain has been established ([11]), the trials used to establish the efficacy of duloxetine in OA excluded patients with psychiatric disorders, including major depressive disorder ([9, 10]). In addition, duloxetine has demonstrated a safety profile whereby most AEs are mild and transitory in nature ([12]). Duloxetine was initially approved in Canada for the symptomatic relief of major depressive disorder and the management of neuropathic pain associated with diabetic peripheral neuropathy ([8]). Subsequently, Health Canada has approved duloxetine for the additional indications of generalized anxiety disorder, fibromyalgia, knee OA, and chronic low back pain ([7]). The objective of this study was to assess the cost effectiveness of duloxetine compared to other oral post-acetaminophen OA pain treatments from a Quebec societal perspective.

Box 1. Significance & Innovations

  • Duloxetine previously has not been assessed for cost effectiveness in osteoarthritis in Canada. This article provided a unique analysis of the cost effectiveness of duloxetine for management of osteoarthritis symptoms in a Canadian population.
  • The cost effectiveness of oral therapies from multiple drug classes was estimated in a single model, providing readers with a wide range of cost-effectiveness information in a single study.
  • Results from this study highlighted the cost effectiveness of duloxetine in the management of osteoarthritis symptoms.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. ADDITIONAL DISCLOSURES
  10. REFERENCES

Background

A cost-effectiveness analysis (CEA) is an economic analysis that compares the costs versus the health effects of ≥2 medical treatments. The outcome of a CEA is the incremental cost-effectiveness ratio (ICER), the difference in costs between treatments divided by the difference in benefits ([13]). Costs are represented in units of currency, while benefits may take disease-specific or general health forms. A special case of the CEA is the cost-utility analysis (CUA), in which benefits are represented by quality-adjusted life-years (QALYs). The QALY is a measure that represents both the quality and quantity of life ([14]). Quality of life is measured as a utility: a valuation of a particular state of health compared to the ideal of perfect health on a 0–1 scale (where 0 = dead and 1 = perfect health). The ICER of a CUA takes the form of cost per QALY. If a treatment has higher costs but lower benefits than another treatment, that treatment is dominated by the other treatment and an ICER is not needed. Willingness to pay represents the value society places on a year of perfect health, a value that frequently is tied to the wealth of the society ([15]). It is also expressed as cost per QALY.

A basic characteristic of a CEA is its perspective, i.e., the society or health care system it models. Common perspectives consist of those that include only medical costs, such as a health care system or an insurer perspective. In contrast, a societal perspective accounts for all costs that are relevant to the disease and its management, including medical costs and costs incurred at the employer and individual patient levels, such as loss of productivity and caregiver burden ([13]).

This CUA evaluated the cost effectiveness of duloxetine compared to other oral treatments for OA symptom management from the Quebec societal perspective. Based on factors such as the structure of the health care system, regulatory approval, and willingness to pay, individual nations vary with respect to which medications are made available for treating patients with OA. Because availability of medications in Canada varies by province, this discussion and model focused exclusively on treatments that are available in Quebec via Régie de l'assurance maladie du Québec (RAMQ), the provincial health insurance system for Quebec ([16]).

CEAs in Canada frequently cite $50,000/QALY as a willingness-to-pay threshold, although $100,000 and multiple thresholds may be cited. Willingness to pay may be influenced by a number of factors, including the strength of the evidence ([17, 18]). Our analysis conservatively assumed $50,000/QALY as the willingness-to-pay threshold.

Population and model design

The modeled population consisted of patients with OA with chronic moderate to severe pain uncontrolled by acetaminophen. In the base case, the cohort began at age 55 years (the mean age of OA patients in Canada) without a history of GI or CV events, received duloxetine or a comparator treatment for up to 12 months, and was followed until age 100 years or death.

A total of 6 therapies were included in the analysis, including representative NSAIDs, opioids, and duloxetine (Table 1). The structure was a semi-Markov model implemented in Microsoft Excel using 3-month cycles for the first 3 years (the maximum length of initial treatment) and annual cycles thereafter. Costs and ICERs were denominated in 2011 Canadian dollars ($), with an annual discount rate of 5.0% applied to costs and outcomes as recommended by the Canadian Agency for Drugs and Technologies in Health ([19]).

Table 1. Treatments*
TherapyDrug classDose
  1. SSNRI = selective serotonin and norepinephrine reuptake inhibitor; COX-2 = cyclooxygenase 2; NSAID = nonsteroidal antiinflammatory drug.

DuloxetineSSNRI60 mg every day
CelecoxibCOX-2 inhibitor NSAID200 mg
DiclofenacNonselective NSAID100–150 mg
NaproxenNonselective NSAID750 mg
HydromorphoneStrong opioid3–9 mg twice a day
OxycodoneStrong opioid10–30 mg twice a day

The model was based on an earlier model documented in Appendix D of the 2008 OA treatment guidelines published for the National Institute for Health and Clinical Excellence (NICE) by the National Collaborating Centre for Chronic Conditions of the Royal College of Physicians ([20]). The NICE model was chosen because it was recent, included GI and CV AEs, and utilized a structure that was flexible and thorough in the modeling of oral treatments for OA. The NICE model may also be viewed as an extension of the earlier model by Maetzel et al created for the Canadian Coordinating Office for Health Technology Assessment ([21]). The NICE model was extended to allow opioids and duloxetine as comparators by adding the ability to model titration, discontinuation, and an expanded number of AEs. Although the NICE used acetaminophen as a post-discontinuation therapy, a basket of all comparators composed the post-discontinuation therapy in our model weighted by their market shares in Quebec; duloxetine was included in the basket with an assumed market share of 5%. The model assumed that all patients who discontinued initial therapy would switch among treatments in the post-discontinuation therapy basket until death. The model considered treatment cost, efficacy, GI and CV AEs, fractures, transient AEs, titration, discontinuation, and the benefit offered by a co-prescribed proton-pump inhibitor (PPI).

AEs were classified as persistent (with an associated risk of mortality and long-term sequelae) or transient (resolving upon treatment discontinuation). Health states in the model were defined by persistent AEs, i.e., a 3-month state in which the AE took place and a subsequent post-AE state in which long-term sequelae were modeled. Persistent AEs included symptomatic ulcer, complicated GI bleed, stroke, myocardial infarction, heart failure, and fracture.

Utilities

The model included treatment efficacy, AEs, and age as components of each calculated utility value, with a utility structure similar to that of the NICE model. Treatment-specific utilities were estimated using a Bayesian mixed treatment comparison of change from baseline total Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) Likert scores. Summary change from baseline WOMAC scores was estimated from 26 OA randomized controlled trials (RCTs) of at least 12 weeks' duration, adjusted for baseline WOMAC scores and then converted to utilities using the transfer-to-utility technique from Barton et al ([22]). Since no RCTs were available for hydromorphone, its efficacy and treatment-specific utility were assumed to be equivalent to oxycodone.

Persistent AE utility weights were adopted from the NICE model ([20]), whereas transient AE utility weights were taken from the literature. Utility weights for age and sex were derived from the 2007/2008 Canadian Community Health Survey ([23]).

Transition probabilities

Transition probabilities to persistent GI and CV AE health states were taken from the NICE model ([20]) or derived from Vestergaard for fracture ([24]). The probabilities from the NICE were based on indirect comparisons of 3 landmark NSAID trials ([25-27]). As in the NICE model, NSAID doses in our model were limited to doses that are prescribed in practice, as opposed to the relatively high doses found in the trials. The model adopted the NICE assumption that rates of persistent GI and CV AEs decreased by 25% as the dose decreased by 50% ([20]). Probabilities of transient AEs were estimated from a meta-analysis of 28 trials using a maximum likelihood simulation technique when AE rates fell below publication thresholds ([28]). This method assumed that trial rates for an AE were part of a single binomial distribution truncated by the reporting thresholds of each study. Probabilities of treatment discontinuation were estimated from a meta-analysis of 25 RCTs of 12–13 weeks' duration using DerSimonian-Laird frequentist methods. If no studies were available for a treatment, the rate was assumed to be equal to another drug of the same class. The 3 meta-analyses described here included nearly all the same RCTs; differences in trial inclusion depended on reporting variations among the trials.

Transition probabilities to GI AE health states were reduced ([20]) by the assumption that PPI coprescriptions would be used at rates common in Canada ([29]). Furthermore, it was assumed that transition probabilities for both GI and CV AEs would be influenced by age, with higher rates at higher ages ([30, 31]).

Costs

Treatment costs (Table 2) were calculated based on trial dosing and an IMS-Brogan analysis of RAMQ drug costs ([33]). Costs of consultation were taken from the Ontario Schedule of Benefits for Physician Services. Costs for transient and persistent AEs were calculated from utilization as described by the NICE ([20]) and costs from Quebec and Ontario sources ([20, 33-37]).

Table 2. Treatment costs
TreatmentFirst 3-month drug costFirst 3-month physician costSubsequent 3-month costDiscontinuation drug costDiscontinuation physician cost
  1. a

    Provided by Lilly Canada.

  2. b

    Calculated from the Ministry of Health and Long-Term Care (2010) ([34]), guided by expert opinion solicited by questionnaire.

  3. c

    Calculated from IMS-Brogan (2010) ([33]).

  4. d

    Calculated from IMS-Brogan (2010) ([33]), using tapering calculated by the Washington State Department of Social and Health Services, 2010 ([32]).

Duloxetine 60 mg$335.26a$65.32b$340.31c$0.00d$44.63b
Celecoxib 200 mg$126.04c$0.00$126.04c$0.00$0.00
Diclofenac 100–150 mg$47.78c$0.00$47.78c$0.00$0.00
Hydromorphone 3–9 mg twice a day$83.43c$83.63b$94.26c$27.73d$63.46b
Naproxen 750 mg$36.14c$0.00$36.14c$0.00$0.00
Oxycodone 10–30 mg twice a day$224.87c$83.63b$257.22c$99.19d$63.46b

Costs for loss of productivity were estimated using the human capital approach. Two types of loss of productivity costs were modeled: loss of productivity as a result of having OA and loss of productivity caused by treatment-related AEs. Loss of productivity costs related to disease state were based on the study by Maetzel ([38]). The loss of productivity costs from treatment-related AEs were based on workers' compensation insurance guidelines for the US regarding days until return to work after the initial occurrence of a persistent AE ([39]). The value of days of work lost was estimated from Canadian sources for the number of days and hours in a work week, the hourly wage, and the average age at retirement ([40-43]). The model was able to calculate loss of productivity costs using minimum or average hourly wages, with the minimum wage used for the base case.

Outcomes

The primary outcome measure was the cost per QALY ICER. A number of secondary outcomes such as components of cost were also available. A cost-effectiveness plane, which plotted incremental costs (y-axis) versus incremental QALYs (x-axis), was used to compare the relative cost effectiveness of various treatments.

Sensitivity analysis

Sensitivity analyses were performed to quantify uncertainty and determine the key drivers of model results. Twenty-five one-way sensitivity analyses of comparisons of a comparator to duloxetine were conducted. One-way sensitivity analyses measure the impact that discrete changes in a given parameter will have on the results of the model. This allows the user to estimate the level of confidence a decision maker can have in the results of the economic evaluation. In probabilistic sensitivity analysis, cost-effectiveness results were calculated 500 times, each time replacing 18 input variables with newly sampled values from predefined distributions.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. ADDITIONAL DISCLOSURES
  10. REFERENCES

Cost and cost-utility results

In the base case, the model estimated that after AEs were considered, naproxen was the least expensive but least effective treatment, while oxycodone was the most expensive treatment and duloxetine was the most effective treatment. ICERs for diclofenac (versus naproxen), celecoxib (versus diclofenac), and duloxetine (versus celecoxib) were $16,491, $28,258, and $36,291, respectively (Table 3). The cost-effectiveness plane (Figure 1) depicts the incremental costs and QALYs in relation to naproxen, the least expensive treatment. The line connecting naproxen, diclofenac, celecoxib, and duloxetine forms the cost-effectiveness frontier, while hydromorphone and oxycodone were dominated, as seen by their location above and to the left of the frontier.

Table 3. Results of the base-case incremental cost-effectiveness analysis*
TreatmentCost over naproxenaQALYs over naproxenaICER vs. baselinebIncremental costcIncremental QALYsbICER
  1. QALYs = quality-adjusted life years; ICER = incremental cost-effective ratio.

  2. a

    Costs and QALYs discounted at 5.0%. “Baseline” is the least expensive treatment.

  3. b

    “Baseline” is the least expensive treatment.

  4. c

    Costs and QALYs discounted at 5.0%.

Oxycodone$1,7220.0173$99,456  Dominated
Hydromorphone$1,3940.0165$84,636  Dominated
Duloxetine$9370.0284$32,960$806 vs. celecoxib0.0222 vs. celecoxib$36,291 vs. celecoxib
Celecoxib$1310.0062$21,056$68 vs. diclofenac0.0024 vs. diclofenac$28,258 vs. diclofenac
Diclofenac$630.0038$16,491$63 vs. naproxen0.0038 vs. naproxen$16,491 vs. naproxen
Naproxen (baseline)
image

Figure 1. Cost-effectiveness plane of the base-case analysis based on the Quebec societal perspective. QALYs = quality-adjusted life years.

Download figure to PowerPoint

Sensitivity analysis results

Twenty-five one-way sensitivity analyses of duloxetine evaluated against a comparator were conducted; the results of these analyses were highly dependent on the drug class of the comparator. When using an NSAID as a comparator, the model was most sensitive to NSAID-related CV AE rates, PPI usage rates, and the discount rate, as the effects of NSAID-related persistent AEs continue into the distant future. When the comparator was a non-NSAID, other parameters had a greater impact on the model results. For instance, with hydromorphone as a comparator, the 3 most sensitive inputs were length of transient AE treatment, hydromorphone discontinuation rate, and cost of duloxetine.

Based on the probabilistic sensitivity analysis of duloxetine versus celecoxib (Figure 2), incremental costs clustered around $800 with nearly all values ±$100 (12%). Incremental QALYs clustered near 0.022 with most values ±0.008 (36%). A greater degree of variation was associated with the estimation of quality of life than with costs. The mean cost per QALY (as shown by the diamond) and nearly all of the 500 iterations were associated with ICERs below a willingness-to-pay threshold of $50,000 (as shown by the diagonal line). The cost-effectiveness acceptability curve (Figure 3) shows a median ICER of $36,529 (consistent with the base-case mean ICER of $36,291) and a 95% probability of being below a willingness-to-pay threshold of $50,000 per QALY.

image

Figure 2. Probabilistic sensitivity analysis of duloxetine versus celecoxib, with the white diamond showing the base-case scenario. QALYs = quality-adjusted life years.

Download figure to PowerPoint

image

Figure 3. Cost-effectiveness acceptability curve for the base-case analysis showing willingness to pay for duloxetine versus celecoxib. ICER = incremental cost-effective ratio; QALY = quality-adjusted life year.

Download figure to PowerPoint

Subgroup analyses

Subgroup analyses were performed on a high-risk population of patients with a history of CV or GI events and a cohort of patients ages ≥65 years. These subgroups were chosen because duloxetine is associated with a lower risk of CV and GI AEs than NSAIDs and because the risk of CV and GI AEs rises with age ([30, 31]). In the analysis of patients with a history of CV or GI events, naproxen remained the least expensive treatment. The ICER for duloxetine versus celecoxib decreased to $15,619, which indicates duloxetine is a particularly good value for patients at high risk of CV and GI events. For the cohort of patients ages ≥65 years, the ICER for duloxetine versus celecoxib decreased to $20,463, indicating that duloxetine is also more cost effective for an older population.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. ADDITIONAL DISCLOSURES
  10. REFERENCES

The analysis used in this study extends the model created for the 2008 NICE OA guidelines. Both models use similar structures and incorporate many identical data values, including transition probabilities and utilities associated with persistent AEs. Because of these similarities, the 2 models share many strengths and limitations. The strengths of this model include its use of a comprehensive set of utilities and utility weights, AE estimates derived from a NICE meta-analysis, and its Markov structure.

One limitation of our model was the use of 3 years as the maximum length for the initial oral treatment of OA, which is a chronic condition. However, treatment from the post-discontinuation therapy continued for life. Also, several other models, including those by Bessette et al ([44]), Chen et al ([45]), and Maetzel et al ([21]), have used similar maximum treatment lengths. The NICE model ([20]) utilized a base-case treatment duration of 3 months, but in sensitivity analysis allowed for lifetime treatment.

Our model adopted the probabilities of NSAID-related CV and GI AEs from the NICE OA model. Therefore, we also adopted the NICE assumption that rates of NSAID-related GI and CV AEs decrease by 25% as the dose decreases by 50%. This assumption was tested in the sensitivity analysis of the NICE model, which found that the overall cost-effectiveness results were robust to changes in this assumption ([20]).

A limited number of RCTs were available for the meta-analyses used to estimate treatment efficacy, discontinuation rates, and transient AE rates. Only the estimates for naproxen and celecoxib were based on at least 4 studies, whereas there has been only 1 minor study for oxycodone and no studies for hydromorphone in OA. For approximately one-third of AEs, there were no studies that reported an AE rate for the treatment. In these cases, the model assumed that the efficacy and/or rates of AEs were equal to those of other drugs in the same class.

Constipation was a particularly important AE in the analyses of opioids. The rates of constipation used in the model were consistent with the literature. However, the rate of severe constipation that required medical attention was based on a single study ([46]).

An additional limitation was that the model failed to include several aspects of chronic conditions that are medically managed. These aspects include disease progression, loss of efficacy over time, and the associated need for dose escalation. Treatment costs associated with opioids may have been underestimated because opioid tolerance was not included in the model. With opioid tolerance developing over time, patients may require higher doses, thereby increasing opioid drug costs, outpatient physician costs, and the cost of opioid-related AEs ([47]).

Estimates of loss of productivity costs were based on limited available information and were likely to be conservative. The majority of productivity cost estimates were derived from time lost from work because of upfront management of persistent AEs. Consequently, this calculation does not reflect all societal costs as a result of OA therapy, such as the costs incurred by patients who take time off from work for physician visits associated with titration, discontinuation, and transient AEs. This calculation also fails to account for costs related to caregivers, transportation, or post–AE period costs in light of a lack of reliable sources for these data. The costs for the post–AE period could become significant, especially in patients who have experienced a stroke. Indirect costs associated with opioid addiction and dependence are absent from the model, which likely underestimates opioid-associated costs from a Quebec societal perspective.

The model estimated that duloxetine was a cost-effective treatment for OA with an ICER of $36,291 versus celecoxib from a Quebec societal perspective. Based on the probabilistic sensitivity analysis estimates, there was a 95.0% probability that duloxetine would fall below a willingness-to-pay threshold of $50,000. Results also showed that duloxetine dominated strong opioids in the base case as well as in all scenarios under one-way sensitivity analyses. For patients at higher risk of CV or GI AEs, including the RAMQ population age ≥65 years, duloxetine was estimated to be particularly cost effective, with a cost per QALY no greater than $22,500.

This model represented a positive examination of the cost effectiveness of duloxetine in the management of OA-associated symptoms in a Canadian population. Economic studies that have evaluated the role of duloxetine in patients with chronic low back pain, fibromyalgia, and painful diabetic peripheral neuropathy also have demonstrated evidence of the cost effectiveness of duloxetine in the management of those patients ([48-50]). Potential areas of further research include expansion of the analysis to include loss of efficacy, the need for dosage escalation, and a more comprehensive array of indirect costs.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. ADDITIONAL DISCLOSURES
  10. REFERENCES

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

Study conception and design. Wielage, Lee, Klein, Happich.

Acquisition of data. Wielage, Patel, Bansal.

Analysis and interpretation of data. Wielage, Patel, Bansal, Klein, Happich.

ROLE OF THE STUDY SPONSOR

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. ADDITIONAL DISCLOSURES
  10. REFERENCES

The role of Eli Lilly and Company was limited to the participation of authors Lee and Happich. Publication of this article was not contingent upon approval by Eli Lilly and Company.

ADDITIONAL DISCLOSURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. ADDITIONAL DISCLOSURES
  10. REFERENCES

Authors Wielage, Patel, and Klein are employees of Medical Decision Modeling Inc., a contract research organization and vendor to Eli Lilly and Company. Author Bansal was employed by Medical Decision Modeling Inc. at the time of the study and is currently employed by Thought Semantics, LLC.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ROLE OF THE STUDY SPONSOR
  9. ADDITIONAL DISCLOSURES
  10. REFERENCES
  • 1
    Goldring S, Goldring M. Clinical aspects, pathology and pathophysiology of osteoarthritis. J Musculoskelet Neuronal Interact 2006;6:3768.
  • 2
  • 3
    Hunter DJ, McDougall JJ, Keefe FJ. The symptoms of osteoarthritis and the genesis of pain. Rheum Dis Clin North Am 2008;34:62343.
  • 4
    National Collaborating Centre for Chronic Conditions.Osteoarthritis: national clinical guideline for care and management in adults.London:Royal College of Physicians;2008.
  • 5
    Zhang W, Doherty M, Arden N, Bannwarth B, Bijlsma J, Gunther KP, et al. EULAR evidence based recommendations for the management of hip osteoarthritis: report of a task force of the EULAR Standing Committee for International Clinical Studies Including Therapeutics (ESCISIT). Ann Rheum Dis 2005;64:66981.
  • 6
    Zhang W, Moskowitz RW, Nuki G, Abramson S, Altman RD, Arden N, et al. OARSI recommendations for the management of hip and knee osteoarthritis, part II: OARSI evidence-based, expert consensus guidelines. Osteoarthritis Cartilage 2008;16:13762.
  • 7
    Eli Lilly Canada Inc.Product monograph Cymbalta.Toronto:Eli Lilly Canada;2011.
  • 8
  • 9
    Chappell A, Ossanna M, Liu-Seifert H, Iyengar S, Skljarevski V, Li L, et al. Duloxetine, a centrally acting analgesic, in the treatment of patients with osteoarthritis knee pain: a 13-week, randomized, placebo-controlled trial. Pain 2009;146:25360.
  • 10
    Chappell AS, Desaiah D, Liu-Seifert H, Zhang S, Skljarevski V, Belenkov Y, et al. A double-blind, randomized, placebo-controlled study of the efficacy and safety of duloxetine for the treatment of chronic pain due to osteoarthritis of the knee. Pain Pract 2011;11:3341.
  • 11
    Trivedi MH. The link between depression and physical symptoms. Prim Care Companion J Clin Psychiatry 2004;6:126.
  • 12
    Brunton S, Wang F, Edwards SB, Crucitti AS, Ossanna MJ, Walker DJ, et al. Profile of adverse events with duloxetine treatment: a pooled analysis of placebo-controlled studies. Drug Saf 2010;33:393407.
  • 13
    Russell LB, Gold MR, Siegel JE, Daniels N, Weinstein MC. The role of cost-effectiveness analysis in health and medicine. JAMA 1996;276:11727.
  • 14
    Walley T. Pharmacoeconomics and economic evaluation of drug therapies. In: Haycox A, Boland A, editors.Pharmacoeconomics. 1st ed.London:Churchill Livingstone;2004. p.6775.
  • 15
    World Health Organization.Macroeconomics and health: investing in health for economic development. Report of the WHO Commission on Macroeconomics and Health.Geneva:World Health Organization;2001.
  • 16
    Régie de l'assurance maladie du Québec. List of medications.2011. URL: http://www.prod.ramq.gouv.qc.ca/DPI/PO/Commun/PDF/Liste_Med/Liste_Med/liste_med_cor1_2010_06_01_en.pdf.
  • 17
    Laupacis A, Feeny D, Detsky AS, Tugwell PX. How attractive does a new technology have to be to warrant adoption and utilization? Tentative guidelines for using clinical and economic evaluations. CMAJ 1992;146:47381.
  • 18
    Braithwaite RS, Meltzer DO, King JT, Leslie D, Roberts MS. What does the value of modern medicine say about the $50,000 per quality-adjusted life-year decision rule?Med Care 2008;46:34956.
  • 19
    Canadian Agency for Drugs and Technologies in Health.Guidelines for the economic evaluation of health technologies: Canada. 3rd ed.Ottawa:Canadian Agency for Drugs and Technologies in Health;2006.
  • 20
    National Collaborating Centre for Chronic Conditions.Osteoarthritis: national clinical guideline for care and management in adults. Appendix D.London:Royal College of Physicians;2008.
  • 21
    Maetzel A, Krahn M, Naglie G.The cost-effectiveness of celecoxib and rofecoxib in patients with osteoarthritis or rheumatoid arthritis.Ottawa:Canadian Coordinating Office for Health Technology Assessment;2001.
  • 22
    Barton GR, Sach TH, Jenkinson C, Avery AJ, Doherty M, Muir KR. Do estimates of cost-utility based on the EQ-5D differ from those based on the mapping of utility scores?Health Qual Life Outcomes 2008;6:51.
  • 23
    Statistics Canada. Canadian Community Health Survey: public use microdata file 2007/2008 (cycle 4.1).2009. URL: http://www5.statcan.gc.ca/bsolc/olc-cel/olc-cel?catno=82M0013X&lang=eng.
  • 24
    Vestergaard P. Pain-relief medication and risk of fractures. Curr Drug Saf 2008;3:199203.
  • 25
    Silverstein FE, Faich G, Goldstein JL, Simon LS, Pincus T, Whelton A, et al. Gastrointestinal toxicity with celecoxib vs nonsteroidal anti-inflammatory drugs for osteoarthritis and rheumatoid arthritis: the CLASS study. A randomized controlled trial. JAMA 2000;284:124755.
  • 26
    Combe B, Swergold G, McLay J, McCarthy T, Zerbini C, Emery P, et al. Cardiovascular safety and gastrointestinal tolerability of etoricoxib vs diclofenac in a randomized controlled clinical trial (the MEDAL study). Rheumatology (Oxford) 2009;48:42532.
  • 27
    Farkouh ME, Kirshner H, Harrington RA, Ruland S, Verheugt FW, Schnitzer TJ, et al. Comparison of lumiracoxib with naproxen and ibuprofen in the Therapeutic Arthritis Research and Gastrointestinal Event Trial (TARGET), cardiovascular outcomes: randomised controlled trial. Lancet 2004;364:67584.
  • 28
    Wielage RC, Samsa GP, Klein TM, Happich M. A maximum likelihood simulation technique for estimating adverse event rates from published trials PRM145. In:International Society for Pharmacoeconomics and Outcomes Research 15th Annual European Congress.Berlin:ISPOR;2012.
  • 29
    Canadian Institute for Health Information.Proton pump inhibitor use in seniors: an analysis focusing on drug claims, 2001 to 2008.Canada:Canadian Institute for Health Information;2009.
  • 30
    Hippisley-Cox J, Coupland C, Logan R. Risk of adverse gastrointestinal outcomes in patients taking cyclo-oxygenase-2 inhibitors or conventional non-steroidal anti-inflammatory drugs: population based nested case-control analysis. BMJ 2005;331:13106.
  • 31
    Hippisley-Cox J, Coupland C, Vinogradova Y, Robson J, May M, Brindle P. Derivation and validation of QRISK, a new cardiovascular disease risk score for the United Kingdom: prospective open cohort study. BMJ 2007;335:136.
  • 32
    Washington State Department of Social & Health Services.Calculator: tapering plan for client with chronic, non-cancer pain.Lakewood (WA):Washington State Department of Social & Health Services;2009.
  • 33
    IMS-Brogan Inc.Delta PATM. Provincial formulary database, Quebec.Ottawa:IMS-Brogan;2011.
  • 34
    Ministry of Health and Long-Term Care. Ontario schedule of benefits for physician services.2010. URL: http://www.health.gov.on.ca/english/providers/program/ohip/sob/physserv/physserv_mn.html.
  • 35
    Ontario Case Costing Initiative. OCCI costing analysis tool.2010. URL: http://www.occp.com/.
  • 36
    Fraser Health. Hospice palliative care program symptom guidelines: nausea and vomiting.2006.
  • 37
    Swegle JM, Logemann C. Management of common opioid-induced adverse effects. Am Fam Physician 2006;74:134754.
  • 38
    Maetzel A. The economic burden associated with osteoarthritis, rheumatoid arthritis, and hypertension: a comparative study. Ann Rheum Dis 2004;63:395401.
  • 39
    Work Loss Data Institute.Official disability guidelines. 15th ed.Encinitas (CA):WLDI;2010.
  • 40
    Kozhaya N.The retirement age in Quebec: a worrying situation.Montreal:Montreal Economic Institute;2007.
  • 41
    Gouvernement du Québec. Wages: commission des normes du travail du Québec.2010. URL: http://www.cnt.gouv.qc.ca/en/wages-pay-and-work/wages/index.html.
  • 42
    Tremblay DG, Najem E, Paquet R. Work-family balancing and working time: what measures are available to Canadian workers and what measures should employers develop?Glob J Bus Res 2007;1:97113.
  • 43
    Statistics Canada. Average hourly wages of employees by selected characteristics and profession, unadjusted data, by province (monthly) (Quebec).2011. URL: http://www.statcan.gc.ca/tables-tableaux/sum-som/l01/cst01/labr69f-eng.htm.
  • 44
    Bessette L, Risebrough N, Mittmann N, Roussy JP, Ho J, Zlateva G. Cost-utility of celecoxib use in different treatment strategies for osteoarthritis and rheumatoid arthritis from the Quebec healthcare system perspective. J Med Econ 2009;12:24658.
  • 45
    Chen YF, Jobanputra P, Barton P, Bryan S, Fry-Smith A, Harris G, et al. Cyclooxygenase-2 selective non-steroidal anti-inflammatory drugs (etodolac, meloxicam, celecoxib, rofecoxib, etoricoxib, valdecoxib and lumiracoxib) for osteoarthritis and rheumatoid arthritis: a systematic review and economic evaluation. Health Technol Assess 2008;12:1278.
  • 46
    Hjalte F, Berggren AC, Bergendahl H, Hjortsberg C. The direct and indirect costs of opioid-induced constipation. J Pain Symptom Manage 2010;40:696703.
  • 47
    Trescot AM, Helm S, Hansen H, Benyamin R, Glaser SE, Adlaka R, et al. Opioids in the management of chronic non-cancer pain: an update of American Society of the Interventional Pain Physicians' (ASIPP) guidelines. Pain Physician 2008;11:S562.
  • 48
    O'Connor A, Noyes K, Holloway R. A cost-utility comparison of four first-line medications in painful diabetic neuropathy. Pharmacoeconomics 2008;26:104564.
  • 49
    Beard SM, Roskell N, Le TK, Zhao Y, Coleman A, Ang D, et al. Cost effectiveness of duloxetine in the treatment of fibromyalgia in the United States. J Med Econ 2011;14:46376.
  • 50
    Wielage RC, Bansal M, Andrews JS, Wohlreich MM, Klein RW, Happich M. The cost-effectiveness of duloxetine in chronic low back pain: a US private payer perspective. Value Health 2013;16:33444.