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. ADDITIONAL DISCLOSURE
  9. REFERENCES
  10. Supporting Information

Objective

To determine if a pharmacist-initiated multidisciplinary strategy provides value for money compared to usual care in participants with previously undiagnosed knee osteoarthritis.

Methods

Pharmacies were randomly allocated to provide either 1) usual care and a pamphlet or 2) intervention care, which consisted of education, pain medication management by a pharmacist, physiotherapy-guided exercise, and communication with the primary care physician. Costs and quality-adjusted life-years (QALYs) were determined for patients assigned to each treatment and incremental cost-effectiveness ratios (ICERs) were determined.

Results

From the Ministry of Health perspective, the average patient in the intervention group generated slightly higher costs compared with usual care. Similar findings were obtained when using the societal perspective. The intervention resulted in ICERs of $232 (95% confidence interval [95% CI] −1,530, 2,154) per QALY gained from the Ministry of Health perspective and $14,395 (95% CI 7,826, 23,132) per QALY gained from the societal perspective, compared with usual care.

Conclusion

A pharmacist-initiated, multidisciplinary program was good value for money from both the societal and Ministry of Health perspectives.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ADDITIONAL DISCLOSURE
  9. REFERENCES
  10. Supporting Information

Despite much research on evidence-based therapy in osteoarthritis (OA), many care gaps remain and patients often experience symptoms without a proper diagnosis and/or proper treatment ([1-3]). Pharmacists have been shown to have the ability to identify community-dwelling adults with undiagnosed OA of the knee and launch interventions that lead to improvements in pain, function, and health-related quality of life ([4, 5]).

The Pharmacist-Initiated Intervention Trial in Osteoarthritis (PhIT-OA) study was a cluster randomized controlled trial designed to investigate whether pharmacists, upon finding untreated knee OA in clients who frequented their pharmacies, could improve quality of care and quality of life by launching a multidisciplinary intervention compared to usual care ([4]). The multidisciplinary intervention included the administration of a validated knee OA screening questionnaire, education, pain medication management, physiotherapy-guided exercise, and communication with the client's primary care physician. The main results of the study were an improvement in the Arthritis Foundation's quality indicators of care pass rate (difference of 45.2%; 95% confidence interval 34.5, 55.9) and significant improvements in pain, function, and quality of life all favoring the multidisciplinary intervention.

Here we present the cost-utility analysis of the PhIT-OA study. Observed quality-adjusted life-years (QALYs) were compared with observed total costs to a third-party payer (the provincial Ministry of Health) and to society, including costs of medication, costs of other health care, costs incurred by the patients, and productivity costs. The aim of this analysis was to show which treatment strategy provides the best value for money in participants with previously undiagnosed knee OA.

Box 1. Significance & Innovations

  • We performed an economic analysis based on a randomized controlled trial whose results were published previously.
  • Empirical measures of costs and quality-adjusted life-years in patients in both arms of the trial were determined.
  • This is the first prospective, randomized study in osteoarthritis to look at whether a pharmacist-initiated, multidisciplinary intervention offers good value for money to payers.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ADDITIONAL DISCLOSURE
  9. REFERENCES
  10. Supporting Information

Patient enrollment criteria have been described in detail elsewhere ([4]). The Clinical Research Ethics Board at the University of British Columbia approved the study protocol and all patients gave written informed consent before inclusion. The primary study end point was an assessment of each participant's overall quality of OA care based on the Arthritis Foundation's quality indicators for the management of OA ([6]). At least 13 pharmacies and 65 patients per randomization group were required to obtain a power of 90% at a significance level of 0.05.

Intervention

Pharmacies were randomly allocated to provide either 1) usual care, which consisted of an educational pamphlet on knee OA created by The Arthritis Society, or 2) intervention care, which consisted of the administration of a validated knee OA screening questionnaire by a pharmacist, education, pain medication management by a pharmacist, physiotherapy-guided exercise, and communication with the client's primary care physician.

Utilities and QALYs

Health state utility values provide a valuation of the patient's health on a scale anchored from 0–1, where 0 = death and 1 = full health. Patients described their quality of life using the Health Utilities Index Mark 3 (HUI3) and the Paper Adaptive Test-5D (PAT-5D), which were administered at baseline and at 3 and 6 months. The PAT-5D is a new, adaptive generic quality of life instrument developed to assess societal preferences for musculoskeletal conditions ([7]). The HUI3 is a generic, preference-based measure used for health assessments that include OA ([8]). The utilities reported by each patient were multiplied by the duration of the trial (6 months) to estimate the number of QALYs experienced.

Costs

Costs were adjusted to and are reported as 2009 Canadian dollars. Costs were included from 2 perspectives: the Ministry of Health (the perspective of the Canadian health care system) and societal. The Ministry of Health perspective included all physician visits and treatments/medications, laboratory tests, and imaging, excluding any visits and treatments paid out of pocket. In addition, for the intervention group patients, 2.5 hours (at $60.86/hour) of a pharmacists' salary was included to account for the initial OA screening, the pain medication management service on a monthly basis, and communication with the primary care physician. Since provincial ministries of health in Canada are funding medication management and enhanced pharmacist activities ([9]), these costs were included in the Ministry of Health perspective. The societal perspective included all Ministry of Health and out-of-pocket costs (e.g., massage therapy, acupuncture, over-the-counter medications) and lost productivity due to knee OA. Also, since physiotherapy is not funded in British Columbia, the costs for physiotherapists (per visit) were included in the societal perspective.

Health care utilization was collected for each patient according to the study protocol. Patients answered questions regarding health care resource utilization at 3 and 6 months after randomization and reported on health care (physician visits, laboratory tests, hospital admissions, imaging studies, medication, and home care), work (work time, absences, and unpaid work), and other societal costs (expenses from knee pain or OA, household help, and informal care). The components of resource utilization included health care professional cost items and equipment cost items. A unit cost was assigned for each component of resource utilization. The prices of health professional visits as well as the cost per equipment item were based on the 2009 British Columbia Medical Services Plan ([10]).

Productivity losses were captured using the human capital approach, where average age- and sex-adjusted earnings were utilized to determine the daily loss of income when reported by participants ([11, 12]).

Statistical analysis

The incremental cost-effectiveness ratio (ICER) was calculated for each of the societal and Ministry of Health perspectives, based both on the PAT-5D and HUI3 instruments. The uncertainty in the estimation of the costs and effectiveness was modeled using nested imputation and bootstrapping. In every economic evaluation, it is important to address how to handle missing data. Complete case analysis is in general inefficient and can lead to bias ([13]). Therefore, it is important to employ an appropriate technique that will permit complete case analysis using the entire data set. The approach suggested by Briggs and colleagues, Oostenbrink et al, and the International Society for Pharmacoeconomics and Outcomes Research was followed ([13-16]). In the PhIT-OA trial, the rate of missing data cases was 14% for costs, 18% for the PAT-5D, and 12% for the HUI3. The missing data were assumed to be missing at random, since it is assumed that the missing values do not depend on the values of unobserved variables. Using the Markov chain Monte Carlo (MCMC) procedure, the missing values were imputed and the complete data set was bootstrapped within each treatment group. The MCMC procedure creates 5 imputed samples and each of these samples is bootstrapped 10,000 times within each treatment group. This procedure was done in the SAS software PROC MI procedure, which is a multiple imputation procedure that creates multiply imputed data sets for incomplete p-dimensional multivariate data ([17]). Total cost and QALYs were averaged for patients within each treatment group within each bootstrap sample. Expected values of the cost and effectiveness outcomes were then averaged over the 5 imputed samples. Lastly, the ICER was calculated for each bootstrap sample. Each of the 10,000 bootstrapped effect and cost differences (between the treatment and control groups) was plotted on the cost-effectiveness plane ([18]). The cost-effectiveness acceptability curves (CEACs) were constructed for each of the perspectives based on both the PAT-5D and the HUI3. The CEACs were constructed by plotting the proportion of the costs and effects pairs that are cost effective for a range of maximum willingness-to-pay values (λ). For comparison purposes, a sensitivity analysis where the data set was restricted to patients with complete data on costs and effectiveness was also conducted.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ADDITIONAL DISCLOSURE
  9. REFERENCES
  10. Supporting Information

The total costs for each health resource item stratified by treatment arm in the PhIT-OA trial are shown in Table 1. Results of the base-case cost-effectiveness analysis, including average costs and QALYs stratified by each of the treatment arms, and the ICERs for both the societal and Ministry of Health perspectives, are reported in Table 2.

Table 1. Total costs (in 2009 Canadian $) for the societal and Ministry of Health perspectives*
 Usual care, societal perspectiveIntervention, societal perspectiveUsual care, Ministry of Health perspectiveIntervention, Ministry of Health perspective
CostNo. of visits/ itemsaNo. (%) of patients (n = 66)CostNo. of visits/ itemsaNo. (%) of patients (n = 73)CostNo. of visits/ itemsaNo. (%) of patients (n = 66)CostNo. of visits/itemsaNo. (%) of patients (n = 73)
  1. Based on the 2009 British Columbia Medical Services Plan. CT = computed tomography; MRI = magnetic resonance imaging.

  2. a

    “Items” pertains to the number of pieces of equipment (aids or devices).

Health professional            
Acupuncturist18522 (3)000 (0)000 (0)000 (0)
Chiropractor2,8355913 (20)1,6353310 (14)000 (0)000 (0)
Family doctor1,491.164434 (52)1,220.043631 (42)1,491.164434 (52)1,220.043631 (42)
Massage therapist1,437.5234 (6)62.511 (1)000 (0)000 (0)
Neurosurgeon161.9411 (2)161.9411 (1)161.9411 (2)161.9411 (1)
Orthopedic specialist396.4844 (6)99.1211 (1)396.4844 (6)99.1211 (1)
Osteopath12011 (2)000 (0)12011 (2)000 (0)
Physiotherapist808124 (6)25,27838751 (70)000 (0)000 (0)
Podiatrist2311 (2)2311 (1)2311 (2)2311 (1)
Rheumatologist163.2411 (2)000 (0)163.2411 (2)000 (0)
Resource            
CT scan000 (0)87.611 (1)000 (0)87.611 (1)
Injection to knee22.7211 (2)000 (0)22.7211 (2)000 (0)
MRI2,68533 (5)3,58044 (5)2,68533 (5)3,58044 (5)
Ultrasound000 (0)66.622 (3)000 (0)66.622 (3)
Radiograph1,415.444139 (59)1,098.93330 (41)1,415.444139 (59)1,098.93330 (41)
Equipment (aids or devices)a            
Brace (ankle, knee, wrist)14233 (5)7043 (4)14233 (5)7043 (4)
Bath bar2211 (2)000 (0)2211 (2)000 (0)
Cane8033 (5)4011 (1)8033 (5)4011 (1)
Heating pads (neck or back)000 (0)000 (0)000 (0)000 (0)
Orthotics62933 (5)30011 (1)62933 (5)30011 (1)
Paraffin heater5011 (2)000 (0)5011 (2)000 (0)
Walk-in shower5,00011 (2)000 (0)5,00011 (2)000 (0)
Costs of paid work, mean ± SD761.85 ± 576.934 (6)994.12 ± 1,006.113 (4)
Costs of paid help, mean ± SD695.06 ± 500.993 (5)996.05 ± 948.326 (8)
Table 2. Results of the base-case cost-effectiveness analysis*
 Societal perspectiveMinistry of Health perspective
Usual care (n = 62), x (95% CI)aIntervention group (n = 73)ICER (95% CI)Usual care (n = 62), x (95% CI)aIntervention group (n = 73)ICER (95% CI)
  1. Costs are measured in 2009 Canadian dollars. 95% CI = 95% confidence interval; ICER = incremental cost-effectiveness ratio; PAT-5D = Paper Adaptive Test-5D; HUI3 = Health Utilities Index Mark 3; QALY = quality-adjusted life-year.

  2. a

    Four patients are missing costs and QALYs at all time points, and therefore were excluded from the analysis.

Cost, based on PAT-5D432 (337, 544)755 (681, 836) 115 (90, 142)120 (90, 154) 
Cost, based on HUI3441 (344, 553)755 (681, 836) 112 (87, 139)120 (90, 154) 
QALY, based on PAT-5D0.4238 (0.4178, 0.4297)0.4469 (0.4428, 0.4506) 0.4237 (0.4177, 0.4296)0.4473 (0.4433, 0.4509) 
QALY, based on HUI30.3646 (0.3549, 0.3738)0.3878 (0.3809, 0.3945) 0.3642 (0.3544, 0.3734)0.3863 (0.3795, 0.3927) 
Using PAT-5D to derive QALY  14,395 (7,826, 23,132)  232 (−1,530, 2,154)
Using HUI3 to derive QALY  14,903 (6,669, 30,116)  431 (−1,604, 2,901)

From the Ministry of Health perspective, the average patient in the intervention group generated $120 (PAT-5D)/$120 (HUI3) in costs compared with $115 (PAT-5D)/$112 (HUI3) for usual care (differences are due to missing values in either the PAT-5D or HUI3). From the societal perspective, the average patient in the intervention group generated $755 (PAT-5D)/$755 (HUI-3) in costs compared with $432 (PAT-5D)/$441 (HUI3) for usual care. With QALYs calculated using HUI3 health state utility values, compared with usual care, the intervention resulted in ICERs of $431 per QALY gained from the Ministry of Health perspective and $14,903 per QALY gained from the societal perspective. Using PAT-5D health state utility values, the intervention resulted in ICERs of $232 per QALY gained from the Ministry of Health perspective and $14,395 per QALY gained from the societal perspective, compared with usual care. Results of the bootstrap/imputation analyses are shown in Figures 1-4 (scatterplots on the cost-effectiveness plane) and Supplementary Figures 1–4 (CEACs; available in the online version of this article at http://onlinelibrary.wiley.com/doi/10.1002/acr.22232/abstract).

image

Figure 1. Scatterplot displaying, on the incremental cost-effectiveness plane, the incremental cost and effectiveness pairs resulting from 10,000 bootstrapped samples comparing the intervention with usual care using the Paper Adaptive Test-5D from the Ministry of Health perspective. Points in the upper right quadrant represent iterations for which the intervention was more costly and more effective than usual care, whereas points in the lower right quadrant represent iterations where the intervention was the dominant strategy (i.e., more effective but less costly). Can$ = Canadian dollars; QALY = quality-adjusted life-year.

Download figure to PowerPoint

image

Figure 2. Scatterplot displaying, on the incremental cost-effectiveness plane, the incremental cost and effectiveness pairs resulting from 10,000 bootstrapped samples comparing the intervention with usual care using the Paper Adaptive Test-5D from the societal perspective. Points in the upper right quadrant represent iterations for which the intervention was more costly and more effective than usual care. Can$ = Canadian dollars; QALY = quality-adjusted life-year.

Download figure to PowerPoint

image

Figure 3. Scatterplot displaying, on the incremental cost-effectiveness plane, the incremental cost and effectiveness pairs resulting from 10,000 bootstrapped samples comparing the intervention with usual care using the Health Utilities Index Mark 3 from the Ministry of Health perspective. Points in the upper right quadrant represent iterations for which the intervention was more costly and more effective than usual care, whereas points in the lower right quadrant represent iterations where the intervention was the dominant strategy (i.e., more effective but less costly). Can$ = Canadian dollars; QALY = quality-adjusted life-year.

Download figure to PowerPoint

image

Figure 4. Scatterplot displaying, on the incremental cost-effectiveness plane, the incremental cost and effectiveness pairs resulting from 10,000 bootstrapped samples comparing the intervention with usual care using the Health Utilities Index Mark 3 from the societal perspective. Points in the upper right quadrant represent iterations for which the intervention was more costly and more effective than usual care. Can$ = Canadian dollars; QALY = quality-adjusted life-year.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ADDITIONAL DISCLOSURE
  9. REFERENCES
  10. Supporting Information

The PhIT-OA study demonstrated that for patients with undiagnosed knee OA, a pharmacist-initiated, multidisciplinary strategy involving case identification, pain medication management, physiotherapist-directed exercise, and involvement of primary care physicians improved quality of care over usual management. In addition, pain, function, and quality of life were superior in those randomized to the intervention than in those randomized to usual care management after 6 months. In the current analysis, we investigated whether the pharmacist-initiated, multidisciplinary strategy represented good value for money compared to usual care from both the societal and Ministry of Health perspectives. From both perspectives, the multidisciplinary strategy appeared to fall within commonly accepted limits of what is considered to be cost-effective use of resources.

However, there is some uncertainty in the findings. For instance, considering the Ministry of Health perspective, there is ∼90% probability that the pharmacist-initiated intervention would be cost effective if the decision-making threshold for a QALY is $2,000. Similarly, from the societal perspective, there is ∼90% probability that the pharmacist-initiated intervention would be cost effective if the decision-making threshold for a QALY is $20,000. For both perspectives, compared to other funded strategies in Canadian health care, even with the observed uncertainty, this intervention appears to offer value for money. Since no other clinical trial has examined a similar intervention for OA of the knee, we believe this is the only information available to guide decision makers on choosing among the 2 options examined here.

Two health state utility value methods were used in this study (the HUI3 and the PAT-5D). Despite being somewhat different in their domains and levels, they appeared to yield similar ICER values when used in the denominator to determine QALYs. This finding is reassuring as to the robustness of the analysis.

Other strengths of this analysis include prospective collection of data on both resource use and effectiveness outcomes, which should have minimized the bias that would result in retrospective data collection. The nested imputation and bootstrapping used in this analysis enabled full incorporation of the uncertainty resulting from missing values and limited sample size of the study.

Our study has a number of limitations. First, the time horizon of the current analysis ended after 6 months, which was in line with the duration of the PhIT-OA study. Since compliance with exercise and other aspects of the intervention may wane over time, it is possible that longer followup may have changed the economic conclusions. Second, the difference in QALYs between the 2 treatment arms tended to be small (∼0.02); however, the 95% confidence limits did not overlap when either the HUI3 or the PAT-5D was used. Third, although we did include absenteeism from work (indirect costs) in our analysis, its inclusion was not a major contributor to the results. Only 3 people reported any work loss secondary to their OA and, as such, the contribution to the average costs was minimal. We did not include reduced productivity while at work (presenteeism), which likely would have favored the intervention arm considering the significant improvements in pain, function, and quality of life experienced by patients in that group. Fourth, our results are likely specific to the Canadian health care system, since different services and practitioners are publicly funded in other countries and patterns of practice tend to differ. Finally, since patients were aware of the allocated group (although assessors were blinded), the results may have been influenced by their treatment preferences.

In conclusion, a pharmacist-initiated, multidisciplinary program was superior to usual care in terms of both quality of care and in improving pain, function, and quality of life. In addition, it appears that this intervention is good value for money from both the societal and Ministry of Health perspectives.

AUTHOR CONTRIBUTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ADDITIONAL DISCLOSURE
  9. REFERENCES
  10. Supporting Information

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. Marra 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. Marra, Cibere, Gastonguay, Esdaile.

Acquisition of data. Marra, Grindrod, Gastonguay.

Analysis and interpretation of data. Marra, Grubisic, Woolcott, Gastonguay.

ADDITIONAL DISCLOSURE

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ADDITIONAL DISCLOSURE
  9. REFERENCES
  10. Supporting Information

Author Woolcott is currently an employee of Pfizer, but was not at the time of the study.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ADDITIONAL DISCLOSURE
  9. REFERENCES
  10. Supporting Information
  • 1
    Glazier RH, Badley EM, Wright JG, Coyte PC, Williams JI, Harvey B, et al. Patient and provider factors related to comprehensive arthritis care in a community setting in Ontario, Canada. J Rheumatol 2003;30:184650.
  • 2
    Grindrod KA, Marra CA, Colley L, Cibere J, Tsuyuki RT, Esdaile JM, et al. After patients are diagnosed with knee osteoarthritis, what do they do?Arthritis Care Res (Hoboken) 2010;62:5105.
  • 3
    Ganz DA, Chang JT, Roth CP, Guan M, Kamberg CJ, Niu F, et al. Quality of osteoarthritis care for community-dwelling older adults. Arthritis Rheum 2006;55:2417.
  • 4
    Marra CA, Cibere J, Grubisic M, Grindrod KA, Gastonguay L, Thomas JM, et al. Pharmacist-initiated intervention trial in osteoarthritis: a multidisciplinary intervention for knee osteoarthritis. Arthritis Care Res (Hoboken) 2012;64:183745.
  • 5
    Marra CA, Cibere J, Tsuyuki RT, Soon JA, Esdaile JM, Gastonguay L, et al. Improving osteoarthritis detection in the community: pharmacist identification of new, diagnostically confirmed osteoarthritis. Arthritis Rheum 2007;57:123844.
  • 6
    MacLean CH, Saag KG, Solomon DH, Morton SC, Sampsel S, Klippel JH. Measuring quality in arthritis care: methods for developing the Arthritis Foundation's quality indicator set. Arthritis Rheum 2004;51:193202.
  • 7
    Pouchot J, Kopec JA, Latteier J, Sayre EC, Al-Harthy F, Garbuz D, et al. Validity of a new adaptive measure of health-related quality of life in arthritis [abstract].Qual Life Res 2004;13:1590.
  • 8
    Blanchard C, Feeny D, Mahon JL, Bourne R, Rorabeck C, Stitt L, et al. Is the Health Utilities Index responsive in total hip arthroplasty patients?J Clin Epidemiol 2003;56:104654.
  • 9
    Canadian Pharmacists Association. Summary of pharmacists' expanded scope of practice activities across Canada.2013. URL: http://blueprintforpharmacy.ca/docs/kt-tools/pharmacists'-expanded-scope_summary-chart–cpha–october-2013.pdf.
  • 10
    Medical Services Plan, Medical Services Commission.Payment schedule.British Columbia:Ministry of Health, Government of British Columbia;2009.
  • 11
    Liljas B. How to calculate indirect costs in economic evaluations. Pharmacoeconomics 1998;13:17.
  • 12
    Drummond MF, O'Brien B, Stoddart GL, Torrance GW.Methods for the economic evaluation of health care programmes.Oxford:Oxford University Press;1997.
  • 13
    Briggs A, Clark T, Wolstenholme J, Clarke P. Missing…presumed at random: cost-analysis of incomplete data. Health Econ 2003;12:37792.
  • 14
    Oostenbrink JB, Al MJ. The analysis of incomplete cost data due to dropout. Health Econ 2005;14:76376.
  • 15
    Oostenbrink JB, Al MJ, Rutten-van Molken MP. Methods to analyse cost data of patients who withdraw in a clinical trial setting. Pharmacoeconomics 2003;21:110312.
  • 16
    Ramsey S, Willke R, Briggs A, Brown R, Buxton M, Chawla A, et al. Good research practices for cost-effectiveness analysis alongside clinical trials: the ISPOR RCT-CEA Task Force. Value Health 2005;8:52133.
  • 17
    Yuan YC. Multiple imputation for missing data: concepts and new development. Version 9.0.1999. URL: http://www.ats.ucla.edu/stat/sas/library/multipleimputation.pdf.
  • 18
    Briggs A, Fenn P. Confidence intervals or surfaces? Uncertainty on the cost-effectiveness plane. Health Econ 1998;7:72340.

Supporting Information

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. AUTHOR CONTRIBUTIONS
  8. ADDITIONAL DISCLOSURE
  9. REFERENCES
  10. Supporting Information

Additional Supporting Information may be found in the online version of this article.

FilenameFormatSizeDescription
ACR_22232_sm_SupplFigures.docx25KSupplementary Figures

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.