| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Department of Anesthesia, Indiana University School of Medicine, Indianapolis, Indiana.
Address correspondence to Thomas R. Vetter, MD, Department of Anesthesia, Riley Hospital for Children, Room 2001, 702 Barnhill Drive, Indianapolis, Indiana 46202. Address e-mail to tvetter{at}iupui.edu.
Abstract
BACKGROUND: The relative efficiency of a health care intervention or health status improvement realized for a given amount of resources expended can be determined using cost-effectiveness analysis or cost-utility analysis.
METHODS: An extensive chronic pain-focused search was undertaken of the MEDLINE, EMBASE, and SCI-EXPANDED databases. A total of 1822 unique citations were generated, with 142 studies subsequently categorized as incorporating one of seven recognized types of health care economic evaluation.
RESULTS: Of the 142 identified chronic pain-related economic evaluations published between 1988 and 2006, 30 incorporated a cost-effectiveness analysis and 29 incorporated a cost-utility analysis. The data are consistent with the previously reported chronological pattern of an increased overall diffusion of cost-utility analysis studies from the general medical and health services research literature into the medical subspecialty journals. However, only a few studies combined the economic analysis alongside a randomized controlled trial, the economic end-points in the trials had limited time horizons, and there was failure to address the protracted costs versus benefits of treating long-term and often recurrent chronic pain conditions.
CONCLUSIONS: Although it would appear worthwhile for researchers and clinicians to consider cost-effectiveness analysis and cost-utility analysis in their trial designs and treatment algorithms for chronic pain conditions, methodological improvements can be made in trial designs.
The relative efficiency of a health care intervention, defined as the health status improvement realized for a given amount of resources expended, can be determined using cost-effectiveness analysis (CEA) or cost-utility analysis (CUA) (1–4). CUA is a form of CEA in which the clinical outcome is measured in quality-adjusted life years (Table 1) (5,7,8). This study was undertaken to assess the historical application of health care economic evaluation methods, specifically CUA, in the chronic pain medicine literature.
|
METHODS
A comprehensive search was undertaken in June 2006 of the PubMed (MEDLINE) database (1966–present). The search strategy combined the MeSH-terms ("Costs and Cost Analysis" OR "Cost-Benefit Analysis" OR "Cost of Illness") AND ("Facial Pain" OR "Low Back Pain" OR "Back Pain" OR "Myofascial Pain Syndromes" OR "Pain, Intractable" OR "Complex Regional Pain Syndromes" OR "Facial Neuralgia" OR "Facial Pain" OR "Neuralgia" OR "Reflex Sympathetic Dystrophy" OR "Headache" OR "Causalgia" OR "Arthralgia" OR "Fibromyalgia"). A second search was performed of the Excerpta Medica Database (EMBASE) (1980–present) using the same MeSH-terms so as to identify any other pertinent chronic pain-related economic analysis studies. A third search was done of the Science Citation Index Expanded (SCI-EXPANDED) database (1987–present) using the keywords ("Cost-Effectiveness" AND "Pain"). Finally, reference lists were inspected of 10 recently published systematic reviews, focusing on the application of economic analysis in the specific areas of complementary and alternative medicine, low back pain, rheumatology, migraine, spinal cord stimulation, and multidisciplinary pain management.
No attempt was made to identify unpublished primary studies or findings presented only in abstract form. Neither expert opinions nor data from the pharmaceutical industry were elicited. No hand-searching of relevant journals was performed. Only English language studies were identified so as to readily permit adequate study interpretation.
A three-step assessment of the initially retrieved citations was undertaken (Fig. 1). The retrieved citations were compiled in EndNote® 9.0, and duplicate citations were noted. The published abstracts from the 1822 unique citations generated via the combined MEDLINE, EMBASE, and SCI-EXPANDED searches were individually read by the author. A full-text review of the content of the 162 potentially relevant studies resulted in 142 being identified as incorporating 1 of the 7 previously recognized types of health care economic evaluations (5). These 142 studies were categorized into one of the seven types of partial or full economic evaluation, specifically: a cost-description, cost-outcome description, cost analysis, cost-minimization analysis, cost-benefit analysis, cost-effectiveness analysis, or cost-utility analysis (Table 1) (5,9).
|
RESULTS
In contrast to a modest increase in the number of chronic pain-related health care economic evaluation studies published between 1988 and 2000, there has been a marked increase in the last 5 yr. More rigorous analysis methods have also been used (Fig. 2).
|
Of the 142 identified chronic pain-related economic evaluations that were published between 1988 and 2006, 30 incorporated a CEA and 29 incorporated a CUA. Of the 29 chronic pain-related CUA, 10 relied upon previously published cost and outcome data and applied either basic decision analytic modeling or Markov modeling. Of the 18 (62%) that conducted a stochastic CUA alongside a randomized controlled trial, 14 (78%) were published together with their respective clinical outcome data. The other four clinical trial-based CUA were published in different journals, from 2 to 7 yr apart. In 17 of the 18, the time horizon of the economic evaluation, and hence the included costs were predicated upon and limited to the duration of the concomitant clinical trial. Moreover, the majority (78%) limited their time horizon to 1 yr or less, several intentionally, so as to avoid the issue of discounting of costs and quality-adjusted life years. As a result, the 18 had a relatively short time horizon (median of 12 mo, range of 4–36 mo).
Sensitivity analyses were included in all 10 of the deterministic decision analysis and Markov modeling studies, but in only 11 of the 18 stochastic CUA.
None of the 29 chronic-pain related CUA studies identified in the present literature search was multinational in scope. However, two of the CUA studies did use a purchasing power parity conversion factor, one from Dutch gilders to United States dollars for early rheumatoid arthritis therapy and the other from Canadian dollars to Euros for spinal cord stimulation for failed back surgery syndrome.
DISCUSSION
The number of deterministic and stochastic chronic pain-related CUA appears to be increasing. The present data are consistent with the previously reported S-shaped chronological pattern of the overall increased diffusion of CUA from the general medical and health service research literature into the corresponding medical subspecialty journals (10). However, it is worth noting that only one of the 29 identified chronic pain-related CUA was published in a dedicated pain medicine journal and none in an anesthesiology journal; this despite the significant contribution by these specialties to chronic pain clinical care and research.
There are two fundamental approaches to conducting a CEA or CUA (11,12). In deterministic decision analytic modeling, the required cost and clinical outcome data are retrospectively obtained from existing sources and coalesced (13,14). Alternatively, in the stochastic "piggy-backed" or trial-based approach (15,16), the economic data are prospectively collected alongside a "pragmatic" randomized controlled trial that seeks to mimic real life (12,17). Using either approach, the economic evaluation results can be reported as an incremental cost-effectiveness ratio that is plotted on the cost-effectiveness plane (18) and/or as a cost-effectiveness acceptability curve (19,20). A sensitivity analysis of the base-case scenario assumptions should be undertaken and reported in almost all circumstances (21).
Conducting a CUA alongside or within a clinical trial has only recently become more common (22). This was borne out by a systematic review of 533 CUA published between 1976 and 2001, which revealed that only 8% had been performed alongside a clinical trial (23). A key issue with such conjoint studies is the concomitant reporting of the clinical outcome and the economic evaluation findings. Of the 8%, 84% were published separately, on average nearly 2 yr after the associated clinical findings, and in journals with significantly lower readership and influence (i.e., journal impact factor) (23). This contrasts markedly with the 78% joint publication rate among the 18 presently identified chronic pain-related stochastic CUA. This may be due in part to changes in editorial policies such as at the British Medical Journal whose editor, in 2002, advised would-be authors to "Send us your clinical and economic results together, and we will be delighted. Send somebody else your clinical results and us your economic results, and we will send them back, politely" (24).
Unlike many cardiology and oncology trial-based economic evaluations with societal perspectives and protracted time horizons (e.g., longer-term survival versus death), most of the 18 identified stochastic chronic pain-related CUA applied a third-party insurer perspective and used a limited time horizon of <1 yr (25). While admittedly a matter of practicality, this methodology resulted in a failure to address the protracted costs versus benefits of treating long-term and often episodic or recurrent chronic pain conditions (26).
Despite these methodological and cognitive limitations, there appears to be a place for continued CEA and CUA in chronic pain medicine research (27,28). It would also appear worthwhile for clinicians to consider any available and pertinent CEA and CUA findings in their treatment algorithms for chronic pain conditions (6,29). To do otherwise will likely disenfranchise invested pain medicine researchers and clinicians from the prioritization and decision-making processes that are already underway within governmental agencies and third-party payers and that are aimed at identifying how to optimally allocate finite financial resources in the face of a steadily increasing demand for health care services (30). Moreover, the next logical step in the advancement of health care appears to be the melding of the well established principles of evidence-based medicine with both patient-centered outcomes and CUA data so as to create value-based medicine (31).
ACKNOWLEDGMENTS
The author thank Victoria L. Phillips, DPhil, and Kimberly J. Rask, MD, PhD, in the Department of Health Policy and Management at the Rollins School of Public Health, Emory University, Atlanta, GA, for their scholarly support.
Footnotes
Accepted for publication February 13, 2007.
Reprints will not be available from author.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
