The Value of Specialty Pharmaceuticals – A Systematic Review
Published Online: June 19, 2014
Martin Zalesak, MD, PhD; Joyce S. Greenbaum, BA; Joshua T. Cohen, PhD; Fotios Kokkotos, PhD; Adam Lustig, MS; Peter J. Neumann, ScD; Daryl Pritchard, PhD; Jeffrey Stewart, BA; and Robert W. Dubois, MD, PhD
Novel specialty pharmaceuticals—typically biological therapies which may cost tens of thousands of dollars for a course of treatment—hold great promise for patients living with complex and chronic conditions. 1 The improved efficacy and the potential to redefine treatment modalities, however, are not without cost. High research and development costs, special handling and distribution networks, and necessary enhancements to patient support programs all contribute to the high price of specialty pharmaceuticals.2 Due in part to the high and often rising cost of these products, payers increasingly demand evidence of their value.3,4 It is predicted that by 2017 specialty pharmaceuticals will represent more than half of total pharmaceutical sales, intensifying the need to clearly understand their clinical and functional value.5
To holistically characterize the value of specialty pharmaceuticals, one must look beyond cost and take into account benefits. To this end, we systematically reviewed published studies involving specialty pharmaceuticals for the top 3 disease areas by pharmaceutical spending in the United States, namely rheumatoid arthritis (RA), multiple sclerosis (MS), and breast cancer (BC). For each disease area, we compared the clinical and functional efficacy of specialty pharmaceutical treatments with conventional therapies representing the previously available standard of care. To evaluate therapeutic benefits in an economic context, we also reviewed available cost-effectiveness findings for the specialty pharmaceuticals. This combined economic and clinical approach allowed us to comprehensively assess the value of specialty pharmaceuticals from both a patient and payer perspective.
Definition of Specialty Pharmaceuticals
A variety of definitions exist for “specialty pharmaceutical.” We performed a literature search to identify a “specialty pharmaceutical” definition that was well accepted in the medical literature and would provide a reasonable framework for our review. Based on articles reviewed,1-4,6-9 we defined specialty pharmaceuticals to be pharmaceutical treatments that: (1) are high cost (generally accepted as having prescription price exceeding $600 per month); (2) require close monitoring, including personalized or frequent adjustment of dosing; and (3) require special handling, such as careful temperature control, or restrictions on where the medication can be administered, prepared, or distributed.
Selection of Disease Areas
We selected disease areas by first building a comprehensive list of specialty pharmaceuticals marketed in the United States and identifying the corresponding disease area(s) for each therapy. A full list of specialty pharmaceuticals appears in eAppendix A (available at www.ajmc.com). We excluded orphan diseases because they typically lack a conventional therapy with which the specialty pharmaceutical can be compared. We excluded diseases for which specialty pharmaceuticals provide only acute or supportive care because our goal was to evaluate specialty pharmaceuticals in cases where they incur the highest costs—namely when used chronically.
Next, we reviewed market research reports and other materials to find the top 10 disease areas by specialty pharmaceutical spend in the United States. We initially focused our analysis on the 4 top areas in this list but ultimately excluded the disease area with the fourth-highest pharmaceutical spend—human immunodeficiency virus (HIV)—because we could find no studies comparing specialty therapies for HIV with nonspecialty therapies. Since HIV therapies were not available prior to the introduction of specialty pharmaceuticals to treat this disease, specialty pharmaceutical benefits to HIV patients have been substantial. Exclusion of HIV left 3 disease areas: rheumatoid arthritis (RA), multiple sclerosis (MS), and breast cancer (BC).
Selection of Pharmaceuticals
For each of the 3 disease areas, we included specialty pharmaceuticals if they were: (1) marketed in the United States; (2) identified as a specialty pharmaceutical per our constructed definition; and (3) specified by guidelines for treatment in the disease area at time of publication.
Selection of Metrics
For each of the 3 disease areas, we selected 1 clinical metric (eg, efficacy) and 1 functional metric (eg, quality of life). For the clinical metric, we first identified outcomes mentioned in clinical trials catalogued in the US government’s clinical trials registry for the pertinent disease area. We next searched PubMed (National Center for Biotechnology Information, US National Library of Medicine, Bethesda, Maryland) for articles that used each metric and also mentioned both a specialty pharmaceutical in that disease area and the disease area itself. We retained the clinical end point used in the greatest number articles as the end point for our literature review. We selected the functional metric using a similar process but the initial list of metrics was determined from review articles focused on functional outcomes in the disease area.10-17 Our economic evaluation focused on analyses estimating cost-per-QALY (quality-adjusted life year), a gold standard metric in the health economics literature.18
We identified original clinical and functional outcome articles for each disease area from review articles. We identified review in PubMed using the search phrase: “(disease AND metric)” and applying PubMed’s systematic review filter. To this list, we added original articles published too recently to appear in the reviews. We identified these articles from a supplemental search of PubMed that omitted the systematic review filter. We excluded original articles if they were: (1) published before January 1990; (2) not in English; (3) not an original article; (4) not relevant (ie, did not provide original outcome data associated with the relevant drugs) based on the title or abstract; or (5) not available for electronic download. Articles were likewise excluded if they: (6) did not report quantitative results; or (7) did not compare the selected specialty pharmaceuticals with nonspecialty treatments.
We identified cost-effectiveness articles from the Tufts Medical Center Cost-Effectiveness Analysis Registry,19 which contains 3488 cost-utility analyses on a wide variety of diseases and treatments. For each disease area, we searched the registry for articles evaluating pertinent specialty pharmaceuticals. Articles that did not compare a pertinent specialty drug with a conventional treatment were excluded.
Two reviewers independently examined each clinical and functional metric article to determine if its results were positive (favorable), negative (unfavorable), or neutral. Results were deemed positive if they indicated that the specialty drug confers a statistically significant improvement relative to the alternative conventional treatment; negative if they indicated a statistically significant advantage for the conventional treatment; or neutral if they did not achieve statistical significance, regardless of the direction of the finding.
Given the heterogenous nature of the studies reviewed, a qualitative, as opposed to quantitative, assessment of the literature was necessary. Patient populations, time frames, treatment administration, and methods of measuring outcomes across studies were diverse, precluding the use of meta-analytic techniques. However, the more standardized nature of the cost-effectiveness papers allowed quantitative aggregation of results, and subsequent reporting of summary statistics.
We summarized economic results by reporting the mean, median, minimum, and maximum cost-effectiveness ratios reported by pertinent articles. Because each article can report multiple ratios (eg, for different populations, or evaluating different doses of the specialty pharmaceutical), we reported both unweighted and weighted summary statistics. For the weighted results, we assigned each ratio a weight of 1/n, where n is the number of ratios reported by that article. Doing so equalizes the statistical contribution from each article regardless of how many ratios it reports.
Case Example Disease Areas
The literature review, along with an examination of payer materials,2,8,9 yielded 266 specialty pharmaceuticals corresponding to 85 disease areas marketed in the United States. Exclusion of 31 orphan diseases and 22 conditions for which specialty pharmaceuticals provide only acute or supportive care yielded 32 diseases for inclusion. The specialty pharmaceuticals and disease areas we evaluated are listed in eAppendix B. Table 1 reports the estimated spend on specialty pharmaceuticals for the top 10 disease areas, as determined from various analyst and market research reports.20-29 The top 3 areas—RA, MS, and BC—were identified as the targets for our analyses.
PDF is available on the last page.