RSS Feeds
-
THE STATUTORY DRUG COMPENDIA ARE NOT JUST OF CONCERN TO ONCOLOGISTS
Submit responseTo the Editor: The triad of articles and accompanying editorial on the statutory drug compendia that appeared in the March 3, 2009 issue of Annals (1- 4) are welcome additions to the literature on this important topic. In particular, we wish to applaud the work of Abernethy et al. in solidly indentifying some of the inadequacies of the compendia in evaluations of off-label indications for products used in oncology, and would like to make a few comments that extend concern to the mental health field.
Prompted by recent critical data pertaining to off-label prescribing in the medical literature (5), and related resource utilization and cost concerns in our practices, we recently reviewed citations in DrugDex for the seven “atypical” antipsychotics currently on the market. To its credit, Thomson-Reuter’s DrugDex is the only available compendium, recognized by the US federal government as a guide to reimbursement decisions, which provides detailed evaluations for off-label indications of psychoactive drugs. We were surprised to find that there were 85 separate off-label indications listed for these agents, more than 10 per drug on average. However, we were disappointed to find that, while over 90% of these off-label indications for atypical antipsychotics received favorable ratings that would likely make them eligible for reimbursement, the great majority of the citations supporting these indications were case reports, case series and open-label observational studies. What we would consider high quality controlled data was seldom cited. Frankly, many of the citations entered into DrugDex in support of off-label indications for atypical antipsychotics are reminiscent of the “evidence” often presented in pharmaceutical company detailing.
There are the practical implications of these observations. In 2008, IMS-Health, a corporation that tracks the volume and dollar value of physician prescriptions, reported that antipsychotic drugs are now the number one revenue-generating therapeutic class, with a 14.6 billion dollar US market, outstripping for the first time even lipid-lowering agents and proton pump inhibitors (6). These revenues are comparable to the combined revenues from high-ticket items like antineoplastic monoclonal antibodies, erythropoietins and some of the biological response modifiers (6).
Although manufacturer revenues cannot be equated directly with health system costs, these financial data speak volumes. The content of the statutory compendia are currently of primary concern to those who prescribe and reimburse prescriptions for oncologics, but there seems little doubt that psychoactive drugs, and expensive antipsychotic drugs in particular, will be the focus of future utilization management efforts by administrators of both public and privately supported health care, particularly with the advent of Medicare Part D prescription drug benefits.
Given the likely importance of the compendia system in future reimbursement decisions across therapeutic classes, it is critical that evaluations of off-label indications in the compendia be based on solid clinical research, and whenever possible controlled observations, free as possible from commercial bias. We join with the authors (1 - 4) in advocating greater physician involvement in the development and use of independent, evidence-based compendia, and evidence-based regulatory reform of off-label drug use in general.
References
1. Abernethy AP, Raman G, Balk EM, et al. Reliability of compendia methods for off-label oncology indications. Annals of Internal Medicine. 2009; 150: 336 - 343.
2. Tillman K, Burton B, Jacques LB, Phurrough SE. Compendia and anticancer therapy under Medicare. Ann Intern Med. 2009; 150: 348 - 350.
3. Gillick MR. Controlling off-label medication use. Ann Intern Med. 2009; 150: 344- 347.
4. Sox HC. Evaluating off-label uses of anticancer drugs: time for a change. Editorial in Annals of Internal Medicine. 2009; 150 (5): 353- 354.
5. Walton SW, Schumock GT, Lee K-V, Alexander GC, Meltzer D, et al. Prioritizing future research on off-label prescribing: Results of a quantitative evaluation. Pharmacotherapy. 2008; 28(12):1443.
6. IMS Health. Top Therapeutic Classes by US Sales. Top-line Industry Data. 2008; http://www.imshealth.com/portal/site/imshealth/menuitem.a46c6d4df3db4b3d88f611019418c22a/?vgnextoid=85f4a56216a10210VgnVCM100000ed152ca2RCRD&cpsextcurrchannel=1
Conflict of Interest:
Dr. Kruszewski does not have any current business or financial arrangements with any pharmaceutical company. Prior to 2000, Dr. Kruszewski participated on the speaker’s bureaus of the following companies: Pfizer, Inc. GlaxoSmithKline, Janssen (Johnson and Johnson), AstraZeneca, Wallace Labs, Eli Lilly, GE-Amersham Biosciences; and previously Served on an Eli Lilly Northeast Advisory Panel (1998.) Dr. Kruszewski has been an expert in complex litigation involving OxyContin (2005-2006), Neurontin (2004-present) and, recently in multiple jurisdiction litigation for olanzapine with 10 US states’ Attorneys General (2008-present). Dr. Kruszewski’s false claim litigation against Southwood Psychiatric was intervened upon and settled by the United States Department of Justice in 2009. Dr. Kruszewski testified in 2008 Daubert Hearing, Federal Court, Boston MA, regarding the mechanism of action and harm regarding gabapentin; and he again testified in the Frye Hearing in New York state (2009) regarding the ability of sertraline to cause adverse mood and behaviors. Dr. Kruszewski has less that $25,000.00 holding in Celgene Pharmaceuticals. Dr. Paczynski does not have any current business or financial arrangements with any pharmaceutical company. He had brief experience as a speaker for Astra-Zeneca on two occasions in 2002 and 2003.
-
"Author's response"
Submit responseDr. McEvoy cites two concerns in his rapid response: (1) In Table 3, we do not show that AHFS DI included reviews of bevacizumab for breast or lung cancer, in 2006; and, (2) In Table 3, we do not show that AHFS DI included reviews of docetaxel for gastric or ovarian cancer, or rituximab for chronic lymphocytic leukemia, in addition to the 2 bevacizumab indications, in 2008 (1).
Over the past 3 days we have carefully reviewed the communications and available source documents for Table 3.
Mr. McEvoy is correct that the indications for bevacizumab we list are contained in the 2005 print edition of AHFS-DI, however, our assessment was based on the bevacizumab that we accessed as late as February 17, 2006 when we also accessed the AHFS-DI through STAT!Ref Online Medical Electronic Library. We had also checked the ASHP web site for updated monographs which are ccasionally posted for AHFS-DI subscribers. Furthermore, neither our major medical center library’s nor our own purchased copy of the 2006 AHFS-DI “red book” had yet arrived. Thus the 2006 version of AHFS-DI was effectively unavailable to us.
Mr. McEvoy was aware of these data when we presented them to the public forum presentation to the MedCAC and CMS in March 2006. Further, this table, in the 2006 version, was available in the original Technology Assessment white paper report available on the AHRQ website from May 2007. He previously levied the same criticism about the bevacizumab indications and received the same response from us.
While we have no reason to doubt that the editors of AHFS-DI had reviewed the information and concluded that bevacizumab was indicated for breast and lung cancer by the time we reviewed their publication, never- the-less, Mr. McEvoy fails to acknowledge that the publication and distribution of this information to the consumer had not taken place by the date of our assessment (up to Feb 17, 2006). Indeed, our assessment of the compendia was designed to evaluate not only the quality of the information contained, the quality of the editorial process used to generate it, but also its timeliness to the consumer or end user. Thus we stand by our assessment of the indications listed in Table 3 as of 2/17/2006.
For the 2008 update we had access to all of the compendia documents except AHFS DI; I have source documentation to confirm the inputs for Table 3. On May 22, 2008, we were working to get the AHFS DI information for the updated review. Dr. McCrory had access to AHFS DI Essentials (which is less comprehensive than the print edition) through an institutional subscription at the Veterans Administration Hospital (Durham, NC), but that was not enough to finish the tables. Hence, we chose to contact AHFS directly. According to e-mail communications, on July 7, 2008, Kelly Moore, RN, a research assistant with our team, spoke with AHFS editorial staff via telephone to explain what we needed and why. AHFS then sent Ms. Moore the documents that were used to generate the tables in the paper; hence, these tables were updated with information sent to us by AHFS directly. I do not have access to the information that AHFS sent to Ms. Moore; however, several days later Ms. Moore sent me updated tables for the manuscript, including the table that later became Table 3. Tracked changes in the document demonstrate how Ms. Moore made updates to the table in response to information sent by AHFS in July. She did not show any updates to the AHFS DI column in what is labeled Table 3 in the published manuscript.
As a result of this review, I note a potential correction for Table 1. Our manuscript indicates that the 2008 update information for AHFS DI came from an institutional subscription viewed on 7 July 2008. In fact, according to these source documents, the more accurate statement is that we viewed AHFS DI Essentials on 22 May 2008 and had direct communication with the publisher for supplemental data on 7 July 2008.
References
1. McEvoy, G. Letter. Ibid Pg. 336
Conflict of Interest:
None declared
-
Correction Request
Submit responseIn a recent article by Abernethy and colleagues (1), there are errors in Table 3 for the agent and cancer combination indications listed for the American Hospital Formulary Service Drug Information (AHFS DI) compendium. In table 3, the total number (n) of indications in this compendium should be corrected to be 7 not 2. This correct number reflects the changes in reviews published in the AHFS DI compendium between 2006 and 2008, the timeframe defined by the authors. The indications for bevacizumab, which included both breast and lung cancer, were omitted from the table but were published in the 2006 version of the AHFS DI. Likewise, the indications for docetaxel (gastric and ovarian) and rituximab (chronic lymphocytic leukemia) were omitted from the table but were included in the February 2008 version of the AHFS DI.
References
1. Abernethy AP, Raman G, Balk EM, Hammond JM, Orlando LA, Wheeler JL et al. Reliability of compendia methods for off-label oncology indications. Ann Intern Med. 2009 Feb 19.
Conflict of Interest:
None declared
-
Compendia Study Fails to Measure Reliability
Submit responseAbernethy and colleagues take a very timely look at drug compendia in their recent article (1). As one of the compendia examined in the article, we are concerned that the methods—which focus on quantitative rather than qualitative analysis—fail to appropriately assess the reliability of compendia (1,2). Comparing the number of off-label uses and of cited references included in each compendium ignores the essential question of whether the information provided in each is high quality. Table 3 (1) and the more detailed original 2006 study (2), for example, provide little information on the actual quality and validity of cited evidence or whether it substantiates medical acceptance. In addition, the study failed to assess whether the recommendations made by each compendium were congruent with the cited evidence.
Compendia should include clinically relevant information based on carefully selected studies in order to determine medical acceptance. Devoting resources to a highly quantitative approach (i.e., adding numerous citations) can ultimately detract from the scope and usefulness of compendia in determining medical acceptance, yet the authors conclude that compendia should be judged by the proportion of available evidence cited.
The authors criticize all the compendia for not citing a follow-up report (3) to the original phase III trial (4) establishing gemcitabine’s use in bladder cancer. However, it is important to recognize that citing the additional reference (3) does not change the determination of medical acceptance, which was established by the original citation (4). Rather than cataloguing all applicable published references, determining medical acceptance is the statutory charge for compendia (5). Does the more recent reference add to the weight of evidence? Yes. Does it change the medical acceptance and CMS coverage decision? No. When new evidence substantially changes medical acceptance (e.g., survival benefits, toxicity), it should be included.
We also take exception to the authors’ discussion regarding the usefulness of this study as a benchmark for future evaluation of compendia. Future studies should not measure improvement in compendia by the proportion of evidence cited at the expense of focusing on the quality of evidence. Instead, the benchmark that Congress established, and that CMS is charged with implementing, is whether the compendia appropriately reflect current medical acceptance.
The stakes regarding off-label uses are enormous for both patients and the entire health-care system. Therefore, it is critical that the processes that compendia use to evaluate off-label uses for oncology medications meet the highest standards.
References
1. Abernethy AP, Raman G, Balk EM, Hammond JM, Orlando LA, Wheeler JL et al. Reliability of compendia methods for off-label oncology indications. Ann Intern Med. 2009 Feb 19 [Epub ahead of print; PMID: 19221366]
2. Abernethy AP, Hammond JM, Hubbard ML, Patwardhan MB, Orlando LA, McCrory DC, et al.; Agency for Healthcare Research and Quality. Compendia for coverage of off-label uses of drugs and biologics in an anti-cancer chemotherapeutic regimen. Rockville, MD: Agency for Healthcare Research and Quality; 2007. Assessed at www.cms.hhs.gov/determination process/downloads/id46TA.pdf on 10 February 2009.
3. von der Maase H, Hansen SW, Roberts JT, Dogliotti L, Oliver T, Moore MJ, et al. Gemcitabine and cisplatin versus methotrexate, vinblastine, doxorubicin, and cisplatin in advanced or metastatic bladder cancer: results of a large, randomized, multinational, multicenter, phase III study. J Clin Oncol. 2000;18:3068-77. [PMID 11001674]
4. von der Masse H, Sengelov L, Roberts JT, Ricci S, Dogliotti L, Oliver T, et al. Long-term survival results of arandomized trial comparing gemcitabine plus cisplatin, with methotrexate, vinblastine, doxorubicin, plus cisplatin in patients with bladder cancer. J Clin Oncol. 2005;23:4602 -8. [PMID: 16034041]
5. Tillman K, Burton B, Jacques LB, Phurrough SE. Compendia and anticancer therapy under Medicare. Ann Intern Med. 2009 Feb 19 [Epub ahead of print; PMID: 19221368]
Conflict of Interest:
None declared