Letter to the Editor

Abstract

Re: Postma MJ, Kappelhoff BS, van Hulst M, et al. Economic evaluation of dabigatran etexilate for the primary prevention of venous tromboembolic events following major orthopedic surgery in the Netherlands. JME 2012;15(5):878–86.

Dear Editor,

After reviewing this study, it is obvious there is a place for the use of an oral agent for prevention of venous thromboembolism and some cost savings may be appreciated when compared to the use of injectable agents. However, it is difficult to prove this with results from the current study. As a cost minimization study, a thorough analysis of all outcomes should be discussed to assure equivalency1. As Friedman et al.2 report, costs associated with bleeding incidents are significant with these drugs. Bleeding is obviously a major risk with dabigatran use in elderly patients, and it could have a considerable impact on potential costs for the demographic used in this study3. Despite response bias being a common issue with a retrospective survey, it appears that the response rate was actually more than acceptable. It is mentioned that the patients had surgery in 2008, but the time which the patients were interviewed is not mentioned, which may introduce some recall bias to the results. Also, the results report that almost 60% of patients prefer oral over subcutaneous agents, if efficacy and safety would be comparable. It is difficult, in general, to compare one oral drug to several injectable drugs because of the vast differences in administration of the products. A more thorough picture may have been achieved if other oral drugs were included, including warfarin and rivaroxaban. Further, the methods utilized to achieve costing for all items involved are mostly sound, but some figures are questionable. For example, the nursing utilization cost for instruction on use of the subcutaneous agents seems quite high, although it is reported to be based on the instruction protocol. In reality, a nurse would likely not spend more than 10–15 min instructing a patient on these agents. The fact that up to 40 min was accounted for in the sensitivity analysis seems to be an over-estimate. Therefore, it would be ideal to provide a more thorough analysis in these areas to assure that cost savings are truly possible with dabigatran.

Rachel Caggy

Thomas O’Rorke

Priyam Patel

Khalid M. Kamal

Duquesne University, Pittsburgh, PA, USA

References

• Drummond MF, Sculpher MJ, Torrance GW, et al. Methods for the economic evaluation of health care programmes, 3rd edn. New York: Oxford Medical Publications, 2005
   Google Scholar
• Friedman RJ, Dahl OE, Rosencher N, et al. Dabigatran versus enoxaparin for prevention of venous thromboembolism after hip or knee arthroplasty: a pooled analysis of three trials. Thrombs Res 2010;126:175-82
   PubMed Web of Science ®Google Scholar
• Harper P, Young L, Merriman E. Bleeding risk with dabigatran in the frail elderly. NEJM 2012;366:864-6
   PubMed Web of Science ®Google Scholar

Author response to Letter to the Editor

We thank Caggy et al. for their interest in our work. Indeed, of all techniques differentiated in economic evaluation of healthcare, we chose the cost-minimization as the preferred one in this specific situation. Often, this technique is useful in a situation where non-inferiority studies have been designed and performed as the basis for reimbursement. Therefore, in the absence of proven superiority of dabigatran yet proven non-inferiority, cost-minimization could be considered appropriate and valid here1. Motivated by the non-inferiority, focusing on the costs only indeed neglects any potential differences in health effects still existing between dabigatran and enoxaparin. This refers to both (potentially relatively small) differences in venous thromboembolic (VTE) events and in bleeding rates and to differences favoring dabigatran or the comparator2. Notably, recently pooled data on elderly patients show that major bleeding risks for dabigatran were better than for enoxaparin (150 mg) or statistically insignificantly higher (220 mg) with, however, corresponding statistically significantly better efficacy major VTE and VTE-related mortality3. We agree with Caggy et al. that the response rates to our questionnaire were good. Unfortunately, comparison with warfarin and rivaroxaban was impossible at the time of study conduct, as neither was registered in the Netherlands in 2008. Also, the major focus of our study concerned the comparison of an injectable vs an oral administration.

We were not extremely worried about recall bias in the questionnaire as patients who gave informed consent were interviewed only several weeks after surgery, with prior patient information being given according to the procedure approved by the Ethics Committee. All subsequent costing was done according to estimates of time and pricing, with timings derived from protocols, for example, the specific protocol on instructions for self-administration4. Obviously, exact numericals can be argued and might be specific for the Dutch situation in some cases, for example, whether class-room instruction programs are available, etcetera. The Tornado-diagrams illustrate that the results on cost-savings were not highly sensitive to this specific assumption on duration of instruction.

In conclusion, we acknowledge the discussions raised by our colleagues and agree that additional analyses to further substantiate and strengthen our estimates are useful. Also, we argue that there is a place for cost-minimization in economic analyses supporting rational decisions in drug reimbursements.

Maarten J. Postma

Unit of PharmacoEpidemiology & PharmacoEconomics (PE²), Department of Pharmacy, University of Groningen, Groningen, the Netherlands

Bregt S. Kappelhoff

Boehringer Ingelheim, Alkmaar, the Netherlands

Marinus van Hulst

Unit of PharmacoEpidemiology & PharmacoEconomics (PE²), Department of Pharmacy, University of Groningen, Groningen, the Netherlands, and Martini Hospital, Groningen, the Netherlands

Jacobus R. B. J. Brouwers

Department of Geriatrics-Ephor, University Medical Centre Utrecht, The Netherlands, and Unit of Pharmacotherapy & Pharmaceutical Care, Department of Pharmacy, University of Groningen, Groningen, the Netherlands