2576 - Are Biologics Cost-Effective? Analysis Based on Real-Life Rheumatoid Arthritis Patients 

Wednesday, November 9, 2011: 9:45 AM
W194b (McCormick Place West)
Hawre Jalal1, Kaleb Michaud2, Hyon K. Choi3, Young Hee Rho3 and Karen Kuntz1, 1University of Minnesota, Minneapolis, MN, 2National Data Bank for Rheumatic Diseases & University of Nebraska Medical Center, Omaha, NE, 3Boston University School of Medicine, Boston, MA
Presentation Number: 2576

Background/Purpose: Previous cost-effectiveness analyses (CEAs) of biologics for rheumatoid arthritis (RA) based on randomized controlled trials (RCTs) have been limited by short time horizon, failure to properly incorporate adverse events (AE) and discontinuation rates, and poor generalizability to real-world effectiveness.  The cost-effectiveness of biologics in real-life RA patients has not been adequately explored.  Our objective was to conduct CEAs of several commonly adopted treatment strategies involving sequential use of biologics based on an observational study.  

Method: We developed a Markov simulation model to estimate quality-adjusted life years (QALYs) and costs associated with several treatment strategies for RA patients over the lifespan.  We evaluated strategies based on different sequences of 5 biologics (etanercept, infliximab, adalimumab, abatacept, and rituximab) compared to conventional non-biologic DMARDs.  We modeled the discontinuation rate of each biologic as well as the reason (i.e. serious AE, cost of medication, or ineffectiveness).  Those who discontinued because of ineffectiveness were allowed to transition to the next biologic in the sequence based on the probability of switching biologics in an observational study (the National Data Bank [NDB]).  Those who discontinued because of the other reasons switched to nonbiologics.  Markov health states were defined by Health Assessment Questionnaire (HAQ) scores and the types of medication the patients received. Transition probabilities, effectiveness measures (i.e., HAQ score improvements), AE rates, quality of life weights, and discontinuation rates were estimated from the NDB.  Direct and indirect costs (2009 US dollars) were obtained from the literature.  Both costs and effectiveness are discounted by 3% annually.  

Result: Biologics were generally more effective than nonbiologics.  The most cost-effective sequence of biologics was Etanercept -> Infliximab -> Adalimumab (EIA), with an incremental lifetime effectiveness of 0.1 QALYs compared to nonbiologics.  The EIA strategy had an incremental lifetime cost of about $150,000 yielding an incremental cost-effectiveness ratio (ICER) >$1M/QALY, which is well above willingness to pay (WTP) thresholds considered acceptable.  In a sensitivity analysis, we increased the HAQ improvement from rates reported in the NDB to those typically reported in trials (Table).  Biologic strategies became cost-effective at WTP thresholds of $200K/QALY.  However, this ICER is still higher than those reported in RCT-based CEAs.

Conclusion: We used real-life observational data to model sequences of biologic use, discontinuation rates, and AEs. Using these observational data, we were unable to observe the cost-effectiveness reported by RCT-based CEA.  The primary reasons for this discrepancy relate to a lower incremental effectiveness of biologics in real-life compared to RCTs.  


Table. Cost-Effectiveness of Sequential Biologic Use (EIA) According to Assumptions

Sensitivity Analysis


Cost ($)

Incremental  Cost ($)

Effectiveness (QALY)

Incremental Effectiveness (QALY)


Primary Model












A + B






A + C






A + D












A = HAQ progression rates are equivalent to those in RCTs of biologics.

B = There is additional mortality due to serious adverse events.

C = Study population's HAQ ranges between 1.1 and 2.0.

D = Study population's HAQ ranges between 2.1 and 3.0.



Keywords: anti-TNF therapy, economics and rheumatoid arthritis (RA)

Disclosure: H. Jalal, None; K. Michaud, None; H. K. Choi, Takeda, 2, Novartis Pharmaceutical Corporation, 5, URL, 2, Savient, 5, Centocor, Inc. ; Y. H. Rho, None; K. Kuntz, None.