Skip to main content
Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Cost-utility analysis of de-escalating biological disease-modifying anti-rheumatic drugs in patients with rheumatoid arthritis

  • Benjamin Birkner ,

    Roles Conceptualization, Data curation, Formal analysis, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing

    Affiliation Hamburg Center for Health Economics (HCHE), Universität Hamburg, Hamburg, Germany

  • Jürgen Rech,

    Roles Writing – review & editing

    Affiliation Friedrich-Alexander-University Erlangen-Nürnberg (FAU), Department of Internal Medicine 3 – Rheumatology and Immunology, Universitätsklinikum Erlangen, Erlangen, Germany

  • Tom Stargardt

    Roles Conceptualization, Methodology, Project administration, Supervision, Writing – review & editing

    Affiliation Hamburg Center for Health Economics (HCHE), Universität Hamburg, Hamburg, Germany



Recent guideline updates have suggested de-escalating DMARDs when patients with rheumatoid arthritis achieve remission or low disease activity. We aim to evaluate whether it is cost-effective to de-escalate the biological form of DMARDs (bDMARDs).


Using a Markov model, we performed a cost-utility analysis for RA patients on bDMARD treatment. We compared continuing treatment (standard care) to a tapering approach (i.e., an immediate 50% dose reduction), withdrawal (i.e., an immediate 100% dose reduction) and tapering followed by withdrawal of bDMARDs. The parametrization is based on a comprehensive literature review. Results were computed for 30 years with a cycle length of three months. We applied the payer’s perspective for Germany and conducted deterministic and probabilistic sensitivity analyses.


Tapering or withdrawing bDMARD treatment resulted in ICERs of €526,254 (incr. costs -78,845, incr. QALYs -0.1498) or €216,879 (incr. costs -€121,691, incr. QALYs -0.5611) compared to standard care. Tapering followed by withdrawal resulted in a loss of 0.4354 QALYs and savings of €107,969 per patient, with an ICER of €247,987. Deterministic sensitivity analysis revealed that our results remained largely unaffected by parameter changes. Probabilistic sensitivity analysis suggests that tapering, withdrawal and tapering followed by withdrawal were dominant in 39.8%, 28.2% and 29.0% of 10,000 iterations.


Our findings suggest that de-escalating bDMARDs in patients with RA may result in high cost savings but also a decrease in quality of life compared to standard care. If decision makers choose to implement de-escalation in daily practice, our results suggest the tapering approach.

1 Introduction

Rheumatoid Arthritis (RA) is one of the most prevalent chronic inflammatory diseases in North America and Europe [1]. In 2013, the prevalence of RA in Germany was an estimated 1.4%[2]. Treatment follows a treat-to-target approach, which aims for persistent clinical remission or at least low disease activity[3]. Pharmacological treatment consists of analgesics, non-steroidal anti-inflammatory drugs (NSAIDs), glucocorticoids, conventional synthetic disease-modifying antirheumatic drugs (csDMARDs), targeted synthetic DMARDs (tsDMARDs), biological DMARDs (bDMARDs), or combination of these.

The number of patients treated with the highly effective bDMARDs has increased continuously in the US, Europe and Australia since their introduction in the early 2000s[4]. The share of RA patients treated with bDMARDs in Germany, for example, increased from 4.4% in 2002[5] to 27.3% in 2014[6]. Annual mean drug costs per RA patient more than doubled in Germany during this period[5]. In 2016, the 41.3 million defined daily doses of bDMARDs prescribed in Germany[7] accounted for more than two billion euros in health care expenditure.

Given the high cost of long-term bDMARD use, international treatment guidelines have recommended tapering bDMARDs and csDMARDs for patients with RA in stable remission or who have low disease activity[3,810]. Although there is evidence from recent systematic literature reviews of promising clinical outcomes in patients who have de-escalated bDMARDs[1113], health-economic evaluations of de-escalation approaches are scarce. While previous research has shown that de-escalation approaches lead to substantial savings on the cost side, the impact on quality of life (QoL) is much less clear. On the one hand, Vanier et al.[14] and Tran-Duy et al.[15] identified a trade-off between costs and QoL from the French and from the Dutch perspective, respectively. On the other hand, Kobelt[16] found de-escalation to be dominant, i.e., de-escalation to increase QoL. However, all three studies used data from a single trial only (Vanier et al.: STRASS trial, Tran-Duy et al.: POET trial, Kobelt et al.: PRESERVE trial).

In the present study, we focused on bDMARDs due to their high economic impact and modelled three approaches to their de-escalation: tapering (defined as an immediate 50% dose reduction), withdrawal (defined as an immediate 100% dose reduction) and tapering for at least six months followed by withdrawal.

In contrast to previous literature[1416], our analysis (a) considers three different approaches to de-escalating bDMARDs, (b) pools all available evidence from the literature and (c) applies a long-term perspective to capture variations in disease activity. Furthermore, we also generate evidence from the payer’s perspective for Germany. To do so, we developed a Markov model with a time-horizon of 30 years and used random-effects pooling of clinical data for model parameter estimation.

2 Materials and methods

To model the de-escalation of bDMARDs from the payer’s perspective, we compared the costs and outcomes of standard care to those of three possible interventions: (a) tapering, (b) withdrawal and (c) tapering followed by withdrawal. We defined standard care as combination therapy comprising one bDMARD, one csDMARD and concomitant treatment with NSAIDs and glucocorticoids.

We used a cohort-based Markov model because this allowed us to capture recurring events in disease activity over a long time horizon. We included costs and quality-adjusted life years (QALYs) as outcomes and discounted these at 3%. We chose a cycle length of three months and a time horizon of 30 years (120 cycles), applying half-cycle correction. Our hypothetical starting cohort consists of 1,000 patients in the age of 45 years. Initial disease activity of RA is medium to high as guidelines suggest adding bDMARDs at this stage[3,8]. For calculations we used the heemod package for R[17] and Microsoft Excel.

2.1 Model structure

We defined the Markov states in our model according to disease activity: clinical remission (REM), low disease activity (LDA), medium to high disease activity (M/HDA) and death. After each cycle, patients remained in their current state, entered a state with increased or decreased disease activity, or died (Fig 1).

Fig 1. Model structure.

(A) standard care without dose adjustments, (B) tapering, (C) withdrawal, (D) tapering followed by withdrawal of bDMARDs. REM: remission, LDA: low disease activity, M/HDA: medium to high disease activity, TP_REM: remission under tapering, TP_LDA: low disease activity under tapering, WD_REM: remission under immediate withdrawal, WD_LDA: low disease activity under immediate withdrawal. Arrows indicate possible transitions between states at each model cycle of 6 months.

Patients with persistent REM or LDA after 3 months (one cycle) continued to states for sustained remission (S_REM) or low disease activity (S_LDA). Patients with sustained or improved disease activity after six months (two cycles) entered the de-escalation states. Depending on the de-escalation strategy, patients either received a 50% reduced dose of bDMARDs (states: TP_REM, TP_LDA) or were completely withdrawn from bDMARDs immediately (states: WD_REM, WD_LDA). For the approach that involved dose reduction followed by withdrawal, patients had to remain stable in the tapering state for at least six months before entering the withdrawal state (states: S_TP_REM, S_TP_LDA). Model states depicting sustained disease activity rely on the same transition probabilities as their origin states. In cases where a patient’s disease activity worsened under de-escalation, patients were transferred back to the LDA or M/HDA state and resumed bDMARDs treatment at the full dose. These patients were eligible for another round of de-escalation in the event that they subsequently achieved remission or low-disease activity for six months again. There was no restriction on the number of times patients could be transferred back to the LDA or M/HDA states and subsequently become eligible for another round of de-escalation. Lastly, in all of the states, all individuals were at risk of all-cause mortality.

2.2 Parametrization

2.2.1 Transition probabilities.

The transition probabilities in our model were estimated by applying random effects pooling [18] on 14 clinical studies[1932] identified in four systematic reviews[1113,33], as well as eight additional studies[3440] found in a literature search conducted in December 2017 using PubMed and Google Scholar to update the systematic reviews. To be included in our analysis, a study had to be a clinical trial or an observational study, measure disease activity using DAS28 based on erythrocyte sedimentation rate (ESR) or C-reactive protein (CRP), definition of remission as DAS28<2.6 or report results by severity (i.e. DAS8<2.6, DAS28<3.2) and involve pharmacological treatment comprising a combination of one csDMARD and one bDMARD administered in at least one of our settings. A detailed overview of our study pool can be found S2 Table.

We took a three step approach to estimate transition probabilities. First, we categorized studies according to the baseline DAS28 of the underlying study populations (REM: DAS28 < 2.6, LDA: DAS28 < 3.2, M/HDA: DAS28 > 3.2).

Second, we extracted the number of patients per category after 3 months of follow-up. Where these were not available we used the number of patients of the closest follow-up period. Third, we transformed the number of patients per state to transition probabilities from state to state. We calculated transition probabilities per cycle by transforming the transition probability for each study to a constant rate and then further to a transition probability based on our cycle length [41]. We used the following formula to apply both transformations in one step (), where t denotes our cycle length of three months and s the study’s duration in months.

We applied random effects pooling as a final step to estimate transition probabilities for each of the model states (Table 1, pooled estimate). By doing so we considered the studies’ outcomes with respect to individual sample size, reducing the possible impact of biased study results. An overview of the used study populations, calculated transition probabilities as well as pooled estimates of transition probabilities are shown in Table 1.

2.2.2 Mortality.

Data on mortality were based on WHO 2015 all-cause mortality for Germany. We allowed these to vary over time to reflect the ageing of the cohort. A weighted average of gender-specific all-cause mortality ensured a 3:1 ratio of female to male patients to reflect the gender distribution of RA patients in Germany[6]. We assumed a 2.34 times higher mortality risk for patients with moderate/severe RA compared to the general population based on findings for a German cohort[42].

2.2.3 Costs.

We considered two categories of direct costs: pharmaceutical and all other direct costs. Pharmaceutical costs consisted of the average costs to the payer of (a) one csDMARD, (b) one bDMARD and (c) concomitant treatment with one NSAID and one glucocorticoid.

First, we subtracted all discounts and co-payments from pharmacy retail prices for 2017. Second, we divided the calculated costs per package by the amount of milligrams per package. We considered prices for 17 bDMARDs, two csDMARDs, one glucocorticoid and two NSAIDs for the calculation. Third, to calculate the required dose in mg per cycle, we multiplied prices per mg by recommended doses from product information by the European Medicine Agency and recommendations from Smolen et al.[43]. We calculated average costs in each category per year as (a) €14,834 for bDMARDs, (b) €1,086 for csDMARDS and €45 for concomitant medication (for detailed information, see S1 Table).

Other direct costs were taken from Huscher et al.[5], who report average direct costs per disease activity state for patients with RA in Germany in 2011. Non-pharmaceutical costs in our study comprised those of hospitalization, rehabilitation, physician visits, joint replacement surgery, physiotherapy and imaging[5]. We inflated costs by 29.6% to reflect 2017 prices by incorporating growth of healthcare expenditure in Germany since 2010. Thus, all other direct costs for patients aged ≤ 64 years amounted to €1,495 (€2,024 for > 65 years) when in REM, €2,316 (€2,421) when in LDA and €3,713 (€3,881) when in M/HDA.

Annual total direct costs per state per patient aged ≤ 64 years were €17,460 (€17,989 for > 65 years) for REM, €18,281 (€18,386) for LDA and €19,678 (€19,846) for M/HDA under standard care. For tapering, we considered only 50% of the bDMARD costs, yielding total direct costs of €10,043 (€10,572) for TP_REM and €10,864 (€10,969) for TP_LDA. For withdrawal, we calculated costs of €2,626 (€3,155) for patients in WD_REM and €3,447 (€3,552) patients in WD_LDA.

2.2.4 Quality of life.

To measure QALYs, we relied on EQ-5D values by disease state from a Dutch cohort[44]. Depending on disease activity, quality-of-life values were 0.75, 0.71 and 0.60 for patients in any REM, any LDA or the M/HDA state, respectively.

2.3 Sensitivity analyses

We addressed parameter uncertainty by conducting deterministic sensitivity analyses and a Monte Carlo simulation with 10,000 iterations. For the one-way deterministic sensitivity analysis, we varied all cost parameters by 45% based on findings by Fautrel et al.[45] for French patients under combination treatment with csDMARDs and bDMARDs. Upper and lower bounds for utilities were based on the reported 95% confidence intervals of Welsing et al.[44], whereas bounds for the transition probabilities were the 95% confidence intervals from the random effects pooling. As upper and lower bounds for the discount rate, we used 6% and 0% (adjusted to 3 month cycle length).

For probabilistic sensitivity analysis, we conducted a Monte Carlo simulation with 10,000 iterations. We chose a gamma distribution for costs[45], Dirichlet distribution for transition probabilities and beta distribution for quality of life[46]. Details of all parameters can be found in Table 2.

Table 2. Model parameters.

Base case model parameters, upper and lower bounds for deterministic sensitivity analysis and distribution parameters for probabilistic sensitivity analysis.

To test structural assumptions and further assess heterogeneity in our results, we performed several scenario analyses. First, we implemented a utility decrement to consider the side-effects of bDMARDs since patients treated with these agents may have a lower quality of life than patients receiving csDMARDs only. After mapping SF-36 values from Gerhold et al. to EQ-5D[47,48], we calculated a decrement of 0.0567 for standard care (REM, LDA, M/HDA) and 0.0284 for tapering (TP_REM, TP_LDA). Second, because persistence rates in RA treatment vary widely[49], we included scenarios where patients under standard care will withdraw bDMARDs on their own after two, three or five years of stable remission. Third, we restricted eligibility for de-escalation to one time only. Finally, we varied the time horizon of our model to five, 10, 20 and 40 years.

3 Results

3.1 Base case

After 120 cycles (i.e., 30 years), our base case model yielded negative incremental costs and negative incremental QALYs for all three de-escalation approaches compared to standard care. Incremental costs per patient were -€78,845 for tapering, -€121,691 for withdrawal and -€107,696 for dose reduction followed by withdrawal, while incremental QALYs were -0.1498, -0.5611 and—0.4354, respectively. The incremental cost-effectiveness ratios (ICERs)–i.e. the costs saved per QALY lost–were €526,254 for tapering, €216,879 for withdrawal, and €247,987 for dose reduction followed by withdrawal.

3.2 Sensitivity analyses

Our results remained largely unaffected by changes in parameters related to costs and quality of life (Fig 2). Incremental QALYs were only sensitive to changes in transition probabilities under remission. An increase beyond 14.6% (base case: 7.9%) in the probability of moving from TP_REM to M/HDA shifted our results to being dominant for tapering. Reducing the efficacy of standard care by applying a higher probability of 9.1% (base case: 4.9%) for moving from REM to M/HDA also rendered tapering as the dominant strategy (Fig 2). Pharmaceutical costs for bDMARDs are the most influential cost parameter for all three models. Application of the upper (lower) bound resulted in negative incremental costs of up to €179,002 when bDMARDs are withdrawn (€64,371). We further investigate the impact of setting the upper and lower bounds of bDMARD costs to the most extreme points of our chosen gamma distribution (1%/99%-quantiles). Doing so results in negative incremental costs of up to €291,832 for the upper bound and €28,489 for the lower bound.

Fig 2. Results of deterministic sensitivity analyses for incremental costs and quality-adjusted life years (QALYs).

In our Monte Carlo simulation, tapering, withdrawal, and dose reduction followed by withdrawal were the dominant strategy in 39.8%, 28.2% and 29% of 10,000 iterations, whereas de-escalation was less costly but also less effective in 60.2%, 71.8% and 71% of 10,000 iterations, respectively (Fig 3, panel A-C). The cost-effectiveness-acceptability-curve (CEAC, Fig 3, panel D) shows the probability of a cost-effective outcome (i.e. ICER values smaller than willingness-to-pay (WTP)) for our model results. With model results being less costly and less effective the WTP resembles a requirement to save by the payer. The probability of being cost-effective decreases with increasing requirements to save per QALY lost.

Fig 3. Results of Monte Carlo simulations depicted in incremental-cost-effectiveness-ratio (ICER) planes and cost-effectiveness-acceptability curves (CEAC).

3.3 Scenario and structural sensitivity analyses

When we applied a utility decrement for the side effects of bDMARDs, tapering became the dominant strategy (Table 3). For withdrawal and dose reduction followed by withdrawal, the savings generated per QALY lost were substantially increased. When we considered self-withdrawal after two, three or five years of stable remission, savings and the loss of QALYs associated with the three interventions compared to standard care were reduced. The ICER, however, remained similar to the base case. Restricting patients’ eligibility for de-escalation to one time only resulted in a decrease in incremental costs saved, as well as a decrease in incremental QALYs lost.

Table 3. Results from scenario and structural sensitivity analyses.

4 Discussion

We conducted a cost-utility analysis from the payer’s perspective for three different approaches to de-escalating bDMARD treatment in patients with RA: tapering, withdrawal, and tapering followed by withdrawal. The results of our model suggest that all three approaches lead to high cost savings but also a decrease in quality of life. The ICERs appeared to be very high, with €526,254 savings per QALY lost for tapering, €216,879 for withdrawal and €247,987 for tapering followed by withdrawal. Our results remained stable when we varied our input parameters, relaxed assumptions about adherence, or restricted the eligibility of patients for de-escalation to one time only.

To our knowledge, this is the first study to simultaneously (a) consider three different approaches to de-escalating bDMARDs, (b) pool all relevant available evidence on this subject and (c) apply a long-term perspective of 30 year from the payer’s perspective. We dealt with the likely heterogeneity of results in published efficacy studies by pooling the available evidence. Our modelling approach can be transferred easily to different health systems by adapting the cost and utility parameters according to the individual system characteristics.

Overall, our results are in line with those of three recent studies[1416], which found that de-escalating bDMARDs was associated with high cost-savings but also a decrease or only a very slight increase in quality of life compared to continued treatment. Tran-Duy et al.[15], found that savings of €7,133 were associated with a 0.022 loss in QALYs for a time horizon of one year when comparing withdrawal of bDMARDs with continued treatment in a Dutch population. While we found lower savings per year (€4,056) based on our 30-year time horizon, we found a similar loss in quality of life (0.0187). In a French cohort study with a time horizon of 18 months, tapering TNFi-type bDMARDs by means of a spacing strategy was also associated with saved costs (€8,440) and a loss in QALYs (0.158)[14]. The most similar approach to this in our model–i.e., tapering in the form of a 50% dose reduction followed by withdrawal compared to standard care–was associated with lower cost-savings (€3,225) per year, as well as fewer QALYs lost (0.0189). In a Swedish population, the dose-reduction strategy dominated continued full-dose treatment with etanercept with savings of €6,321 and QALYs gained of 0.010 over a 10 year time horizon[16]. We also identified savings from our tapering strategy after 10 years, though these were substantially higher (€33,395) and accompanied by a small loss of QALYs of 0.0560.

The withdrawal approach in our model showed an ICER of €216,879 after 30 years. This is in contrast to a much higher ICER of €368,269 calculated by Tran-Duy et al. based on the results of the POET trial only, albeit using one-year time horizon and the societal perspective[15]. However, depending on the time horizon, Kobelt et al. found a higher ICER of €262,770 (2 years) and a lower ICER of €81,818 (10 years) when comparing continuous treatment with withdrawal from the societal perspective[16]. When we applied a utility decrement, tapering became the dominant strategy which is in line with results found for a Swedish population[16]. The difference between our ICERs and those in other studies[1416] is likely driven by the different evidence used to model effectiveness: whereas the other three publications relied on the results of one or only a few RCTs, we considered the pooled results of 22 clinical studies.

To calculate cost-effectiveness, one would usually compare the ICER against WTP threshold for a QALY. Although there is no official threshold of this type in Germany, we can apply the thresholds suggested by the World Health Organization (WHO) [50] or WTP per QALY for Germany calculated by Ahlert et al.[51]. If judged against these, our results can be regarded as highly cost-effective. Judging against the WTP per QALY in Germany of €18,420 [51] we find a probability of being cost-effective of 98% for all three interventions. Applying the threshold of one to three times the GDP suggested by the WHO[50] shows 92%/74% for tapering, 90%/60% for withdrawal and 91%/62% for tapering followed by withdrawal (Fig 3, panel D).

However, the WTP for a positive outcome might not match the willingness to accept (WTA) a negative consequence[52]. For the healthcare sector, O’Brien et al.[52] found a WTA/WTP ratio of 1.9 for the introduction of a new drug, whereas Carthy et al.[53] reported a ratio of 6.9 when assessing the WTA/WTP ratio for a two-week hospital stay. If we consider a hypothetical scenario in which the most extreme WTP value calculated by Ahlert et al.[51] (€18,420) is multiplied by the WTA/WTP ratio of 6.9, our model results still suggest that there is a high probability that de-escalation is cost-effective (tapering: 73%, withdrawal: 58%, dose reduction followed by withdrawal: 60%).

Among the three de-escalation strategies we modelled, tapering achieved the highest savings per QALY lost. This would also meet recommendations by international guidelines to reduce drug dose when patients achieve sustained low disease activity[3,8]. In addition, an immediate withdrawal of bDMARDs would be unlikely to gain acceptance due to risk aversion among patients and rheumatologists. Patients with RA have been found to be reluctant to change their treatment (e.g., to try a new drug, modify the dose of an existing treatment, or take a so-called drug holiday) and have expressed fears of losing disease control or experiencing flares and side effects[54].

Our model has a number of important limitations. First, we did not distinguish between DAS28 calculated using ESR or CRP. Considering the Simplified or Clinical Disease Activity Index (SDAI/CDAI) might increase the clinical accuracy of our estimated transition probabilities. We still chose to include studies that used the DAS28 based on either biomarker because (a) there is a very high correlation of the DAS28 and the SDAI/CDAI scores[55] and (b) restricting these to SDAI/CDAI only would have substantially reduced the amount of evidence at hand for estimating the parameters of our model.

Second, we had to combine medium and high disease activity into one state (M/HDA) because the underlying studies rarely reported the number of patients within these states. For the same reason, we chose a cut-off of DAS28 < 3.2 to identify LDA in clinical studies, following the classification of Schett et al.[11].

Third, our models may underestimate direct costs (except pharmaceutical costs) because we rely on cost data from registries that, in turn, might be subject to underreporting. This is also true for costs arising from side-effects of concomitant treatments. However, because pharmaceutical costs in our model are four to ten times higher than other direct costs, any underestimation of the latter is very unlikely to jeopardize our results. In the same context, we may overestimate pharmaceutical costs for concomitant medication because we assumed that pharmaceutical costs are independent of disease activity. For example, glucocorticoids or NSAIDs should also be tapered quickly if symptoms have decreased in severity. Again, with costs of less than €0.01 per mg, the impact on our results is negligible.

Fourth, in the base case, we assumed that patients were compliant over the full model horizon of 30 years. While there is evidence that patients and rheumatologists are reluctant to alter their regimen due to the risk of losing disease control, persistence rates in Germany suggest that compliance in patients with RA is imperfect[49]. Regardless, model results were unaffected when we examined possible self-withdrawal of bDMARDs.

Lastly, we modelled dose reduction followed by withdrawal by combining results from studies on tapering with results from studies on immediate withdrawal because only one RCT with 27 patients examined a de-escalation approach consisting of dose reduction followed by withdrawal[21]. In the same context, we include results from RCTs that adjust both csDMARD and/or bDMARD dose in our study pool[21,31]. Additionally, the term “tapering” often refers to a gradual decrease in dose or gradual increase in the spacing of injections, whereas we define it as an immediate 50% reduction in dose. However, modelling individual dosing patterns is beyond the scope of this analysis, and the included RCTs for tapering use the immediate 50% dose reduction only.

5 Conclusion

Our findings suggest that de-escalating bDMARDs through tapering (i.e., through an immediate 50% dose reduction), withdrawal, or tapering followed by withdrawal in patients with RA may lead to high cost savings compared with standard care. However, these savings are accompanied by a decrease in quality of life. If decision makers choose to implement de-escalation in daily practice, our results suggest following the tapering approach. In addition, tapering followed by withdrawal might be viable options for patients who want to change their treatment but have not yet done so for fear of losing disease control.

Although our study sheds light on whether de-escalating bDMARDs in patients with RA is a viable option from the payer’s perspective, more research using data from clinical practice is necessary to validate our findings.


  1. 1. Alamanos Y, Voulgari PV, Drosos AA. Incidence and Prevalence of Rheumatoid Arthritis, Based on the 1987 American College of Rheumatology Criteria: A Systematic Review. Semin Arthritis Rheum. 2006;36: 182–188. pmid:17045630
  2. 2. Hense S, Ramos AL, Callhoff J, Albrecht K, Zink A, Hoffmann F. Prävalenz der rheumatoiden Arthritis in Deutschland auf Basis von Kassendaten. Z Für Rheumatol. 2016;75: 819–827.
  3. 3. Smolen JS, Landewé R, Bijlsma J, Burmester G, Chatzidionysiou K, Dougados M, et al. EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs: 2016 update. Ann Rheum Dis. 2017; 1–18. pmid:28264816
  4. 4. Jönsson B, Kobelt G, Smolen J. The burden of rheumatoid arthritis and access to treatment: uptake of new therapies. Eur J Health Econ. 2008;8: 61–86. pmid:18097697
  5. 5. Huscher D, Mittendorf T, von Hinüber U, Kötter I, Hoese G, Pfäfflin A, et al. Evolution of cost structures in rheumatoid arthritis over the past decade. Ann Rheum Dis. 2015;74: 738–745. pmid:24406543
  6. 6. Albrecht K, Huscher D, Eidner T, Kleinert S, Späthling-Mestekemper S, Bischoff S, et al. Versorgung der rheumatoiden Arthritis 2014. Z Für Rheumatol. 2017;76: 50–57.
  7. 7. Böger RH, Schmidt G. Antirheumatika und Antiphlogistika. In: Schwabe U, Paffrath D, editors. Arzneiverordnungs-Report 2016. Springer Berlin Heidelberg; 2016. pp. 369–380.
  8. 8. Fiehn C, Holle J, Iking-Konert C, Leipe J, Weseloh C, Frerix M, et al. S2e-Leitlinie: Therapie der rheumatoiden Arthritis mit krankheitsmodifizierenden Medikamenten. Z Für Rheumatol. 2018; 1–19.
  9. 9. Singh JA, Saag KG, Bridges SL, Akl EA, Bannuru RR, Sullivan MC, et al. 2015 American College of Rheumatology Guideline for the Treatment of Rheumatoid Arthritis. Arthritis Rheumatol. 2016;68: 1–26. pmid:26545940
  10. 10. Lau CS, Chia F, Harrison A, Hsieh T-Y, Jain R, Jung SM, et al. APLAR rheumatoid arthritis treatment recommendations. Int J Rheum Dis. 2015;18: 685–713. pmid:26334449
  11. 11. Schett G, Emery P, Tanaka Y, Burmester G, Pisetsky DS, Naredo E, et al. Tapering biologic and conventional DMARD therapy in rheumatoid arthritis: current evidence and future directions. Ann Rheum Dis. 2016;75: 1428–1437. pmid:27261493
  12. 12. Nam JL, Ramiro S, Gaujoux-Viala C, Takase K, Leon-Garcia M, Emery P, et al. Efficacy of biological disease-modifying antirheumatic drugs: a systematic literature review informing the 2013 update of the EULAR recommendations for the management of rheumatoid arthritis. Ann Rheum Dis. 2014;73: 516–528. pmid:24399231
  13. 13. Kuijper T, Lamers-Karnebeek F, Jacobs J, Hazes J, Luime J. Flare Rate in Patients with Rheumatoid Arthritis in Low Disease Activity or Remission When Tapering or Stopping Synthetic or Biologic DMARD: A Systematic Review. J Rheumatol. 2015; 2012–22. pmid:26428204
  14. 14. Vanier A, Mariette X, Tubach F, Fautrel B. Cost-Effectiveness of TNF-Blocker Injection Spacing for Patients with Established Rheumatoid Arthritis in Remission: An Economic Evaluation from the Spacing of TNF-Blocker Injections in Rheumatoid Arthritis Trial. Value Health. 2017;20: 577–585. pmid:28407999
  15. 15. Tran‐Duy A, Moghadam MG, Voshaar MAHO, Vonkeman HE, Boonen A, Clarke P, et al. An Economic Evaluation of Stopping Versus Continuing Tumor Necrosis Factor Inhibitor Treatment in Rheumatoid Arthritis Patients With Disease Remission or Low Disease Activity. Arthritis Rheumatol. 2018;70: 1557–1564. pmid:29745059
  16. 16. Kobelt G. Treating to Target with Etanercept in Rheumatoid Arthritis: Cost-Effectiveness of Dose Reductions When Remission Is Achieved. Value Health. 2014;17: 537–544. pmid:25128046
  17. 17. Filipović-Pierucci A, Zarca K, Durand-Zaleski I. Markov Models for Health Economic Evaluations: The R Package heemod. ArXiv170203252 Stat. 2017 [cited 12 Oct 2017].
  18. 18. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7: 177–188. pmid:3802833
  19. 19. Emery P, Hammoudeh M, FitzGerald O, Combe B, Martin-Mola E, Buch MH, et al. Sustained Remission with Etanercept Tapering in Early Rheumatoid Arthritis. N Engl J Med. 2014;371: 1781–1792. pmid:25372086
  20. 20. Tanaka Y, Hirata S, Kubo S, Fukuyo S, Hanami K, Sawamukai N, et al. Discontinuation of adalimumab after achieving remission in patients with established rheumatoid arthritis: 1-year outcome of the HONOR study. Ann Rheum Dis. 2015;74: 389–395. pmid:24288014
  21. 21. Haschka J, Englbrecht M, Hueber AJ, Manger B, Kleyer A, Reiser M, et al. Relapse rates in patients with rheumatoid arthritis in stable remission tapering or stopping antirheumatic therapy: interim results from the prospective randomised controlled RETRO study. Ann Rheum Dis. 2016;75: 45–51. pmid:25660991
  22. 22. Smolen JS, Nash P, Durez P, Hall S, Ilivanova E, Irazoque-Palazuelos F, et al. Maintenance, reduction, or withdrawal of etanercept after treatment with etanercept and methotrexate in patients with moderate rheumatoid arthritis (PRESERVE): a randomised controlled trial. The Lancet. 2013;381: 918–929.
  23. 23. Smolen JS, Emery P, Fleischmann R, van Vollenhoven RF, Pavelka K, Durez P, et al. Adjustment of therapy in rheumatoid arthritis on the basis of achievement of stable low disease activity with adalimumab plus methotrexate or methotrexate alone: the randomised controlled OPTIMA trial. The Lancet. 2014;383: 321–332.
  24. 24. van Herwaarden N, van der Maas A, Minten MJM, van den Hoogen FHJ, Kievit W, van Vollenhoven RF, et al. Disease activity guided dose reduction and withdrawal of adalimumab or etanercept compared with usual care in rheumatoid arthritis: open label, randomised controlled, non-inferiority trial. BMJ. 2015;350: h1389. pmid:25858265
  25. 25. Keystone EC, Pope JE, Thorne JC, Poulin-Costello M, Phan-Chronis K, Vieira A, et al. Two-year radiographic and clinical outcomes from the Canadian Methotrexate and Etanercept Outcome study in patients with rheumatoid arthritis. Rheumatol Oxf Engl. 2016;55: 327–334. pmid:26361879
  26. 26. Westhovens R, Yocum D, Han J, Berman A, Strusberg I, Geusens P, et al. The safety of infliximab, combined with background treatments, among patients with rheumatoid arthritis and various comorbidities: A large, randomized, placebo-controlled trial. Arthritis Rheum. 2006;54: 1075–1086. pmid:16572442
  27. 27. Emery P, Breedveld FC, Hall S, Durez P, Chang DJ, Robertson D, et al. Comparison of methotrexate monotherapy with a combination of methotrexate and etanercept in active, early, moderate to severe rheumatoid arthritis (COMET): a randomised, double-blind, parallel treatment trial. The Lancet. 2008;372: 375–382.
  28. 28. Schiff M, Keiserman M, Codding C, Songcharoen S, Berman A, Nayiager S, et al. Efficacy and safety of abatacept or infliximab vs placebo in ATTEST: a phase III, multi-centre, randomised, double-blind, placebo-controlled study in patients with rheumatoid arthritis and an inadequate response to methotrexate. Ann Rheum Dis. 2008;67: 1096–1103. pmid:18055472
  29. 29. Kavanaugh A, Fleischmann RM, Emery P, Kupper H, Redden L, Guerette B, et al. Clinical, functional and radiographic consequences of achieving stable low disease activity and remission with adalimumab plus methotrexate or methotrexate alone in early rheumatoid arthritis: 26-week results from the randomised, controlled OPTIMA study. Ann Rheum Dis. 2013;72: 64–71. pmid:22562973
  30. 30. Dougados M, Kissel K, Conaghan PG, Mola EM, Schett G, Gerli R, et al. Clinical, radiographic and immunogenic effects after 1 year of tocilizumab-based treatment strategies in rheumatoid arthritis: the ACT-RAY study. Ann Rheum Dis. 2014;73: 803–809. pmid:24473673
  31. 31. Nam JL, Villeneuve E, Hensor EMA, Wakefield RJ, Conaghan PG, Green MJ, et al. A randomised controlled trial of etanercept and methotrexate to induce remission in early inflammatory arthritis: the EMPIRE trial. Ann Rheum Dis. 2014;73: 1027–1036. pmid:24618266
  32. 32. Tanaka Y, Takeuchi T, Mimori T, Saito K, Nawata M, Kameda H, et al. Discontinuation of infliximab after attaining low disease activity in patients with rheumatoid arthritis: RRR (remission induction by Remicade in RA) study. Ann Rheum Dis. 2010;69: 1286–1291. pmid:20360136
  33. 33. Wiens A, Venson R, Correr CJ, Otuki MF, Pontarolo R. Meta-analysis of the Efficacy and Safety of Adalimumab, Etanercept, and Infliximab for the Treatment of Rheumatoid Arthritis. Pharmacother J Hum Pharmacol Drug Ther. 2010;30: 339–353. pmid:20334454
  34. 34. Kremer JM, Dougados M, Emery P, Durez P, Sibilia J, Shergy W, et al. Treatment of rheumatoid arthritis with the selective costimulation modulator abatacept: Twelve‐month results of a phase iib, double‐blind, randomized, placebo‐controlled trial. Arthritis Amp Rheumatol. 2005;52: 2263–2271. pmid:16052582
  35. 35. Rubbert-Roth A, Tak PP, Zerbini C, Tremblay J-L, Carreño L, Armstrong G, et al. Efficacy and safety of various repeat treatment dosing regimens of rituximab in patients with active rheumatoid arthritis: results of a Phase III randomized study (MIRROR). Rheumatology. 2010;49: 1683–1693. pmid:20463186
  36. 36. Kim H-Y, Hsu P-N, Barba M, Sulaiman W, Robertson D, Vlahos B, et al. Randomized comparison of etanercept with usual therapy in an Asian population with active rheumatoid arthritis: the APPEAL trial. Int J Rheum Dis. 2012;15: 188–196. pmid:22462423
  37. 37. Yoo DH, Hrycaj P, Miranda P, Ramiterre E, Piotrowski M, Shevchuk S, et al. A randomised, double-blind, parallel-group study to demonstrate equivalence in efficacy and safety of CT-P13 compared with innovator infliximab when coadministered with methotrexate in patients with active rheumatoid arthritis: the PLANETRA study. Ann Rheum Dis. 2013;72: 1613–1620. pmid:23687260
  38. 38. Hørslev-Petersen K, Hetland ML, Junker P, Pødenphant J, Ellingsen T, Ahlquist P, et al. Adalimumab added to a treat-to-target strategy with methotrexate and intra-articular triamcinolone in early rheumatoid arthritis increased remission rates, function and quality of life. The OPERA Study: an investigator-initiated, randomised, double-blind, parallel-group, placebo-controlled trial. Ann Rheum Dis. 2014;73: 654–661. pmid:23434570
  39. 39. Machado DA, Guzman RM, Xavier RM, Simon JA, Mele L, Pedersen R, et al. Open-Label Observation of Addition of Etanercept Versus a Conventional Disease-Modifying Antirheumatic Drug in Subjects With Active Rheumatoid Arthritis Despite Methotrexate Therapy in the Latin American Region. JCR J Clin Rheumatol. 2014;20: 25–33. pmid:24356474
  40. 40. Schiff M, Weinblatt ME, Valente R, van der Heijde D, Citera G, Elegbe A, et al. Head-to-head comparison of subcutaneous abatacept versus adalimumab for rheumatoid arthritis: two-year efficacy and safety findings from AMPLE trial. Ann Rheum Dis. 2014;73: 86–94. pmid:23962455
  41. 41. Briggs A, Claxton K, Sculpher M. Decision modelling for health economic evaluation. Reprint. Oxford [u.a.]: Oxford Univ. Press; 2011.
  42. 42. Listing J, Kekow J, Manger B, Burmester G-R, Pattloch D, Zink A, et al. Mortality in rheumatoid arthritis: the impact of disease activity, treatment with glucocorticoids, TNFα inhibitors and rituximab. Ann Rheum Dis. 2013; 514–21. pmid:24291654
  43. 43. Smolen JS, Aletaha D, McInnes IB. Rheumatoid arthritis. The Lancet. 2016;388: 2023–2038.
  44. 44. Welsing PMJ, Severens JL, Hartman M, van Gestel AM, van Riel PLCM, Laan RFJM. The Initial Validation of a Markov Model for the Economic Evaluation of (New) Treatments for Rheumatoid Arthritis. PharmacoEconomics. 2006;24: 1011–1020. pmid:17002483
  45. 45. Fautrel B, Cukierman G, Joubert J-M, Laurendeau C, Gourmelen J, Fagnani F. Healthcare service utilisation costs attributable to rheumatoid arthritis in France: Analysis of a representative national claims database. Joint Bone Spine. 2016;83: 53–56. pmid:26671705
  46. 46. Lundkvist J, Kastäng F, Kobelt G. The burden of rheumatoid arthritis and access to treatment: health burden and costs. Eur J Health Econ. 2008;8: 49–60. pmid:18157732
  47. 47. Dakin H, Burns R, Yang Y. HERC database of mapping studies, Version 5.0 (Last updated: 16th May 2016). 2016.
  48. 48. Ara R, Brazier J. Deriving an Algorithm to Convert the Eight Mean SF-36 Dimension Scores into a Mean EQ-5D Preference-Based Score from Published Studies (Where Patient Level Data Are Not Available). Value Health. 2008;11: 1131–1143. pmid:18489495
  49. 49. Neubauer S, Cifaldi M, Mittendorf T, Ganguli A, Wolff M, Zeidler J. Biologic TNF inhibiting agents for treatment of rheumatoid arthritis: persistence and dosing patterns in Germany. Health Econ Rev. 2014;4: 32. pmid:26208932
  50. 50. Hutubessy R, Chisholm D, Edejer TT-T, WHO-CHOICE. Generalized cost-effectiveness analysis for national-level priority-setting in the health sector. Cost Eff Resour Alloc. 2003;1: 8. pmid:14687420
  51. 51. Ahlert M, Breyer F, Schwettmann L. How you ask is what you get: Framing effects in willingness-to-pay for a QALY. Soc Sci Med. 2016;150: 40–48. pmid:26730880
  52. 52. O’Brien BJ, Goeree R, Gafni A, Torrance GW, Pauly MV, Erder H, et al. Assessing the Value of a New Pharmaceutical: A Feasibility Study of Contingent Valuation in Managed Care. Med Care. 1998;36: 370–384. pmid:9520961
  53. 53. Carthy T, Chilton S, Covey J, Hopkins L, Jones-Lee M, Loomes G, et al. On the Contingent Valuation of Safety and the Safety of Contingent Valuation: Part 2—The CV/SG &quot;Chained&quot; Approach. J Risk Uncertain. 1998;17: 187–214.
  54. 54. Markusse IM, Akdemir G, Huizinga TWJ, Allaart CF. Drug-free holiday in patients with rheumatoid arthritis: a qualitative study to explore patients’ opinion. Clin Rheumatol. 2014;33: 1155–1159. pmid:24469272
  55. 55. Aletaha D, Smolen J. The Simplified Disease Activity Index (SDAI) and the Clinical Disease Activity Index (CDAI): a review of their usefulness and validity in rheumatoid arthritis. Clin Exp Rheumatol. 2005;23: S100–8. pmid:16273793