The impact of non-persistence on the direct and indirect costs in patients treated with subcutaneous tumour necrosis factor-alpha inhibitors in Germany

The impact of non-persistence on the direct and indirect costs in patients treated with... Abstract Objective The goal of the present study was to estimate the treatment costs in immune-mediated rheumatic disease patients initiating treatment with an s.c. biologic agent based on treatment persistence. Methods This is a retrospective cohort study based on the German statutory health insurance funds database. Patients ⩾18 years of age with a diagnosis of AS, PsA or RA treated with s.c. TNF-α inhibitors (TNFis) were included. Persistence was estimated as the duration of time from s.c. TNFi therapy initiation to discontinuation, which was defined as at least 60 days without therapy. We performed 1:1 matching based on a propensity score that was constructed as the conditional probability of being persistent as a function of age, gender, index year, physician specialty and Charlson comorbidity index. Finally, the cost differences between the matched pairs were estimated using the Wilcoxon test. Results After 1:1 matching, 678 persistent and 678 non-persistent patients were available for cost analyses. Using a 2-year time period, the costs for office-based visits per patient were €2319 in the persistent cohort compared with €3094 in the non-persistent cohort (P < 0.001). Co-medication costs were €2828 in the persistent cohort compared with €5498 in the non-persistent cohort, hospitalization costs were €3551 in the persistent cohort compared with €5890 in the non-persistent cohort and sick leave costs were €717 in the persistent cohort compared with €1241 in the non-persistent cohort (all P < 0.001). Conclusion The results of this study indicate that persistence with s.c. TNFi treatment can be associated with several cost offsets for immune-mediated rheumatic disease patients. IMRD, persistence, TNF-α, blocker, treatment costs Rheumatology key messages Biologic treatment persistence is associated with several cost offsets for immune-mediated rheumatic disease patients. In immune-mediated rheumatic disease patients, direct medical costs were lower in the persistent cohort. Introduction AS, PsA and RA are chronic progressive immune-mediated rheumatic diseases (IMRDs) that can cause pain, deformity and disability and progressively impair joint structure and function [1]. IMRDs result in substantial personal as well as economic burden [2]. The primary treatment goal for all three conditions is to maximize health-related quality of life by controlling symptoms and inflammation, preventing progressive structural damage, preserving or normalizing function and social participation and targeting remission. Treatment of rheumatic diseases usually involves a multimodal approach including both pharmacological and non-pharmacological strategies [3, 4]. The advent of biologic medications such as s.c. TNF-α inhibitors (TNFis) has transformed the management of IMRDs [5]. Previous research has provided high-quality evidence that anti-TNF agents improve clinical symptoms in the treatment of IMRDs [6–8]. Taking the prescribed medication for a sufficient period of time is crucial to the success of any therapy. Non-compliance and a lack of persistence with prescription medications are major problems in the clinical management of chronic diseases and s.c. TNFi treatment is potentially a very convenient option for patients. However, patients may likely have drug interruptions or long periods of non-persistence. In IMRDs, treatment persistence can be used to represent drug effectiveness, safety and treatment satisfaction [9–11]. While associations between adherence and costs [12] and the cost impact of switching from an s.c. TNFi to different biologic classes have been examined [5, 13], studies estimating the cost impact of non-persistence in patients treated with s.c. TNFis for IMRDs are lacking. In 2016, Dalén et al. [14] analysed treatment costs in first-line (bio-naïve) IMRD patients in Sweden based on persistent and non-persistent patients. The results of their study indicated that persistence with s.c. TNFi treatment may be associated with some cost offsets for IMRD patients. In 2017, Dalén et al. [15] further showed that, in second-line therapy with s.c. TNFis, persistent patients had lower mean total costs. In the same year, Svedbom et al. [16] described and compared treatment persistence with first- and second-line s.c. TNFis as well as their corresponding costs in Sweden. Although these works are of great interest, no recent data are available about the association between non-persistence with s.c. TNFis and costs in Germany. The goal of the present study was to estimate and compare the direct and indirect treatment costs in IMRD patients initiating treatment with an s.c. biologic agent based on treatment persistence. Methods Data source This study is a retrospective cohort study based on the German statutory health insurance funds database. This database continuously provides data pertaining to about five million statutorily insured people, which corresponds to 7% coverage. Exclusively anonymous information and no personal data (in accordance with § 3 Abs. 6 German Federal Data Protection Act) are used. Each patient with defined characteristics (e.g. specific disease, number and/or durations of hospitalizations, prescriptions of specific medications or any combination of these characteristics) can be tracked within the database via a unique patient ID. Finally, the database has already been used in several epidemiological studies, including cost analyses [17]. General informed consent Informed patient consent German law allows the use of anonymous electronic medical records for research purposes under certain conditions. Because patients were only queried as aggregates and no protected health information was available for queries, no institutional review board approval was required for the use of this database or completion of this study. For this study, we only used anonymized data collected from the claims database. Therefore, this study does not require that we obtain informed consent from patients. Study population Patients ⩾18 years of age with a diagnosis of RA [International Classification of Diseases, 10th Revision (ICD-10) M05, M06], PsA (ICD-10 L40.5 in combination with M07.0, M07.1, M07.2, M07.3 or M09.0) or AS (ICD-10 M45) treated with any of the four s.c. TNF-α blockers of interest (etanercept, adalimumab, certolizumab pegol or golimumab) were selected. Patients were categorized by their main diagnosis at the index date (the diagnosis that is documented within the quarter of the index prescription date). Cases with more than one documented diagnosis were classified as a mixed group. The date of the first qualifying prescription for an s.c. biologic agent served as the patient’s index date. The 12-month period immediately preceding the index date was used as the pre-index period and the 12-month period following the final prescription within the persistence period served as the post-index period, which lasted no later than December 2015. All patients were traceable in the database for at least 12 months after their last prescription within the analysis period. The pre-index period and index date were used to measure patients’ baseline characteristics. Treatment persistence with index s.c. TNF-α blockers and health care resource utilization (such as services and direct and indirect costs) were evaluated in both the persistence and post-index periods. Persistence definition and patient stratification Persistence was defined as the duration of time from s.c. TNFi therapy initiation to discontinuation, which was identified as at least 60 days without s.c. TNFi therapy. A longitudinal dataset of medication supply was created for each individual patient and the number of days of drug supply was calculated based on the defined daily dose information associated with each prescription record. The persistence was evaluated at 24 months after the index date. Following this estimation of persistence, patients were categorized into persistent and non-persistent cohorts. The main outcome of the study was rheumatoid disease–related health care resource utilization costs, including direct costs and services. These costs were estimated for the time within 2 years (730 days) following the first s.c. TNFi prescription and separated by RA, PsA and AS indication. Direct costs included hospitalization costs, doctor and hospital visit costs, prescribed medication costs and sickness benefits (in Germany, sickness funds are obligated to provide sick pay for sick leaves of at least 43 days). Services consisted of the number of hospital stays, duration of hospitalization, medical department of the hospital involved and specialties visited. Covariates Demographic characteristics of patients at the index date included age (as a continuous variable and age groups) and gender. Further baseline variables were year of index date and the specialty of the physician initiating treatment. Furthermore, a revised version of the Charlson comorbidity index (CCI) was used as a generic marker of comorbidity. Statistical methods To control for confounding, 1:1 matching was carried out based on a propensity score that was constructed as the conditional probability of being persistent as a function of age, gender, index year, physician specialty and CCI (logistic regression). Greedy matching was used by choosing a non-persistent patient whose propensity score was closest to that of this randomly selected persistent subject for matching. Descriptive statistics were given and group differences (persistent vs non-persistent) were assessed using the Wilcoxon test or the McNemar’s tests after propensity score matching. Finally, the differences in direct costs, indirect costs and services between matched pairs were estimated using the Wilcoxon test. Results Patient characteristics A total of 2060 patients initiating s.c. TNFi therapy were included in the study. After 2 years of follow-up, 682 patients still received s.c. TNFi therapy and 1378 patients had discontinued therapy. In the cohort of non-persistent patients, the mean persistence duration was 266 days (s.d. 190). We observed no significant difference between persistent and non-persistent patients in terms of age, gender or year of therapy initiation. Non-persistent patients had a slightly worse average health status compared with the persistent cohort (CCI 2.6 vs 2.4) (Table 1). After 1:1 matching, 678 pairs (678 persistent and 678 non-persistent patients) were available for cost analyses. Both cohorts were similar in terms of age, gender, year of therapy initiation, CCI and indication (Table 1). Table 1 Baseline characteristics of study patients before and after matching Variable  Before matching   After matching   Persistent  Non-persistent  P-value  Persistent  Non-persistent  P-value  Total (N)  682   1378    678   678    Indication                  RA (%)  55.7  58.8  0.040  55.9  57.2  0.520      AS (%)  21.7  19.9  0.579  21.8  21.7  1.000      PsA (%)  10.7  8.1  0.277  10.3  8.7  0.454      Mixed (%)  11.9  13.3  0.369  11.9  12.4  0.803  Age, mean (s.d.), years  51.1 (13.2)  51.4 (14.1)  0.669  51.1 (13.2)  51.2 (14.0)  0.905  Gender, male, %  42.4  41.1  0.572  42.5  44.0  0.583  Index year, %                  2010  29.6  25.7  0.305  29.4  29.4  0.854      2011  23.6  24.5  23.6  25.5      2012  24.5  25.9  24.6  23.6      2013  22.3  23.9  22.4  21.5  CCI, mean (s.d.)  2.4 (1.3)  2.6 (1.6)  <0.001  2.0 (1.3)  2.0 (1.3)  0.9227  Variable  Before matching   After matching   Persistent  Non-persistent  P-value  Persistent  Non-persistent  P-value  Total (N)  682   1378    678   678    Indication                  RA (%)  55.7  58.8  0.040  55.9  57.2  0.520      AS (%)  21.7  19.9  0.579  21.8  21.7  1.000      PsA (%)  10.7  8.1  0.277  10.3  8.7  0.454      Mixed (%)  11.9  13.3  0.369  11.9  12.4  0.803  Age, mean (s.d.), years  51.1 (13.2)  51.4 (14.1)  0.669  51.1 (13.2)  51.2 (14.0)  0.905  Gender, male, %  42.4  41.1  0.572  42.5  44.0  0.583  Index year, %                  2010  29.6  25.7  0.305  29.4  29.4  0.854      2011  23.6  24.5  23.6  25.5      2012  24.5  25.9  24.6  23.6      2013  22.3  23.9  22.4  21.5  CCI, mean (s.d.)  2.4 (1.3)  2.6 (1.6)  <0.001  2.0 (1.3)  2.0 (1.3)  0.9227  CCI: Charlson comorbidity index. Table 1 Baseline characteristics of study patients before and after matching Variable  Before matching   After matching   Persistent  Non-persistent  P-value  Persistent  Non-persistent  P-value  Total (N)  682   1378    678   678    Indication                  RA (%)  55.7  58.8  0.040  55.9  57.2  0.520      AS (%)  21.7  19.9  0.579  21.8  21.7  1.000      PsA (%)  10.7  8.1  0.277  10.3  8.7  0.454      Mixed (%)  11.9  13.3  0.369  11.9  12.4  0.803  Age, mean (s.d.), years  51.1 (13.2)  51.4 (14.1)  0.669  51.1 (13.2)  51.2 (14.0)  0.905  Gender, male, %  42.4  41.1  0.572  42.5  44.0  0.583  Index year, %                  2010  29.6  25.7  0.305  29.4  29.4  0.854      2011  23.6  24.5  23.6  25.5      2012  24.5  25.9  24.6  23.6      2013  22.3  23.9  22.4  21.5  CCI, mean (s.d.)  2.4 (1.3)  2.6 (1.6)  <0.001  2.0 (1.3)  2.0 (1.3)  0.9227  Variable  Before matching   After matching   Persistent  Non-persistent  P-value  Persistent  Non-persistent  P-value  Total (N)  682   1378    678   678    Indication                  RA (%)  55.7  58.8  0.040  55.9  57.2  0.520      AS (%)  21.7  19.9  0.579  21.8  21.7  1.000      PsA (%)  10.7  8.1  0.277  10.3  8.7  0.454      Mixed (%)  11.9  13.3  0.369  11.9  12.4  0.803  Age, mean (s.d.), years  51.1 (13.2)  51.4 (14.1)  0.669  51.1 (13.2)  51.2 (14.0)  0.905  Gender, male, %  42.4  41.1  0.572  42.5  44.0  0.583  Index year, %                  2010  29.6  25.7  0.305  29.4  29.4  0.854      2011  23.6  24.5  23.6  25.5      2012  24.5  25.9  24.6  23.6      2013  22.3  23.9  22.4  21.5  CCI, mean (s.d.)  2.4 (1.3)  2.6 (1.6)  <0.001  2.0 (1.3)  2.0 (1.3)  0.9227  CCI: Charlson comorbidity index. Cost analyses Table 2 shows the mean per-patient costs in the persistent and non-persistent cohorts. The costs for office-based visits per patient were €2319 in the persistent cohort and €3094 in the non-persistent cohort (P < 0.001). The cost difference was the highest in the mixed indication group (Table 2). An average persistent patient received prescriptions of different co-medications worth €2828 and a non-persistent patient received prescriptions worth €5498 within the 2-year time period (P < 0.001). In this category, the highest cost difference was found for patients with RA. Hospitalization costs were €3551 per persistent patient and €5890 per non-persistent patients (Table 2), which was probably due to the longer average duration of hospitalization in the non-persistent cohort. Furthermore, sick leave costs were much lower in persistent (€717) vs non-persistent (€1241) patients. Total costs per patient (excluding the cost of s.c. TNFi therapy) were ∼€6634 lower in the persistent cohort (Table 2). Table 2 Average costs per patient within 2 years after initiation of s.c. TNFi therapy in persistent and non-persistent patients Indication  Non-persistent, mean (s.d.), €;  Persistent, mean (s.d.), €  Cost difference  P-value  Office-based visits      Total  3094 (4601)  2319 (2021)  775  <0.001      RA  3388 (5642)  2426 (1884)  962  0.005      AS  2447 (2261)  1809 (1573)  638  0.001      PsA  2152 (1563)  2897 (3321)  −745  0.387      Mixed  3530 (3407)  2249 (1701)  1281  0.012  Prescriptions (excluding TNF-α)      Total  5492 (9876)  2828 (13 332)  2664  <0.001      RA  6914 (10 109)  3183 (15 788)  3731  <0.001      AS  2369 (6719)  2807 (12 982)  −438  0.615      PsA  4621 (12 388)  1998 (3237)  2623  0.985      Mixed  5003 (10 245)  1921 (2376)  3082  0.662  Prescriptions (including TNF-α)      Total  27 222 (13 902)  43 707 (14 254)  −16 485  <0.001      RA  27 979 (13 418)  43 759 (16 403)  −15 780  <0.001      AS  24 561 (13 857)  43 926 (14 772)  −19 365  <0.001      PsA  30 223 (15 197)  43 629 (5319)  −13 407  <0.001      Mixed  26 271 (14 676)  43 127 (4727)  −16 856  <0.001  Service utilization      Total  2475 (8010)  2142 (7216)  333  0.079      RA  3125 (9920)  2142 (6760)  983  0.007      AS  1011 (2478)  2326 (9589)  −1315  0.422      PsA  1722 (4572)  1165 (2904)  557  0.925      Mixed  2564 (5732)  2649 (6884)  −85  0.117  Hospitalizations      Total  5890 (14 951)  3552 (11 135)  2338  <0.001      RA  6869 (13 868)  4189 (12 930)  2680  <0.001      AS  3051 (7323)  2027 (6708)  1024  0.157      PsA  8862 (31 399)  3466 (11 076)  5396  0.249      Mixed  4252 (10 753)  3429 (8027)  823  0.490  Costs associated with sick leave      Total  1241 (3774)  717 (2784)  524  0.011      RA  1342 (3918)  620 (2852)  722  0.007      AS  997 (3514)  1072 (2982)  −75  0.386      PsA  1330 (4525)  594 (1689)  736  0.937      Mixed  1135 (2898)  629 (2844)  506  0.007  Total costs (excluding TNF-α)      Total  18 192 (24 628)  11 558 (20 823)  6634  <0.001      RA  21 637 (25 851)  12 561 (23 801)  9076  <0.001      AS  9875 (12 416)  10 042 (19 240)  −167  0.143      PsA  18 687 (36 880)  10 119 (13 312)  8568  0.444      Mixed  16 485 (20 702)  10 877 (12 175)  5608  0.169  Total costs (including TNF-α)      Total  39 922 (25 670)  52 437 (21 521)  −12 515  <0.001      RA  42 703 (26 436)  53 138 (24 119)  −10 435  <0.001      AS  32 068 (17 211)  51 160 (21 153)  −19 092  <0.001      PsA  44 289 (37 175)  51 751 (14 273)  −7462  <0.001      Mixed  37 753 (21 575)  52 083 (12 929)  −14 330  <0.001  Indication  Non-persistent, mean (s.d.), €;  Persistent, mean (s.d.), €  Cost difference  P-value  Office-based visits      Total  3094 (4601)  2319 (2021)  775  <0.001      RA  3388 (5642)  2426 (1884)  962  0.005      AS  2447 (2261)  1809 (1573)  638  0.001      PsA  2152 (1563)  2897 (3321)  −745  0.387      Mixed  3530 (3407)  2249 (1701)  1281  0.012  Prescriptions (excluding TNF-α)      Total  5492 (9876)  2828 (13 332)  2664  <0.001      RA  6914 (10 109)  3183 (15 788)  3731  <0.001      AS  2369 (6719)  2807 (12 982)  −438  0.615      PsA  4621 (12 388)  1998 (3237)  2623  0.985      Mixed  5003 (10 245)  1921 (2376)  3082  0.662  Prescriptions (including TNF-α)      Total  27 222 (13 902)  43 707 (14 254)  −16 485  <0.001      RA  27 979 (13 418)  43 759 (16 403)  −15 780  <0.001      AS  24 561 (13 857)  43 926 (14 772)  −19 365  <0.001      PsA  30 223 (15 197)  43 629 (5319)  −13 407  <0.001      Mixed  26 271 (14 676)  43 127 (4727)  −16 856  <0.001  Service utilization      Total  2475 (8010)  2142 (7216)  333  0.079      RA  3125 (9920)  2142 (6760)  983  0.007      AS  1011 (2478)  2326 (9589)  −1315  0.422      PsA  1722 (4572)  1165 (2904)  557  0.925      Mixed  2564 (5732)  2649 (6884)  −85  0.117  Hospitalizations      Total  5890 (14 951)  3552 (11 135)  2338  <0.001      RA  6869 (13 868)  4189 (12 930)  2680  <0.001      AS  3051 (7323)  2027 (6708)  1024  0.157      PsA  8862 (31 399)  3466 (11 076)  5396  0.249      Mixed  4252 (10 753)  3429 (8027)  823  0.490  Costs associated with sick leave      Total  1241 (3774)  717 (2784)  524  0.011      RA  1342 (3918)  620 (2852)  722  0.007      AS  997 (3514)  1072 (2982)  −75  0.386      PsA  1330 (4525)  594 (1689)  736  0.937      Mixed  1135 (2898)  629 (2844)  506  0.007  Total costs (excluding TNF-α)      Total  18 192 (24 628)  11 558 (20 823)  6634  <0.001      RA  21 637 (25 851)  12 561 (23 801)  9076  <0.001      AS  9875 (12 416)  10 042 (19 240)  −167  0.143      PsA  18 687 (36 880)  10 119 (13 312)  8568  0.444      Mixed  16 485 (20 702)  10 877 (12 175)  5608  0.169  Total costs (including TNF-α)      Total  39 922 (25 670)  52 437 (21 521)  −12 515  <0.001      RA  42 703 (26 436)  53 138 (24 119)  −10 435  <0.001      AS  32 068 (17 211)  51 160 (21 153)  −19 092  <0.001      PsA  44 289 (37 175)  51 751 (14 273)  −7462  <0.001      Mixed  37 753 (21 575)  52 083 (12 929)  −14 330  <0.001  Table 2 Average costs per patient within 2 years after initiation of s.c. TNFi therapy in persistent and non-persistent patients Indication  Non-persistent, mean (s.d.), €;  Persistent, mean (s.d.), €  Cost difference  P-value  Office-based visits      Total  3094 (4601)  2319 (2021)  775  <0.001      RA  3388 (5642)  2426 (1884)  962  0.005      AS  2447 (2261)  1809 (1573)  638  0.001      PsA  2152 (1563)  2897 (3321)  −745  0.387      Mixed  3530 (3407)  2249 (1701)  1281  0.012  Prescriptions (excluding TNF-α)      Total  5492 (9876)  2828 (13 332)  2664  <0.001      RA  6914 (10 109)  3183 (15 788)  3731  <0.001      AS  2369 (6719)  2807 (12 982)  −438  0.615      PsA  4621 (12 388)  1998 (3237)  2623  0.985      Mixed  5003 (10 245)  1921 (2376)  3082  0.662  Prescriptions (including TNF-α)      Total  27 222 (13 902)  43 707 (14 254)  −16 485  <0.001      RA  27 979 (13 418)  43 759 (16 403)  −15 780  <0.001      AS  24 561 (13 857)  43 926 (14 772)  −19 365  <0.001      PsA  30 223 (15 197)  43 629 (5319)  −13 407  <0.001      Mixed  26 271 (14 676)  43 127 (4727)  −16 856  <0.001  Service utilization      Total  2475 (8010)  2142 (7216)  333  0.079      RA  3125 (9920)  2142 (6760)  983  0.007      AS  1011 (2478)  2326 (9589)  −1315  0.422      PsA  1722 (4572)  1165 (2904)  557  0.925      Mixed  2564 (5732)  2649 (6884)  −85  0.117  Hospitalizations      Total  5890 (14 951)  3552 (11 135)  2338  <0.001      RA  6869 (13 868)  4189 (12 930)  2680  <0.001      AS  3051 (7323)  2027 (6708)  1024  0.157      PsA  8862 (31 399)  3466 (11 076)  5396  0.249      Mixed  4252 (10 753)  3429 (8027)  823  0.490  Costs associated with sick leave      Total  1241 (3774)  717 (2784)  524  0.011      RA  1342 (3918)  620 (2852)  722  0.007      AS  997 (3514)  1072 (2982)  −75  0.386      PsA  1330 (4525)  594 (1689)  736  0.937      Mixed  1135 (2898)  629 (2844)  506  0.007  Total costs (excluding TNF-α)      Total  18 192 (24 628)  11 558 (20 823)  6634  <0.001      RA  21 637 (25 851)  12 561 (23 801)  9076  <0.001      AS  9875 (12 416)  10 042 (19 240)  −167  0.143      PsA  18 687 (36 880)  10 119 (13 312)  8568  0.444      Mixed  16 485 (20 702)  10 877 (12 175)  5608  0.169  Total costs (including TNF-α)      Total  39 922 (25 670)  52 437 (21 521)  −12 515  <0.001      RA  42 703 (26 436)  53 138 (24 119)  −10 435  <0.001      AS  32 068 (17 211)  51 160 (21 153)  −19 092  <0.001      PsA  44 289 (37 175)  51 751 (14 273)  −7462  <0.001      Mixed  37 753 (21 575)  52 083 (12 929)  −14 330  <0.001  Indication  Non-persistent, mean (s.d.), €;  Persistent, mean (s.d.), €  Cost difference  P-value  Office-based visits      Total  3094 (4601)  2319 (2021)  775  <0.001      RA  3388 (5642)  2426 (1884)  962  0.005      AS  2447 (2261)  1809 (1573)  638  0.001      PsA  2152 (1563)  2897 (3321)  −745  0.387      Mixed  3530 (3407)  2249 (1701)  1281  0.012  Prescriptions (excluding TNF-α)      Total  5492 (9876)  2828 (13 332)  2664  <0.001      RA  6914 (10 109)  3183 (15 788)  3731  <0.001      AS  2369 (6719)  2807 (12 982)  −438  0.615      PsA  4621 (12 388)  1998 (3237)  2623  0.985      Mixed  5003 (10 245)  1921 (2376)  3082  0.662  Prescriptions (including TNF-α)      Total  27 222 (13 902)  43 707 (14 254)  −16 485  <0.001      RA  27 979 (13 418)  43 759 (16 403)  −15 780  <0.001      AS  24 561 (13 857)  43 926 (14 772)  −19 365  <0.001      PsA  30 223 (15 197)  43 629 (5319)  −13 407  <0.001      Mixed  26 271 (14 676)  43 127 (4727)  −16 856  <0.001  Service utilization      Total  2475 (8010)  2142 (7216)  333  0.079      RA  3125 (9920)  2142 (6760)  983  0.007      AS  1011 (2478)  2326 (9589)  −1315  0.422      PsA  1722 (4572)  1165 (2904)  557  0.925      Mixed  2564 (5732)  2649 (6884)  −85  0.117  Hospitalizations      Total  5890 (14 951)  3552 (11 135)  2338  <0.001      RA  6869 (13 868)  4189 (12 930)  2680  <0.001      AS  3051 (7323)  2027 (6708)  1024  0.157      PsA  8862 (31 399)  3466 (11 076)  5396  0.249      Mixed  4252 (10 753)  3429 (8027)  823  0.490  Costs associated with sick leave      Total  1241 (3774)  717 (2784)  524  0.011      RA  1342 (3918)  620 (2852)  722  0.007      AS  997 (3514)  1072 (2982)  −75  0.386      PsA  1330 (4525)  594 (1689)  736  0.937      Mixed  1135 (2898)  629 (2844)  506  0.007  Total costs (excluding TNF-α)      Total  18 192 (24 628)  11 558 (20 823)  6634  <0.001      RA  21 637 (25 851)  12 561 (23 801)  9076  <0.001      AS  9875 (12 416)  10 042 (19 240)  −167  0.143      PsA  18 687 (36 880)  10 119 (13 312)  8568  0.444      Mixed  16 485 (20 702)  10 877 (12 175)  5608  0.169  Total costs (including TNF-α)      Total  39 922 (25 670)  52 437 (21 521)  −12 515  <0.001      RA  42 703 (26 436)  53 138 (24 119)  −10 435  <0.001      AS  32 068 (17 211)  51 160 (21 153)  −19 092  <0.001      PsA  44 289 (37 175)  51 751 (14 273)  −7462  <0.001      Mixed  37 753 (21 575)  52 083 (12 929)  −14 330  <0.001  However, as expected, when including the cost of s.c. TNFi prescriptions, the total cost per patient was higher in the persistent cohort than in the non-persistent cohort (Table 2). Discussion This retrospective study of 1356 patients found that the 2-year costs per patient for office-based visits, hospitalizations, co-medications and sick leave costs were higher in patients who discontinued their s.c. TNFi therapy after an average of 9 months compared with patients with at least 24 months of therapy. Moreover, the average duration of hospitalization was longer in non-persistent patients. Since the database used does not include data on symptoms and diagnosis severity, costs can be used as a surrogate parameter for patients’ health status. Evidence regarding the share of patients who continued to do well after discontinuation of anti-TNF therapy is lacking. However, even if only some patients who discontinue s.c. TNFi therapy experience an increase in disease activity, this can result in more physician visits, longer hospitalizations and more days of sick leave. On the one hand, this can be caused by the IMRD symptoms themselves. On the other hand, increased time-averaged disease activity in RA can be associated with a higher risk of cardiovascular events [18] as well as a higher probability of developing infections [19]. The difference in non-TNF medication in our study was ∼54%, which is in line with results of Dalen et al. [14] who reported a difference of 56%. Regardless of the diagnosis, but especially in the case of diseases accompanied by pain, better persistence and greater adherence is associated with a considerable reduction in cost. This does not refer to total costs but rather to co-therapy and hospitalizations. Therefore such cost savings provide indirect but clear evidence regarding quality of life. TNFi products are relatively costly and patients with longer treatment durations, that is, better persistence, receive a greater number of prescriptions, which leads to higher treatment costs, while at the same time reducing other costs (co-therapies, hospital, service utilization, sick leave costs). However, longer therapy can achieve an improvement in the quality of life of IMRD patients. Moreover, the value of good persistence can be considered in terms of benefits to all participants in the health care system, including health care providers and payers. Study limitations Each retrospective study is limited by the availability of data in this database. The specific markers or tests of disease severity and the reasons for discontinuation are not recorded in the database. Furthermore, analyses are based on data of the Betriebskrankenkasse insurance fund, which is only one of many insurance funds in Germany, and the patient characteristics of different funds may vary. Moreover, no privately insured patients are included in the analyses. Finally, this study used mean values for costs per patient. However, mean values are highly sensitive to outliers. In this study, service utilization costs in patients with AS were €1011 (s.d. 2478) in non-persistent and €2326 (s.d. 9589) in persistent patients, due to some outliers with very high costs in persistent patients. On the other side, the median costs were higher in non-persistent vs persistent patients (€262 vs €200), showing the role of outliers. The same statistical effect of outliers can be observed in office-based visit costs of patients with PsA. Conclusion The results of this study indicate that persistence with s.c. TNFi treatment can be associated with several cost offsets for IMRD patients, with an improvement in their quality of life. For health care providers and payers, these findings reinforce the value of persistence, given the additional economic burden associated with non-persistent patients. Funding: This study was funded by Merck & Co., Kenilworth, NJ, USA. The sponsor was involved with the writing, study design, analysis and interpretation of data. Disclosure statement: K.Z., K.K., M.H. and S.D. are employees of QuintilesIMS. QuintilesIMS were paid consultants to Merck & Co., Kenilworth, NJ USA, in conjunction with the development of this manuscript. M.F. and H.F. are employees of Team Gesundheit Gesellschaft für Gesundheitsmanagement. S.K. is an employee of Merck & Co., Kenilworth, NJ, USA, and holds stock and options. References 1 Silman AJ, Hochberg MC. Epidemiology of the rheumatic diseases . New York, USA: Oxford University Press, 2001. 2 Smolen JS, Landewe R, Breedveld FC et al.   EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs. Ann Rheum Dis  2010; 69: 964– 75. Google Scholar CrossRef Search ADS PubMed  3 Braun J, Berg R, Boehm H et al.   2010 update of the ASAS/EULAR recommendations for the management of ankylosing spondylitis. Ann Rheum Dis  2011; 70: 896– 904. Google Scholar CrossRef Search ADS PubMed  4 Gossec L, Smolen JS, Gaujoux-Viala C et al.   European League Against Rheumatism recommendations for the management of psoriatic arthritis with pharmacological therapies. Ann Rheum Dis  2012; 71: 4– 12. Google Scholar CrossRef Search ADS PubMed  5 Stevens SR, Chang TH. History of development of TNF inhibitors . Basel, Switzerland: Springer, 2006: 9– 22. Google Scholar CrossRef Search ADS   6 Maxwell LJ, Zochling J, Boonen A et al.   TNF-alpha inhibitors for ankylosing spondylitis. Cochrane Database Syst Rev  2015; 4: CD005468. 7 Aaltonen KJ, Virkki LM, Malmivaara A et al.   Systematic review and meta-analysis of the efficacy and safety of existing TNF blocking agents in treatment of rheumatoid arthritis. PLoS One  2012; 7: e30275. Google Scholar CrossRef Search ADS PubMed  8 Lemos LL, de Oliveira Costa J, Almeida AM et al.   Treatment of psoriatic arthritis with anti-TNF agents: a systematic review and meta-analysis of efficacy, effectiveness and safety. Rheumatol Int  2014; 34: 1345– 60. Google Scholar CrossRef Search ADS PubMed  9 Aletaha D, Smolen JS. Effectiveness profiles and dose dependent retention of traditional disease modifying antirheumatic drugs for rheumatoid arthritis. An observational study. J Rheumatol  2002; 29: 1631– 8. Google Scholar PubMed  10 Fries JF. Effectiveness and toxicity considerations in outcome directed therapy in rheumatoid arthritis. J Rheumatol Suppl  1996; 44: 102– 6. Google Scholar PubMed  11 Pincus T, Marcum SB, Callahan LF. Long-term drug therapy for rheumatoid arthritis in seven rheumatology private practices: II. Second line drugs and prednisone. J Rheumatol  1992; 19: 1885– 94. Google Scholar PubMed  12 Emery P, Keystone E, Tony HP et al.   IL-6 receptor inhibition with tocilizumab improves treatment outcomes in patients with rheumatoid arthritis refractory to anti-tumour necrosis factor biologicals: results from a 24-week multicentre randomised placebo-controlled trial. Ann Rheum Dis  2008; 67: 1516– 23. Google Scholar CrossRef Search ADS PubMed  13 Salaffi F, Carotti M, Gasparini S, Intorcia M, Grassi W. The health-related quality of life in rheumatoid arthritis, ankylosing spondylitis, and psoriatic arthritis: a comparison with a selected sample of healthy people. Health Qual Life Outcomes  2009; 7: 25. Google Scholar CrossRef Search ADS PubMed  14 Dalén J, Svedbom A, Black CM et al.   Treatment persistence among patients with immune-mediated rheumatic disease newly treated with subcutaneous TNF-alpha inhibitors and costs associated with non-persistence. Rheumatol Int  2016; 36: 987– 95. Google Scholar CrossRef Search ADS PubMed  15 Dalén J, Svedbom A, Black CM, Kachroo S. Second-line treatment persistence and costs among patients with immune-mediated rheumatic diseases treated with subcutaneous TNF-alpha inhibitors. Rheumatol Int  2017; 37: 2049– 58. Google Scholar CrossRef Search ADS PubMed  16 Svedbom A, Dalén J,M, Black CM, Kachroo S. Persistence and costs with subcutaneous TNF-alpha inhibitors in immune-mediated rheumatic disease stratified by treatment line. Patient Pref Adherence  2017; 11: 95– 106. Google Scholar CrossRef Search ADS   17 Friedel H, Delges A, Clouth J, Trautvetter DT. Expenditures of the German statutory health insurance system for patients suffering from acute coronary syndrome and treated with percutaneous coronary intervention. Eur J Health Econ  2010; 11: 449– 55. Google Scholar CrossRef Search ADS PubMed  18 Solomon DH, Reed GW, Kremer JM et al.   Disease activity in rheumatoid arthritis and the risk of cardiovascular events. Arthritis Rheumatol  2015; 67: 1449– 55. Google Scholar CrossRef Search ADS PubMed  19 Au K, Reed G, Curtis JR et al.   High disease activity is associated with an increased risk of infection in patients with rheumatoid arthritis. Ann Rheum Dis  2011; 70: 785– 91. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Rheumatology Oxford University Press

The impact of non-persistence on the direct and indirect costs in patients treated with subcutaneous tumour necrosis factor-alpha inhibitors in Germany

Loading next page...
 
/lp/ou_press/the-impact-of-non-persistence-on-the-direct-and-indirect-costs-in-LWSBp8vFiN
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For permissions, please email: journals.permissions@oup.com
ISSN
1462-0324
eISSN
1462-0332
D.O.I.
10.1093/rheumatology/key099
Publisher site
See Article on Publisher Site

Abstract

Abstract Objective The goal of the present study was to estimate the treatment costs in immune-mediated rheumatic disease patients initiating treatment with an s.c. biologic agent based on treatment persistence. Methods This is a retrospective cohort study based on the German statutory health insurance funds database. Patients ⩾18 years of age with a diagnosis of AS, PsA or RA treated with s.c. TNF-α inhibitors (TNFis) were included. Persistence was estimated as the duration of time from s.c. TNFi therapy initiation to discontinuation, which was defined as at least 60 days without therapy. We performed 1:1 matching based on a propensity score that was constructed as the conditional probability of being persistent as a function of age, gender, index year, physician specialty and Charlson comorbidity index. Finally, the cost differences between the matched pairs were estimated using the Wilcoxon test. Results After 1:1 matching, 678 persistent and 678 non-persistent patients were available for cost analyses. Using a 2-year time period, the costs for office-based visits per patient were €2319 in the persistent cohort compared with €3094 in the non-persistent cohort (P < 0.001). Co-medication costs were €2828 in the persistent cohort compared with €5498 in the non-persistent cohort, hospitalization costs were €3551 in the persistent cohort compared with €5890 in the non-persistent cohort and sick leave costs were €717 in the persistent cohort compared with €1241 in the non-persistent cohort (all P < 0.001). Conclusion The results of this study indicate that persistence with s.c. TNFi treatment can be associated with several cost offsets for immune-mediated rheumatic disease patients. IMRD, persistence, TNF-α, blocker, treatment costs Rheumatology key messages Biologic treatment persistence is associated with several cost offsets for immune-mediated rheumatic disease patients. In immune-mediated rheumatic disease patients, direct medical costs were lower in the persistent cohort. Introduction AS, PsA and RA are chronic progressive immune-mediated rheumatic diseases (IMRDs) that can cause pain, deformity and disability and progressively impair joint structure and function [1]. IMRDs result in substantial personal as well as economic burden [2]. The primary treatment goal for all three conditions is to maximize health-related quality of life by controlling symptoms and inflammation, preventing progressive structural damage, preserving or normalizing function and social participation and targeting remission. Treatment of rheumatic diseases usually involves a multimodal approach including both pharmacological and non-pharmacological strategies [3, 4]. The advent of biologic medications such as s.c. TNF-α inhibitors (TNFis) has transformed the management of IMRDs [5]. Previous research has provided high-quality evidence that anti-TNF agents improve clinical symptoms in the treatment of IMRDs [6–8]. Taking the prescribed medication for a sufficient period of time is crucial to the success of any therapy. Non-compliance and a lack of persistence with prescription medications are major problems in the clinical management of chronic diseases and s.c. TNFi treatment is potentially a very convenient option for patients. However, patients may likely have drug interruptions or long periods of non-persistence. In IMRDs, treatment persistence can be used to represent drug effectiveness, safety and treatment satisfaction [9–11]. While associations between adherence and costs [12] and the cost impact of switching from an s.c. TNFi to different biologic classes have been examined [5, 13], studies estimating the cost impact of non-persistence in patients treated with s.c. TNFis for IMRDs are lacking. In 2016, Dalén et al. [14] analysed treatment costs in first-line (bio-naïve) IMRD patients in Sweden based on persistent and non-persistent patients. The results of their study indicated that persistence with s.c. TNFi treatment may be associated with some cost offsets for IMRD patients. In 2017, Dalén et al. [15] further showed that, in second-line therapy with s.c. TNFis, persistent patients had lower mean total costs. In the same year, Svedbom et al. [16] described and compared treatment persistence with first- and second-line s.c. TNFis as well as their corresponding costs in Sweden. Although these works are of great interest, no recent data are available about the association between non-persistence with s.c. TNFis and costs in Germany. The goal of the present study was to estimate and compare the direct and indirect treatment costs in IMRD patients initiating treatment with an s.c. biologic agent based on treatment persistence. Methods Data source This study is a retrospective cohort study based on the German statutory health insurance funds database. This database continuously provides data pertaining to about five million statutorily insured people, which corresponds to 7% coverage. Exclusively anonymous information and no personal data (in accordance with § 3 Abs. 6 German Federal Data Protection Act) are used. Each patient with defined characteristics (e.g. specific disease, number and/or durations of hospitalizations, prescriptions of specific medications or any combination of these characteristics) can be tracked within the database via a unique patient ID. Finally, the database has already been used in several epidemiological studies, including cost analyses [17]. General informed consent Informed patient consent German law allows the use of anonymous electronic medical records for research purposes under certain conditions. Because patients were only queried as aggregates and no protected health information was available for queries, no institutional review board approval was required for the use of this database or completion of this study. For this study, we only used anonymized data collected from the claims database. Therefore, this study does not require that we obtain informed consent from patients. Study population Patients ⩾18 years of age with a diagnosis of RA [International Classification of Diseases, 10th Revision (ICD-10) M05, M06], PsA (ICD-10 L40.5 in combination with M07.0, M07.1, M07.2, M07.3 or M09.0) or AS (ICD-10 M45) treated with any of the four s.c. TNF-α blockers of interest (etanercept, adalimumab, certolizumab pegol or golimumab) were selected. Patients were categorized by their main diagnosis at the index date (the diagnosis that is documented within the quarter of the index prescription date). Cases with more than one documented diagnosis were classified as a mixed group. The date of the first qualifying prescription for an s.c. biologic agent served as the patient’s index date. The 12-month period immediately preceding the index date was used as the pre-index period and the 12-month period following the final prescription within the persistence period served as the post-index period, which lasted no later than December 2015. All patients were traceable in the database for at least 12 months after their last prescription within the analysis period. The pre-index period and index date were used to measure patients’ baseline characteristics. Treatment persistence with index s.c. TNF-α blockers and health care resource utilization (such as services and direct and indirect costs) were evaluated in both the persistence and post-index periods. Persistence definition and patient stratification Persistence was defined as the duration of time from s.c. TNFi therapy initiation to discontinuation, which was identified as at least 60 days without s.c. TNFi therapy. A longitudinal dataset of medication supply was created for each individual patient and the number of days of drug supply was calculated based on the defined daily dose information associated with each prescription record. The persistence was evaluated at 24 months after the index date. Following this estimation of persistence, patients were categorized into persistent and non-persistent cohorts. The main outcome of the study was rheumatoid disease–related health care resource utilization costs, including direct costs and services. These costs were estimated for the time within 2 years (730 days) following the first s.c. TNFi prescription and separated by RA, PsA and AS indication. Direct costs included hospitalization costs, doctor and hospital visit costs, prescribed medication costs and sickness benefits (in Germany, sickness funds are obligated to provide sick pay for sick leaves of at least 43 days). Services consisted of the number of hospital stays, duration of hospitalization, medical department of the hospital involved and specialties visited. Covariates Demographic characteristics of patients at the index date included age (as a continuous variable and age groups) and gender. Further baseline variables were year of index date and the specialty of the physician initiating treatment. Furthermore, a revised version of the Charlson comorbidity index (CCI) was used as a generic marker of comorbidity. Statistical methods To control for confounding, 1:1 matching was carried out based on a propensity score that was constructed as the conditional probability of being persistent as a function of age, gender, index year, physician specialty and CCI (logistic regression). Greedy matching was used by choosing a non-persistent patient whose propensity score was closest to that of this randomly selected persistent subject for matching. Descriptive statistics were given and group differences (persistent vs non-persistent) were assessed using the Wilcoxon test or the McNemar’s tests after propensity score matching. Finally, the differences in direct costs, indirect costs and services between matched pairs were estimated using the Wilcoxon test. Results Patient characteristics A total of 2060 patients initiating s.c. TNFi therapy were included in the study. After 2 years of follow-up, 682 patients still received s.c. TNFi therapy and 1378 patients had discontinued therapy. In the cohort of non-persistent patients, the mean persistence duration was 266 days (s.d. 190). We observed no significant difference between persistent and non-persistent patients in terms of age, gender or year of therapy initiation. Non-persistent patients had a slightly worse average health status compared with the persistent cohort (CCI 2.6 vs 2.4) (Table 1). After 1:1 matching, 678 pairs (678 persistent and 678 non-persistent patients) were available for cost analyses. Both cohorts were similar in terms of age, gender, year of therapy initiation, CCI and indication (Table 1). Table 1 Baseline characteristics of study patients before and after matching Variable  Before matching   After matching   Persistent  Non-persistent  P-value  Persistent  Non-persistent  P-value  Total (N)  682   1378    678   678    Indication                  RA (%)  55.7  58.8  0.040  55.9  57.2  0.520      AS (%)  21.7  19.9  0.579  21.8  21.7  1.000      PsA (%)  10.7  8.1  0.277  10.3  8.7  0.454      Mixed (%)  11.9  13.3  0.369  11.9  12.4  0.803  Age, mean (s.d.), years  51.1 (13.2)  51.4 (14.1)  0.669  51.1 (13.2)  51.2 (14.0)  0.905  Gender, male, %  42.4  41.1  0.572  42.5  44.0  0.583  Index year, %                  2010  29.6  25.7  0.305  29.4  29.4  0.854      2011  23.6  24.5  23.6  25.5      2012  24.5  25.9  24.6  23.6      2013  22.3  23.9  22.4  21.5  CCI, mean (s.d.)  2.4 (1.3)  2.6 (1.6)  <0.001  2.0 (1.3)  2.0 (1.3)  0.9227  Variable  Before matching   After matching   Persistent  Non-persistent  P-value  Persistent  Non-persistent  P-value  Total (N)  682   1378    678   678    Indication                  RA (%)  55.7  58.8  0.040  55.9  57.2  0.520      AS (%)  21.7  19.9  0.579  21.8  21.7  1.000      PsA (%)  10.7  8.1  0.277  10.3  8.7  0.454      Mixed (%)  11.9  13.3  0.369  11.9  12.4  0.803  Age, mean (s.d.), years  51.1 (13.2)  51.4 (14.1)  0.669  51.1 (13.2)  51.2 (14.0)  0.905  Gender, male, %  42.4  41.1  0.572  42.5  44.0  0.583  Index year, %                  2010  29.6  25.7  0.305  29.4  29.4  0.854      2011  23.6  24.5  23.6  25.5      2012  24.5  25.9  24.6  23.6      2013  22.3  23.9  22.4  21.5  CCI, mean (s.d.)  2.4 (1.3)  2.6 (1.6)  <0.001  2.0 (1.3)  2.0 (1.3)  0.9227  CCI: Charlson comorbidity index. Table 1 Baseline characteristics of study patients before and after matching Variable  Before matching   After matching   Persistent  Non-persistent  P-value  Persistent  Non-persistent  P-value  Total (N)  682   1378    678   678    Indication                  RA (%)  55.7  58.8  0.040  55.9  57.2  0.520      AS (%)  21.7  19.9  0.579  21.8  21.7  1.000      PsA (%)  10.7  8.1  0.277  10.3  8.7  0.454      Mixed (%)  11.9  13.3  0.369  11.9  12.4  0.803  Age, mean (s.d.), years  51.1 (13.2)  51.4 (14.1)  0.669  51.1 (13.2)  51.2 (14.0)  0.905  Gender, male, %  42.4  41.1  0.572  42.5  44.0  0.583  Index year, %                  2010  29.6  25.7  0.305  29.4  29.4  0.854      2011  23.6  24.5  23.6  25.5      2012  24.5  25.9  24.6  23.6      2013  22.3  23.9  22.4  21.5  CCI, mean (s.d.)  2.4 (1.3)  2.6 (1.6)  <0.001  2.0 (1.3)  2.0 (1.3)  0.9227  Variable  Before matching   After matching   Persistent  Non-persistent  P-value  Persistent  Non-persistent  P-value  Total (N)  682   1378    678   678    Indication                  RA (%)  55.7  58.8  0.040  55.9  57.2  0.520      AS (%)  21.7  19.9  0.579  21.8  21.7  1.000      PsA (%)  10.7  8.1  0.277  10.3  8.7  0.454      Mixed (%)  11.9  13.3  0.369  11.9  12.4  0.803  Age, mean (s.d.), years  51.1 (13.2)  51.4 (14.1)  0.669  51.1 (13.2)  51.2 (14.0)  0.905  Gender, male, %  42.4  41.1  0.572  42.5  44.0  0.583  Index year, %                  2010  29.6  25.7  0.305  29.4  29.4  0.854      2011  23.6  24.5  23.6  25.5      2012  24.5  25.9  24.6  23.6      2013  22.3  23.9  22.4  21.5  CCI, mean (s.d.)  2.4 (1.3)  2.6 (1.6)  <0.001  2.0 (1.3)  2.0 (1.3)  0.9227  CCI: Charlson comorbidity index. Cost analyses Table 2 shows the mean per-patient costs in the persistent and non-persistent cohorts. The costs for office-based visits per patient were €2319 in the persistent cohort and €3094 in the non-persistent cohort (P < 0.001). The cost difference was the highest in the mixed indication group (Table 2). An average persistent patient received prescriptions of different co-medications worth €2828 and a non-persistent patient received prescriptions worth €5498 within the 2-year time period (P < 0.001). In this category, the highest cost difference was found for patients with RA. Hospitalization costs were €3551 per persistent patient and €5890 per non-persistent patients (Table 2), which was probably due to the longer average duration of hospitalization in the non-persistent cohort. Furthermore, sick leave costs were much lower in persistent (€717) vs non-persistent (€1241) patients. Total costs per patient (excluding the cost of s.c. TNFi therapy) were ∼€6634 lower in the persistent cohort (Table 2). Table 2 Average costs per patient within 2 years after initiation of s.c. TNFi therapy in persistent and non-persistent patients Indication  Non-persistent, mean (s.d.), €;  Persistent, mean (s.d.), €  Cost difference  P-value  Office-based visits      Total  3094 (4601)  2319 (2021)  775  <0.001      RA  3388 (5642)  2426 (1884)  962  0.005      AS  2447 (2261)  1809 (1573)  638  0.001      PsA  2152 (1563)  2897 (3321)  −745  0.387      Mixed  3530 (3407)  2249 (1701)  1281  0.012  Prescriptions (excluding TNF-α)      Total  5492 (9876)  2828 (13 332)  2664  <0.001      RA  6914 (10 109)  3183 (15 788)  3731  <0.001      AS  2369 (6719)  2807 (12 982)  −438  0.615      PsA  4621 (12 388)  1998 (3237)  2623  0.985      Mixed  5003 (10 245)  1921 (2376)  3082  0.662  Prescriptions (including TNF-α)      Total  27 222 (13 902)  43 707 (14 254)  −16 485  <0.001      RA  27 979 (13 418)  43 759 (16 403)  −15 780  <0.001      AS  24 561 (13 857)  43 926 (14 772)  −19 365  <0.001      PsA  30 223 (15 197)  43 629 (5319)  −13 407  <0.001      Mixed  26 271 (14 676)  43 127 (4727)  −16 856  <0.001  Service utilization      Total  2475 (8010)  2142 (7216)  333  0.079      RA  3125 (9920)  2142 (6760)  983  0.007      AS  1011 (2478)  2326 (9589)  −1315  0.422      PsA  1722 (4572)  1165 (2904)  557  0.925      Mixed  2564 (5732)  2649 (6884)  −85  0.117  Hospitalizations      Total  5890 (14 951)  3552 (11 135)  2338  <0.001      RA  6869 (13 868)  4189 (12 930)  2680  <0.001      AS  3051 (7323)  2027 (6708)  1024  0.157      PsA  8862 (31 399)  3466 (11 076)  5396  0.249      Mixed  4252 (10 753)  3429 (8027)  823  0.490  Costs associated with sick leave      Total  1241 (3774)  717 (2784)  524  0.011      RA  1342 (3918)  620 (2852)  722  0.007      AS  997 (3514)  1072 (2982)  −75  0.386      PsA  1330 (4525)  594 (1689)  736  0.937      Mixed  1135 (2898)  629 (2844)  506  0.007  Total costs (excluding TNF-α)      Total  18 192 (24 628)  11 558 (20 823)  6634  <0.001      RA  21 637 (25 851)  12 561 (23 801)  9076  <0.001      AS  9875 (12 416)  10 042 (19 240)  −167  0.143      PsA  18 687 (36 880)  10 119 (13 312)  8568  0.444      Mixed  16 485 (20 702)  10 877 (12 175)  5608  0.169  Total costs (including TNF-α)      Total  39 922 (25 670)  52 437 (21 521)  −12 515  <0.001      RA  42 703 (26 436)  53 138 (24 119)  −10 435  <0.001      AS  32 068 (17 211)  51 160 (21 153)  −19 092  <0.001      PsA  44 289 (37 175)  51 751 (14 273)  −7462  <0.001      Mixed  37 753 (21 575)  52 083 (12 929)  −14 330  <0.001  Indication  Non-persistent, mean (s.d.), €;  Persistent, mean (s.d.), €  Cost difference  P-value  Office-based visits      Total  3094 (4601)  2319 (2021)  775  <0.001      RA  3388 (5642)  2426 (1884)  962  0.005      AS  2447 (2261)  1809 (1573)  638  0.001      PsA  2152 (1563)  2897 (3321)  −745  0.387      Mixed  3530 (3407)  2249 (1701)  1281  0.012  Prescriptions (excluding TNF-α)      Total  5492 (9876)  2828 (13 332)  2664  <0.001      RA  6914 (10 109)  3183 (15 788)  3731  <0.001      AS  2369 (6719)  2807 (12 982)  −438  0.615      PsA  4621 (12 388)  1998 (3237)  2623  0.985      Mixed  5003 (10 245)  1921 (2376)  3082  0.662  Prescriptions (including TNF-α)      Total  27 222 (13 902)  43 707 (14 254)  −16 485  <0.001      RA  27 979 (13 418)  43 759 (16 403)  −15 780  <0.001      AS  24 561 (13 857)  43 926 (14 772)  −19 365  <0.001      PsA  30 223 (15 197)  43 629 (5319)  −13 407  <0.001      Mixed  26 271 (14 676)  43 127 (4727)  −16 856  <0.001  Service utilization      Total  2475 (8010)  2142 (7216)  333  0.079      RA  3125 (9920)  2142 (6760)  983  0.007      AS  1011 (2478)  2326 (9589)  −1315  0.422      PsA  1722 (4572)  1165 (2904)  557  0.925      Mixed  2564 (5732)  2649 (6884)  −85  0.117  Hospitalizations      Total  5890 (14 951)  3552 (11 135)  2338  <0.001      RA  6869 (13 868)  4189 (12 930)  2680  <0.001      AS  3051 (7323)  2027 (6708)  1024  0.157      PsA  8862 (31 399)  3466 (11 076)  5396  0.249      Mixed  4252 (10 753)  3429 (8027)  823  0.490  Costs associated with sick leave      Total  1241 (3774)  717 (2784)  524  0.011      RA  1342 (3918)  620 (2852)  722  0.007      AS  997 (3514)  1072 (2982)  −75  0.386      PsA  1330 (4525)  594 (1689)  736  0.937      Mixed  1135 (2898)  629 (2844)  506  0.007  Total costs (excluding TNF-α)      Total  18 192 (24 628)  11 558 (20 823)  6634  <0.001      RA  21 637 (25 851)  12 561 (23 801)  9076  <0.001      AS  9875 (12 416)  10 042 (19 240)  −167  0.143      PsA  18 687 (36 880)  10 119 (13 312)  8568  0.444      Mixed  16 485 (20 702)  10 877 (12 175)  5608  0.169  Total costs (including TNF-α)      Total  39 922 (25 670)  52 437 (21 521)  −12 515  <0.001      RA  42 703 (26 436)  53 138 (24 119)  −10 435  <0.001      AS  32 068 (17 211)  51 160 (21 153)  −19 092  <0.001      PsA  44 289 (37 175)  51 751 (14 273)  −7462  <0.001      Mixed  37 753 (21 575)  52 083 (12 929)  −14 330  <0.001  Table 2 Average costs per patient within 2 years after initiation of s.c. TNFi therapy in persistent and non-persistent patients Indication  Non-persistent, mean (s.d.), €;  Persistent, mean (s.d.), €  Cost difference  P-value  Office-based visits      Total  3094 (4601)  2319 (2021)  775  <0.001      RA  3388 (5642)  2426 (1884)  962  0.005      AS  2447 (2261)  1809 (1573)  638  0.001      PsA  2152 (1563)  2897 (3321)  −745  0.387      Mixed  3530 (3407)  2249 (1701)  1281  0.012  Prescriptions (excluding TNF-α)      Total  5492 (9876)  2828 (13 332)  2664  <0.001      RA  6914 (10 109)  3183 (15 788)  3731  <0.001      AS  2369 (6719)  2807 (12 982)  −438  0.615      PsA  4621 (12 388)  1998 (3237)  2623  0.985      Mixed  5003 (10 245)  1921 (2376)  3082  0.662  Prescriptions (including TNF-α)      Total  27 222 (13 902)  43 707 (14 254)  −16 485  <0.001      RA  27 979 (13 418)  43 759 (16 403)  −15 780  <0.001      AS  24 561 (13 857)  43 926 (14 772)  −19 365  <0.001      PsA  30 223 (15 197)  43 629 (5319)  −13 407  <0.001      Mixed  26 271 (14 676)  43 127 (4727)  −16 856  <0.001  Service utilization      Total  2475 (8010)  2142 (7216)  333  0.079      RA  3125 (9920)  2142 (6760)  983  0.007      AS  1011 (2478)  2326 (9589)  −1315  0.422      PsA  1722 (4572)  1165 (2904)  557  0.925      Mixed  2564 (5732)  2649 (6884)  −85  0.117  Hospitalizations      Total  5890 (14 951)  3552 (11 135)  2338  <0.001      RA  6869 (13 868)  4189 (12 930)  2680  <0.001      AS  3051 (7323)  2027 (6708)  1024  0.157      PsA  8862 (31 399)  3466 (11 076)  5396  0.249      Mixed  4252 (10 753)  3429 (8027)  823  0.490  Costs associated with sick leave      Total  1241 (3774)  717 (2784)  524  0.011      RA  1342 (3918)  620 (2852)  722  0.007      AS  997 (3514)  1072 (2982)  −75  0.386      PsA  1330 (4525)  594 (1689)  736  0.937      Mixed  1135 (2898)  629 (2844)  506  0.007  Total costs (excluding TNF-α)      Total  18 192 (24 628)  11 558 (20 823)  6634  <0.001      RA  21 637 (25 851)  12 561 (23 801)  9076  <0.001      AS  9875 (12 416)  10 042 (19 240)  −167  0.143      PsA  18 687 (36 880)  10 119 (13 312)  8568  0.444      Mixed  16 485 (20 702)  10 877 (12 175)  5608  0.169  Total costs (including TNF-α)      Total  39 922 (25 670)  52 437 (21 521)  −12 515  <0.001      RA  42 703 (26 436)  53 138 (24 119)  −10 435  <0.001      AS  32 068 (17 211)  51 160 (21 153)  −19 092  <0.001      PsA  44 289 (37 175)  51 751 (14 273)  −7462  <0.001      Mixed  37 753 (21 575)  52 083 (12 929)  −14 330  <0.001  Indication  Non-persistent, mean (s.d.), €;  Persistent, mean (s.d.), €  Cost difference  P-value  Office-based visits      Total  3094 (4601)  2319 (2021)  775  <0.001      RA  3388 (5642)  2426 (1884)  962  0.005      AS  2447 (2261)  1809 (1573)  638  0.001      PsA  2152 (1563)  2897 (3321)  −745  0.387      Mixed  3530 (3407)  2249 (1701)  1281  0.012  Prescriptions (excluding TNF-α)      Total  5492 (9876)  2828 (13 332)  2664  <0.001      RA  6914 (10 109)  3183 (15 788)  3731  <0.001      AS  2369 (6719)  2807 (12 982)  −438  0.615      PsA  4621 (12 388)  1998 (3237)  2623  0.985      Mixed  5003 (10 245)  1921 (2376)  3082  0.662  Prescriptions (including TNF-α)      Total  27 222 (13 902)  43 707 (14 254)  −16 485  <0.001      RA  27 979 (13 418)  43 759 (16 403)  −15 780  <0.001      AS  24 561 (13 857)  43 926 (14 772)  −19 365  <0.001      PsA  30 223 (15 197)  43 629 (5319)  −13 407  <0.001      Mixed  26 271 (14 676)  43 127 (4727)  −16 856  <0.001  Service utilization      Total  2475 (8010)  2142 (7216)  333  0.079      RA  3125 (9920)  2142 (6760)  983  0.007      AS  1011 (2478)  2326 (9589)  −1315  0.422      PsA  1722 (4572)  1165 (2904)  557  0.925      Mixed  2564 (5732)  2649 (6884)  −85  0.117  Hospitalizations      Total  5890 (14 951)  3552 (11 135)  2338  <0.001      RA  6869 (13 868)  4189 (12 930)  2680  <0.001      AS  3051 (7323)  2027 (6708)  1024  0.157      PsA  8862 (31 399)  3466 (11 076)  5396  0.249      Mixed  4252 (10 753)  3429 (8027)  823  0.490  Costs associated with sick leave      Total  1241 (3774)  717 (2784)  524  0.011      RA  1342 (3918)  620 (2852)  722  0.007      AS  997 (3514)  1072 (2982)  −75  0.386      PsA  1330 (4525)  594 (1689)  736  0.937      Mixed  1135 (2898)  629 (2844)  506  0.007  Total costs (excluding TNF-α)      Total  18 192 (24 628)  11 558 (20 823)  6634  <0.001      RA  21 637 (25 851)  12 561 (23 801)  9076  <0.001      AS  9875 (12 416)  10 042 (19 240)  −167  0.143      PsA  18 687 (36 880)  10 119 (13 312)  8568  0.444      Mixed  16 485 (20 702)  10 877 (12 175)  5608  0.169  Total costs (including TNF-α)      Total  39 922 (25 670)  52 437 (21 521)  −12 515  <0.001      RA  42 703 (26 436)  53 138 (24 119)  −10 435  <0.001      AS  32 068 (17 211)  51 160 (21 153)  −19 092  <0.001      PsA  44 289 (37 175)  51 751 (14 273)  −7462  <0.001      Mixed  37 753 (21 575)  52 083 (12 929)  −14 330  <0.001  However, as expected, when including the cost of s.c. TNFi prescriptions, the total cost per patient was higher in the persistent cohort than in the non-persistent cohort (Table 2). Discussion This retrospective study of 1356 patients found that the 2-year costs per patient for office-based visits, hospitalizations, co-medications and sick leave costs were higher in patients who discontinued their s.c. TNFi therapy after an average of 9 months compared with patients with at least 24 months of therapy. Moreover, the average duration of hospitalization was longer in non-persistent patients. Since the database used does not include data on symptoms and diagnosis severity, costs can be used as a surrogate parameter for patients’ health status. Evidence regarding the share of patients who continued to do well after discontinuation of anti-TNF therapy is lacking. However, even if only some patients who discontinue s.c. TNFi therapy experience an increase in disease activity, this can result in more physician visits, longer hospitalizations and more days of sick leave. On the one hand, this can be caused by the IMRD symptoms themselves. On the other hand, increased time-averaged disease activity in RA can be associated with a higher risk of cardiovascular events [18] as well as a higher probability of developing infections [19]. The difference in non-TNF medication in our study was ∼54%, which is in line with results of Dalen et al. [14] who reported a difference of 56%. Regardless of the diagnosis, but especially in the case of diseases accompanied by pain, better persistence and greater adherence is associated with a considerable reduction in cost. This does not refer to total costs but rather to co-therapy and hospitalizations. Therefore such cost savings provide indirect but clear evidence regarding quality of life. TNFi products are relatively costly and patients with longer treatment durations, that is, better persistence, receive a greater number of prescriptions, which leads to higher treatment costs, while at the same time reducing other costs (co-therapies, hospital, service utilization, sick leave costs). However, longer therapy can achieve an improvement in the quality of life of IMRD patients. Moreover, the value of good persistence can be considered in terms of benefits to all participants in the health care system, including health care providers and payers. Study limitations Each retrospective study is limited by the availability of data in this database. The specific markers or tests of disease severity and the reasons for discontinuation are not recorded in the database. Furthermore, analyses are based on data of the Betriebskrankenkasse insurance fund, which is only one of many insurance funds in Germany, and the patient characteristics of different funds may vary. Moreover, no privately insured patients are included in the analyses. Finally, this study used mean values for costs per patient. However, mean values are highly sensitive to outliers. In this study, service utilization costs in patients with AS were €1011 (s.d. 2478) in non-persistent and €2326 (s.d. 9589) in persistent patients, due to some outliers with very high costs in persistent patients. On the other side, the median costs were higher in non-persistent vs persistent patients (€262 vs €200), showing the role of outliers. The same statistical effect of outliers can be observed in office-based visit costs of patients with PsA. Conclusion The results of this study indicate that persistence with s.c. TNFi treatment can be associated with several cost offsets for IMRD patients, with an improvement in their quality of life. For health care providers and payers, these findings reinforce the value of persistence, given the additional economic burden associated with non-persistent patients. Funding: This study was funded by Merck & Co., Kenilworth, NJ, USA. The sponsor was involved with the writing, study design, analysis and interpretation of data. Disclosure statement: K.Z., K.K., M.H. and S.D. are employees of QuintilesIMS. QuintilesIMS were paid consultants to Merck & Co., Kenilworth, NJ USA, in conjunction with the development of this manuscript. M.F. and H.F. are employees of Team Gesundheit Gesellschaft für Gesundheitsmanagement. S.K. is an employee of Merck & Co., Kenilworth, NJ, USA, and holds stock and options. References 1 Silman AJ, Hochberg MC. Epidemiology of the rheumatic diseases . New York, USA: Oxford University Press, 2001. 2 Smolen JS, Landewe R, Breedveld FC et al.   EULAR recommendations for the management of rheumatoid arthritis with synthetic and biological disease-modifying antirheumatic drugs. Ann Rheum Dis  2010; 69: 964– 75. Google Scholar CrossRef Search ADS PubMed  3 Braun J, Berg R, Boehm H et al.   2010 update of the ASAS/EULAR recommendations for the management of ankylosing spondylitis. Ann Rheum Dis  2011; 70: 896– 904. Google Scholar CrossRef Search ADS PubMed  4 Gossec L, Smolen JS, Gaujoux-Viala C et al.   European League Against Rheumatism recommendations for the management of psoriatic arthritis with pharmacological therapies. Ann Rheum Dis  2012; 71: 4– 12. Google Scholar CrossRef Search ADS PubMed  5 Stevens SR, Chang TH. History of development of TNF inhibitors . Basel, Switzerland: Springer, 2006: 9– 22. Google Scholar CrossRef Search ADS   6 Maxwell LJ, Zochling J, Boonen A et al.   TNF-alpha inhibitors for ankylosing spondylitis. Cochrane Database Syst Rev  2015; 4: CD005468. 7 Aaltonen KJ, Virkki LM, Malmivaara A et al.   Systematic review and meta-analysis of the efficacy and safety of existing TNF blocking agents in treatment of rheumatoid arthritis. PLoS One  2012; 7: e30275. Google Scholar CrossRef Search ADS PubMed  8 Lemos LL, de Oliveira Costa J, Almeida AM et al.   Treatment of psoriatic arthritis with anti-TNF agents: a systematic review and meta-analysis of efficacy, effectiveness and safety. Rheumatol Int  2014; 34: 1345– 60. Google Scholar CrossRef Search ADS PubMed  9 Aletaha D, Smolen JS. Effectiveness profiles and dose dependent retention of traditional disease modifying antirheumatic drugs for rheumatoid arthritis. An observational study. J Rheumatol  2002; 29: 1631– 8. Google Scholar PubMed  10 Fries JF. Effectiveness and toxicity considerations in outcome directed therapy in rheumatoid arthritis. J Rheumatol Suppl  1996; 44: 102– 6. Google Scholar PubMed  11 Pincus T, Marcum SB, Callahan LF. Long-term drug therapy for rheumatoid arthritis in seven rheumatology private practices: II. Second line drugs and prednisone. J Rheumatol  1992; 19: 1885– 94. Google Scholar PubMed  12 Emery P, Keystone E, Tony HP et al.   IL-6 receptor inhibition with tocilizumab improves treatment outcomes in patients with rheumatoid arthritis refractory to anti-tumour necrosis factor biologicals: results from a 24-week multicentre randomised placebo-controlled trial. Ann Rheum Dis  2008; 67: 1516– 23. Google Scholar CrossRef Search ADS PubMed  13 Salaffi F, Carotti M, Gasparini S, Intorcia M, Grassi W. The health-related quality of life in rheumatoid arthritis, ankylosing spondylitis, and psoriatic arthritis: a comparison with a selected sample of healthy people. Health Qual Life Outcomes  2009; 7: 25. Google Scholar CrossRef Search ADS PubMed  14 Dalén J, Svedbom A, Black CM et al.   Treatment persistence among patients with immune-mediated rheumatic disease newly treated with subcutaneous TNF-alpha inhibitors and costs associated with non-persistence. Rheumatol Int  2016; 36: 987– 95. Google Scholar CrossRef Search ADS PubMed  15 Dalén J, Svedbom A, Black CM, Kachroo S. Second-line treatment persistence and costs among patients with immune-mediated rheumatic diseases treated with subcutaneous TNF-alpha inhibitors. Rheumatol Int  2017; 37: 2049– 58. Google Scholar CrossRef Search ADS PubMed  16 Svedbom A, Dalén J,M, Black CM, Kachroo S. Persistence and costs with subcutaneous TNF-alpha inhibitors in immune-mediated rheumatic disease stratified by treatment line. Patient Pref Adherence  2017; 11: 95– 106. Google Scholar CrossRef Search ADS   17 Friedel H, Delges A, Clouth J, Trautvetter DT. Expenditures of the German statutory health insurance system for patients suffering from acute coronary syndrome and treated with percutaneous coronary intervention. Eur J Health Econ  2010; 11: 449– 55. Google Scholar CrossRef Search ADS PubMed  18 Solomon DH, Reed GW, Kremer JM et al.   Disease activity in rheumatoid arthritis and the risk of cardiovascular events. Arthritis Rheumatol  2015; 67: 1449– 55. Google Scholar CrossRef Search ADS PubMed  19 Au K, Reed G, Curtis JR et al.   High disease activity is associated with an increased risk of infection in patients with rheumatoid arthritis. Ann Rheum Dis  2011; 70: 785– 91. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2018. Published by Oxford University Press on behalf of the British Society for Rheumatology. All rights reserved. For permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

RheumatologyOxford University Press

Published: Apr 12, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off