Treatment persistence is a proxy for efficacy, safety and patient satisfaction, and a switch in treatment or treatment discontinuation has been associated with increased indirect and direct costs in inflammatory arthritis (IA). Hence, there are both clinical and economic incentives for the identification of factors associated with treatment persistence. Until now, studies have mainly leveraged traditional regression analysis, but it has been suggested that novel approaches, such as statistical learning techniques, may improve our understanding of factors related to treatment persistence. Therefore, we set up a study using nationwide Swedish high-coverage administrative register data with the objective to identify patient groups with distinct persistence of subcutaneous tumor necrosis factor inhibitor (SC-TNFi) treatment in IA, using recursive partitioning, a statistical learning algorithm. IA was defined as a diagnosis of rheumatic arthritis (RA), ankylosing spondylitis/unspecified spondyloarthritis (AS/uSpA) or psoriatic arthritis (PsA). Adult swedish biologic-naïve patients with IA initiating biologic treatment with a SC-TNFi (adalimumab, etanercept, certolizumab or golimumab) between May 6, 2010, and December 31, 2017. Treatment persistence of SC-TNFi was derived based on prescription data and a defined standard daily dose. Patient characteristics, including age, sex, number of health care contacts, comorbidities and treatment, were collected at treatment initiation and 12 months before treatment initiation. Based on these characteristics, we used recursive partitioning in a conditional inference framework to identify patient groups with distinct SC-TNFi treatment persistence by IA diagnosis. A total of 13,913 patients were included. Approximately 50

Introduction Sickle cell disease (SCD) describes a group of inherited disorders of hemoglobin. Globally, SCD occurs in approximately 300,000-400,000 births annually and is most prevalent in malaria-endemic countries. However, migration has impacted the epidemiology of SCD but data on the matter are scarce. The objective of this study was to describe the epidemiology, treatment uptake, and economic burden of SCD in Sweden, a country with substantial immigration over the last decades. Methods This nationwide retrospective observational registry cohort study identified patients with SCD from 2001 to 2018 and followed them from 2006 to 2018. Using data from high-quality population-based Swedish registers, we estimated prevalence, treatment uptake, and SCD-related health care resource use, sick leave and disability pension. Results Between 2006 and 2018 the number of patients with SCD increased from 504 to 670; inpatient hospital stays and outpatient visits increased by 200% and 300%, respectively. Patients with pain crises had approximately twice the number of inpatient episodes and outpatient visit per year, and had higher productivity losses compared to patients without crises. Conclusion In an era of emerging treatments for SCD, we have, to the best of our knowledge, for the first time comprehensively described epidemiological and economic aspects of SCD in a country where the disease is still rare and not well recognized by the healthcare system.

Background Survival heterogeneity and limited trial follow-up present challenges for estimating lifetime benefits of oncology therapies. This study used CheckMate 067 (NCT01844505) extended follow-up data to assess the predictive accuracy of standard parametric and flexible models in estimating the long-term overall survival benefit of nivolumab plus ipilimumab (an immune checkpoint inhibitor combination) in advanced melanoma. Methods Six sets of survival models (standard parametric, piecewise, cubic spline, mixture cure, parametric mixture, and landmark response models) were independently fitted to overall survival data for treatments in CheckMate 067 (nivolumab plus ipilimumab, nivolumab, and ipilimumab) using successive data cuts (28, 40, 52, and 60 mo). Standard parametric models allow survival extrapolation in the absence of a complex hazard. Piecewise and cubic spline models allow additional flexibility in fitting the hazard function. Mixture cure, parametric mixture, and landmark response models provide flexibility by explicitly incorporating survival heterogeneity. Sixty-month follow-up data, external ipilimumab data, and clinical expert opinion were used to evaluate model estimation accuracy. Lifetime survival projections were compared using a 5% discount rate. Results Standard parametric, piecewise, and cubic spline models underestimated overall survival at 60 mo for the 28-mo data cut. Compared with other models, mixture cure, parametric mixture, and landmark response models provided more accurate long-term overall survival estimates versus external data, higher mean survival benefit over 20 y for the 28-mo data cut, and more consistent 20-y mean overall survival estimates across data cuts. Conclusion This case study demonstrates that survival models explicitly incorporating survival heterogeneity showed greater accuracy for early data cuts than standard parametric models did, consistent with similar immune checkpoint inhibitor survival validation studies in advanced melanoma. Research is required to assess generalizability to other tumors and disease stages. Highlights Given that short clinical trial follow-up periods and survival heterogeneity introduce uncertainty in the health technology assessment of oncology therapies, this study evaluated the suitability of conventional parametric survival modeling approaches as compared with more flexible models in the context of immune checkpoint inhibitors that have the potential to provide lasting survival benefits. This study used extended follow-up data from the phase III CheckMate 067 trial (NCT01844505) to assess the predictive accuracy of standard parametric models in comparison with more flexible methods for estimating the long-term survival benefit of the immune checkpoint inhibitor combination of nivolumab plus ipilimumab in advanced melanoma. Mixture cure, parametric mixture, and landmark response models provided more accurate estimates of long-term overall survival versus external data than other models tested. In this case study with immune checkpoint inhibitor therapies in advanced melanoma, extrapolation models that explicitly incorporate differences in cancer survival between observed or latent subgroups showed greater accuracy with both early and later data cuts than other approaches did.

Introduction EQ-5D-3L preference-based value sets are predominately based on hypothetical health states and derived in cross-sectional settings. Therefore, we derived an experience-based value set from a prospective observational study. Methods The International Costs and Utilities Related to Osteoporotic fractures Study (ICUROS) was a multinational study on fragility fractures, prospectively collecting EQ-5D-3L and Time trade-off (TTO) within two weeks after fracture (including pre-fracture recall), and at 4, 12, and 18 months thereafter. We derived an EQ-5D-3L value set by regressing the TTO values on the ten impairment levels in the EQ-5D-3L. We explored the potential for response shift and whether preferences for domains vary systematically with prior impairment in that domain. Finally, we compared the value set to 25 other EQ-5D-3L preference-based value sets. Results TTO data were available for 12,954 EQ-5D-3L health states in 4683 patients. All coefficients in the value set had the expected sign, were statistically significant, and increased monotonically with severity of impairment. We found evidence for response shift in mobility, self-care, and usual activities. The value set had good agreement with the only other experience- and preference-based value set, but poor agreement with all hypothetical value sets. Conclusions We present an experience- and preference-based value set with high face validity. The study indicates that response shift may be important to account for when deriving value sets. Furthermore, the study suggests that perspective (experienced versus hypothetical) is more important than country setting or demographics for valuation of EQ-5D-3L health states.

Abstract BACKGROUND Sickle cell disease (SCD) is an autosomal recessive disorder characterized by abnormal hemoglobin. SCD causes hemolytic anemia, vaso-occlusion leading to vaso-occlusive crises (VOC) and contributing to organ damage and early death. SCD is most prevalent in sub-Saharan Africa and the Middle East, but also countries such as Brazil, India and US, have comparatively high frequencies of SCD. Global migration has contributed to a greater geographical spread. The prevalence of SCD in Sweden is unknown. OBJECTIVE The primary objectives of this study were to estimate the 1-year prevalence of SCD and SCD-associated resource use in Sweden. Secondary objectives were to estimate birth incidence, treatment patterns and survival. PATIENTS Patients with an ICD-10 diagnosis code for SCD (any D57 [excluding D57.3, sickle cell trait]) were identified from the Swedish Patient Registry (between January 1 st 2001 and June 30 th 2018). Patients were assessed for 1-year prevalence and resource use per calendar year for a follow-up period of 13 years (2006-2018). METHODS Patients were considered prevalent from birth or immigration to death or emigration. Resource use from specialized care, including all events recorded in the registry with any D57 as the main diagnosis was assessed in the follow up period 2006-2018 as number of outpatient visits and inpatient stays. Costs for this hospital resource use were estimated through remuneration amounts based on diagnosis related groups. Data on sick leave days and days with disability pension due to SCD in patients in working age (18-65 years) were retrieved from the Swedish Social Security Agency and costed with the mean salary in Sweden, plus social security contributions. Costs are reported in 2019 Swedish Krona (SEK, ≈$ 0.1). RESULTS One-year prevalence of all SCD diagnosis increased from 504 patients (5.53 per 100,000 population) in 2006 to 670 patients (6.55 per 100,000 population) in 2018. The 1-year prevalence of SCD patients ever recorded with an ICD-10 code for SCD with VOC (D57.0) increased from 139 patients (1.53 per 100,000 population) in 2006 to 260 patients (2.54 per 100,000 population) in 2018. The proportion of prevalent patients that were born in Sweden decreased over the years, from approximately 55% in the beginning of the study period to 45% in the end of the study period. The mean and median age of the SCD population decreased over the study period. Individuals with SCD and VOC were, on average younger than the other SCD (D57) subgroups. Birth incidence was captured by calendar year 2006-2018 and was highest in 2007 with 15 children born with SCD. For Swedish-born children with SCD during the patient identification time (n=123), the mean time to identification in the registers was 2.6 years (SD 2.7, range 0-16 years). Hospital outpatient visits and inpatient stays with SCD (all events with D57 recorded) as main diagnosis increased from 57 to 189, and 250 to 1,003, respectively, over the years 2006-2018. This corresponded to costs of inpatient care increasing from 1.4 million (M) SEK in 2006 to 7.3 M SEK in 2018 and costs of outpatient visits increasing from 0.9 M SEK in 2006 to 4.6 M SEK in 2018. The vast majority of costs were incurred in individuals ever recorded with a SCD with VOC diagnosis (D57.0). The most frequent hospital treatment was blood transfusion, with 8-11% of patients receiving transfusion in each year studied, especially common in SCD and VOC diagnosis. The prescribed treatment with the highest increase of uptake over the study period were hydroxyurea, vitamins and paracetamol in all SCD. Individuals in working age had on average 2.3 days of sick leave per patient-year due to SCD (D57), and approximately 4% of these patients received disability benefits because of their SCD. During the follow-up period, the median age at death was 74 years for all SCD and 69 years for SCD with crisis, this is 7-10 years and 12-15 years less compared to the Swedish general population respectively. CONCLUSION This study demonstrates that the prevalence, hospital resource use and associated costs have increased substantially in Sweden. In an era of emerging treatments for SCD we have for the first time comprehensively described epidemiological-, disease-related and economical aspects of SCD in Sweden. Disclosures Hernlund: ICON: Current Employment. Ivergård: ICON: Current Employment. Svedbom: ICON: Current Employment. Dibbern: Novartis: Current Employment. Stenling: Novartis: Current Employment. Sjöö: Novartis: Ended employment in the past 24 months. Vertuani: Novartis: Current Employment. Glenthøj: Saniona: Research Funding; Bristol Myers Squibb: Consultancy; Agios: Consultancy; Novo Nordisk: Honoraria; Novartis: Consultancy; Alexion: Research Funding; Sanofi: Research Funding; Bluebird Bio: Consultancy.

Abstract Acute myeloid leukemia (AML) is associated with a high economic and clinical burden. Recently novel therapies have been added to standard treatment regimens. Here, we evaluated the economic impact of AML up until the introduction of these novel therapies. Individual data on 2954 adult patients diagnosed from 2007 to 2015 from five Swedish national population‐based registers were used, enabling analyses from diagnosis to either death or 5‐year follow‐up for survival, inpatient and outpatient costs, costs of prescribed drugs, sick leave, and early retirement. Costs per patient were stratified by age group, treatment options, and FLT3‐ITD status. The expected 5‐year costs per patient differed substantially between age groups. Patients aged 18–59 years had an expected mean cost per patient of €170,748, while age groups 60–69 years, 70–79 years, and >80 years incurred an expected mean cost of €92,252, €48,344, and €24,118, respectively, over 5 years. Patients <60 years undergoing stem cell transplantation had the highest costs (€228,525 over 5 years). About 60% of costs for these patients were from hospitalizations and 20% from sick leave and early retirement; cost per day was highest from the first admission to complete remission. This study provides a baseline for socioeconomic evaluations of novel therapies in AML in Sweden.

Introduction: AML affects all ages with an incidence rate of 5 per 100,000, but is much more frequent in older population. The overall lifetime risk of AML is estimated to be 0.5-1%. Long-term overall survival in younger (age < 60 years) is about 50%, but much worse among older population. Although AML therapy is one of the most resource-intensive cancer treatments, there are few estimates of the resource use and economic burden by treatment phase. Methods: This study was a retrospective database study performed on Swedish national data. Adult patients (age ≥18 years) diagnosed with AML in Sweden between 2007 and 2015 were identified in the Swedish Cancer Registry, along with vital status. Data on resource use were collected from national registers for inpatient- and outpatient specialized care and prescribed drugs. Information on diagnostics and treatment was accessed from the Swedish national AML Registry (SwAMLR). Data on sick leave (SL) and early retirement (ER) came from the Swedish Social Insurance Agency (absent days costed with the mean salary in Sweden). Hospital care resource use was costed using diagnosis-related group (DRG) remunerations, and include cost of inpatient drugs. The mean cost from the defined start of the treatment phase until the end of the treatment phase was divided by the mean number of days for the corresponding treatment phase to estimate the mean cost per day. The defined treatment phases were restricted to a maximum of 5 years. All costs are represented in US$. Results: Of the 2,954 patients identified in the Swedish Cancer Registry, 1,772 patients with a median age of 64 years were identified in the SwAMLR as fit for receiving high-dose chemotherapy . Of these, 1,243 were recorded with both curative intent of treatment and dates for achieving complete remission. Mean costs from the first AML-related hospital admission until the date of complete remission amount to $27,244. The mean number of days for the corresponding period were 45.16, resulting in a mean cost per day of $603 from first admission to first complete remission. The corresponding cost per day for patients recorded with curative intent but no complete remission (n=428) are $494. Time was counted from first AML-related admission until 90 days after first admission, or SCT or death, whichever occurred first. Costs after complete remission to either relapse, SCT, death or re-induction (n=1,237) amount to $50,793 for a mean of 438.63 days ($116/day). This treatment phase includes long-term survivors, whereas the costs from SCT, relapse or re-induction are not included. From relapse to death, the total cost is almost twofold for patients with re-induction (n=350) compared to palliative treatment (n=254). Cost per day amount to $179 for patients with palliative treatment and $256 for patients with re-induction treatment, respectively. The cost per day from date of SCT to death (n=511) is estimated to $192, incurred over a long period of time (mean number of days 844.02). The age of transplant recipients ranged between 18-71 years, with a median of 52 years. Conclusions: Costs of AML up to remission are feasible to estimate through DRG-costing methods, and studies have shown these costs are intense. Indeed this study shows that the highest cost per day is observed from first admission to complete remission. In addition results from our study show that there are high costs incurred also in the long-term, i.e. after remission. Of the included treatment phases the total cost from date of SCT to death is the largest, amounting to over $160,000. Approximately 20% are due to SL/ER, which is the second largest cost component after inpatient costs accounting for 60% of the total costs. Table. Disclosures Hernlund: ICON: Employment. Redig:ICON: Employment. Paulsson:Novartis: Employment. Vertuani:Novartis: Employment.

The International Costs and Utilities Related to Osteoporotic fractures Study is a multinational observational study set up to describe the costs and quality of life (QoL) consequences of fragility fracture. This paper aims to estimate and compare QoL after hip, vertebral, and distal forearm fracture using time-trade-off (TTO), the EuroQol (EQ) Visual Analogue Scale (EQ-VAS), and the EQ-5D-3L valued using the hypothetical UK value set.

To compare health utility estimated with the EQ-5D and the Time-trade-off (TTO) instruments after vertebral fracture. The International Costs and Utilities Related to Osteoporotic fractures Study (ICUROS) is a prospective multinational study with the aim of estimating costs and health related quality of life (HRQoL) related to osteoporotic fractures. In the study, two instruments were used to estimate patients' perceived health utility: the direct individual preference based TTO instrument and the indirect EQ-5D instrument, valued by societal preferences. Health utility was measured before fracture (recollection), and 2 weeks, 4 months and 12 months after fracture. In the 225 patients included in the interim analysis, estimates of health utility elicited from the two instruments varied significantly and the EQ-5D systematically provided lower estimates compared to the direct TTO instrument. The smallest mean outcome difference was 0.10 (TTO: 0.91 and EQ-5D: 0.81), elicited before fracture and the largest difference was 0.20 (TTO: 0.44 and EQ-5D: 0.23), elicited after two weeks. The correlation coefficient for the health utility loss over 12 months, using Spearman's correlation, for the two instruments was 0.36 (p<0.05). Osteoporotic vertebral fractures give rise to significant loss of health utility, irrespective of instrument used. However, there are substantial differences in the extent of the health utility decrease assessed by the two instruments. The main explanation for these differences is that the instruments rely on different reference populations; respondents of direct TTO act as their own reference and express preferences regarding their own health state, whereas the EQ-5D reflects the preferences of healthy individuals assessing hypothetical health states.