Economic Burden of Dementia Related Psychosis Among Medicare Beneficiaries: A State-Transition Markov Analysis of Total Annual Direct Costs

Study ObjectiveDementia related psychosis (DRP), characterized by debilitating symptoms such as hallucinations and delusions, is estimated to affect 2.4M people with dementia in the US. Patients with DRP may have twice the rate of dementia progression compared to patients with no DRP. Given that dementia disproportionally impacts the elderly, a comprehensive cost of-illness analysis may add to the current understanding of the overall economic burden of DRP prevalence. The objective of this study was to estimate the cost of DRP from a Centers for Medicare and Medicaid Services (CMS) perspective.MethodsA five state-transition Markov analysis, adapted from Green et al, was conducted to assess the annual direct DRP cost burden to CMS. Patients entering the model were allowed to transition between three at-home health-states (mild dementia plus psychosis, moderate dementia plus psychosis, severe dementia plus psychosis), one long-term care/nursing home (LTC/NH) stay, or death (absorbent health state) at any given time. Since the model accommodates tunnel health-states based on dementia, psychosis and functioning severity levels, patients stayed in the same health-state or transitioned to a more severe health-state or progressed to death (i.e., absorbent health-state) between each 30-day cycle. Prevalence, disease-severity, and state-transition probability estimates were derived from literature while direct costs of DRP were derived from a Medicare claims analysis. Costs were discounted at 3%. Model robustness was tested to check if results were sensitive to changes in inputs and assumptions.ResultsOf the estimated 61.5M Medicare beneficiaries, about 6.87M may suffer from dementia. In the base-case scenario, an estimated total of 2.2M prevalent DRP patients enter the model based on dementia severity into one of the 4 non-absorbent health-states: three at-home (10% mild, 60% moderate, and 10% severe) and one LTC-NH (10%). Total Medicare annual direct DRP costs are estimated to be approximately $119.98B ($113.96B-$125.96B) and about $54K PPPY (Per-Patient-Per-Year) costs (2019 USD). NH costs and patient volume at higher severity levels are significant cost drivers. Sensitivity analysis results show that the model is sensitive to disease severity and disease progression.ConclusionsThese results suggest that DRP imposes a significant direct cost burden despite its low prevalence. In this analysis, per-patient per year (PPPY) cost of DRP prevalence was estimated to be slightly higher than incident PPPY DRP costs. These differences may be attributed to the number of patients at higher severity levels and the time spent in a severe health state as well as cost of LTC/NH stays. Especially given the ageing population in the US, DRP could become an increasing public health concern. There is a significant need for education and awareness about DRP cost burden.FundingAcadia Pharmaceuticals Inc.

Age dependency of EQ-5D-Youth health states valuations on a visual analogue scale

Background Examine whether the use of different ages has an impact on the valuation of EQ-5D-Y health states for a hypothetical child or adolescent. Methods A survey was administered during regular classes among a convenience sample of university students in the Netherlands. Respondents first valued 6 EQ-5D-Y health states (2 mild, 2 moderate, 2 severe) describing a hypothetical child/adolescent of a certain age on a visual analogue scale (VAS). After 1 h respondents valued the same six health states again but this time the age of the child was different. Age differed between 4, 10 and 16 year old. Results Number of respondents was 311. No significant differences in valuation of the six health states were found between the age of 10 and 16. One moderate health state was valued significantly better for a 4-year old compared to a 10 and a 16 year old. The same applied for one severe health state that was valued higher for a 4-year old compared to a 16-year old. Conclusion Our study shows that, except for one moderate and one severe health state, other EQ-5D-Y health states were not valued significantly different when description of age differed. It is possible that problems in specific health domains are considered more severe for older children/adolescents compared to younger children who might still be dependent on their caregivers. Future research should examine whether our findings are also present in a broader set of EQ-5D-Y health states, with a choice-based method like TTO or DCE, and a more heterogeneous sample.

Crosswalk EQ-5D-5L Value Set for Slovenia

IntroductionDue to the availability of the EQ-5D-5L instrument official translation into Slovenian its use is widespread in Slovenia. However, the health profiles obtained in many studies cannot be ascribed their appropriate values as the EQ-5D-5L value set does not yet exist in Slovenia. Our aim was to estimate an interim EQ-5D-5L value set for Slovenia using the crosswalk methodology developed by the EuroQol Group on the basis of the EQ-5D-3L Slovenian TTO value set. Our secondary aim was to compare the interim values obtained with the EQ-5D-3L Slovenian values.MethodsTo obtain a Slovenian interim EQ-5D-5L value set, we applied the crosswalk methodology developed by the EuroQol Group to the Slovenian EQ-5D-3L TTO value set. We examined the differences between values by comparing the mean 3L and 5L value scores and the distribution of values across all respondents.ResultsBy definition, 3-level and 5-level versions have the same range (from 1 to −0.495) and a health state coded 22222 in the 3-level version corresponds to 33333 in the 5-level version. While the addition of a “slight” severity level (22222) in the 5-level version has a low informational value, the addition of a “severe” health state (44444) covers larger range of the scale. The 5-level version results in fewer health states being valued below 0 and above 0.8.ConclusionThe EQ-5D-5L value set, based on the crosswalk methodology, should be used until a value set for the EQ-5D-5L is derived from preferences elicited directly from a representative sample of the Slovenian general population.