|Title:||Design of a prospective payment system for inpatient psychiatric care in Switzerland|
|Authors :||Riguzzi, Marco|
|Conference details:||12. Workshop on Costs and Assessment in Psychiatry, Venice, March 27-29, 2015|
|License (according to publishing contract) :||Licence according to publishing contract|
|Type of review:||No review|
|Subjects :||Tariff structure; Psychiatric care; Prospective payment system|
|Subject (DDC) :||338: Production |
362: Health and social services
|Abstract:||Background: Inpatient psychiatric care in Switzerland is currently reimbursed by uniform per-diem rates. The new Health Insurance Act mandates the introduction of a national tariff system that reflects differences in patient resource use and introduces per-case reimbursements. In this study we developed a prospective payment system (PPS) with a patient classification system and a mixed tariff structure including per-case payments and per-diem rates. Data: Primary data on patient characteristics, cost per episode and daily time spent by medical staff were collected in a survey of 9’888 patients from 17 inpatient psychiatric facilities in 2013, two of which were facilities specialized in disorders related to drug use. Patient characteristics include the main psychiatric diagnosis (ICD-10) and a weekly assessment of the severity of illness with the Health of Nations Outcome Scales questionnaire (HoNOS) covering dimensions such as aggressive behavior and the independence in activities of daily living. Methods: Predictors of total cost per episode were examined with a multivariate linear regression model. Explanatory variables included main psychiatric diagnosis, HoNOS scores, socio-demographic characteristics, type of insurance plan, compulsory hospitalization and hospital fixed effects. Different mixed tariff structures with (staggered) per-case payments and per-diem rates were evaluated. Per-case payments of varying weight were evaluated and per-diem rates were allowed to vary over length of stay. Goodness-of-fit was measured by an out-of-sample evaluation of the mean absolute percentage error (MAPE) between hypothetical revenues and observed costs per episode. Results: Regression analysis leads to a patient classification system with 48 psychiatric cost groups (PCGs) based on psychiatric diagnoses and the HoNOS assessment of severity of illness. The resulting tariff system yields high per-diem payments for the first 7 days of stay, a per-case payment at day 8 and a reduced per-diem rate thereafter. The reduced rate is not cost-covering, constituting an incentive to reduce length of stay, while the combination with per-case payment satisfies cost-neutrality of the tariff system as a whole. As some clinics are concerned with special cases of long-term patients, a cost-covering per-diem tariff is introduced after 60 days of stay (90/120 days are alternative scenarios). Compared to the uniform per-diem tariff presently in operation in Switzerland, a per-diem system with 48 PCGs improves the accuracy of the tariff system in predicting actual costs per case. This means the MAPE is lowered from 21.3% to 19.0% in a system where per-case payment is calibrated to zero. The introduction of a per-case payment deteriorates the accuracy of the system. Discussion: The proposed PPS combines incentives for an appropriate provision of care with incentives for a reduction of length of stay. The latter type of incentive however is linked to reduced accuracy of the tariff system. Hence, for policy-makers there clearly is a trade-off between accuracy and setting incentives. A system with per-case payments might also lead to an uneven distribution of revenues across clinics (in particular, if average length of stay varies across clinics).|
|Departement:||School of Engineering|
|Organisational Unit:||Institute of Data Analysis and Process Design (IDP)|
|Publication type:||Conference Other|
|Appears in Collections:||Publikationen School of Engineering|
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.