Amélie Prigent1,2,3,4, Blaise Kamendje-Tchokobou5,6, Karine Chevreul7,8,9,10,11. 1. AP-HP, URC Eco Ile-de-France, 1, place du parvis Notre Dame, 75004, Paris, France. amelie.prigent@urc-eco.fr. 2. University Paris Diderot, Sorbonne Paris Cité, ECEVE, UMR 1123, 75010, Paris, France. amelie.prigent@urc-eco.fr. 3. Inserm, ECEVE, U1123, 75010, Paris, France. amelie.prigent@urc-eco.fr. 4. Fondation FondaMental, French National Science Foundation, Créteil, 94000, France. amelie.prigent@urc-eco.fr. 5. Centre Hospitalier Départemental Georges Daumézon, Fleury-Les-Aubrais, 45400, France. 6. Agence Régionale de Santé Centre-Val de Loire, Cité administrative Coligny, 131 Rue du Faubourg Bannier, 45000, Orléans, France. 7. AP-HP, URC Eco Ile-de-France, 1, place du parvis Notre Dame, 75004, Paris, France. 8. University Paris Diderot, Sorbonne Paris Cité, ECEVE, UMR 1123, 75010, Paris, France. 9. Inserm, ECEVE, U1123, 75010, Paris, France. 10. Fondation FondaMental, French National Science Foundation, Créteil, 94000, France. 11. Department of Public Health, AP-HP, Robert-Debré University Hospital, 75019, Paris, France.
Abstract
PURPOSE: Health-related quality of life (HRQoL) is a widely used concept in the assessment of health care. Some generic HRQoL instruments, based on specific algorithms, can generate utility scores which reflect the preferences of the general population for the different health states described by the instrument. This study aimed to investigate the relationships between utility scores and potentially associated factors in patients with mental disorders followed in inpatient and/or outpatient care settings using two statistical methods. METHODS: Patients were recruited in four psychiatric sectors in France. Patient responses to the SF-36 generic HRQoL instrument were used to calculate SF-6D utility scores. The relationships between utility scores and patient socio-demographic, clinical characteristics, and mental health care utilization, considered as potentially associated factors, were studied using OLS and quantile regressions. RESULTS: One hundred and seventy six patients were included. Women, severely ill patients and those hospitalized full-time tended to report lower utility scores, whereas psychotic disorders (as opposed to mood disorders) and part-time care were associated with higher scores. The quantile regression highlighted that the size of the associations between the utility scores and some patient characteristics varied along with the utility score distribution, and provided more accurate estimated values than OLS regression. CONCLUSIONS: The quantile regression may constitute a relevant complement for the analysis of factors associated with utility scores. For policy decision-making, the association of full-time hospitalization with lower utility scores while part-time care was associated with higher scores supports the further development of alternatives to full-time hospitalizations.
PURPOSE: Health-related quality of life (HRQoL) is a widely used concept in the assessment of health care. Some generic HRQoL instruments, based on specific algorithms, can generate utility scores which reflect the preferences of the general population for the different health states described by the instrument. This study aimed to investigate the relationships between utility scores and potentially associated factors in patients with mental disorders followed in inpatient and/or outpatient care settings using two statistical methods. METHODS:Patients were recruited in four psychiatric sectors in France. Patient responses to the SF-36 generic HRQoL instrument were used to calculate SF-6D utility scores. The relationships between utility scores and patient socio-demographic, clinical characteristics, and mental health care utilization, considered as potentially associated factors, were studied using OLS and quantile regressions. RESULTS: One hundred and seventy six patients were included. Women, severely ill patients and those hospitalized full-time tended to report lower utility scores, whereas psychotic disorders (as opposed to mood disorders) and part-time care were associated with higher scores. The quantile regression highlighted that the size of the associations between the utility scores and some patient characteristics varied along with the utility score distribution, and provided more accurate estimated values than OLS regression. CONCLUSIONS: The quantile regression may constitute a relevant complement for the analysis of factors associated with utility scores. For policy decision-making, the association of full-time hospitalization with lower utility scores while part-time care was associated with higher scores supports the further development of alternatives to full-time hospitalizations.
Entities:
Keywords:
Mental disorders; Mental health care; Quantile regression; Utility scores
Authors: A V Diez-Roux; C I Kiefe; D R Jacobs; M Haan; S A Jackson; F J Nieto; C C Paton; R Schulz; A V Roux Journal: Ann Epidemiol Date: 2001-08 Impact factor: 3.797