Patient Care Pathways Between Ambulatory and Hospital Settings in Mental Health Care

In France, mental health care is provided by public and private hospitals and is organized around three main categories of care: ambulatory, full-time and part-time, each presenting several types of care (e.g. for full time care: hospitalization or treatment in a therapeutic apartment). This wide variety of treatment methods is associated with significant variations in the use of each of them, which can lead patients with similar mental disorders to follow different care pathways. This suggests that one of them may have a better quality of care than the others and therefore that the variation in care pathways is associated with a loss of equity and decreased chances for some patients, as well as a non-optimal use of resources and inefficiency for the health care system. Several categories of factors are associated with the variation in care pathways: the clinical and socioeconomic characteristics of patients, the characteristics of the mental health care providers and those of the health and health and social care available which can provide care complementary to that of psychiatry, as well as the characteristics of the environment in terms of living conditions (urban development and unemployment rate for example).

While it is not possible to act in the short term on the characteristics of patients or of their living conditions, it seems essential to identify the characteristics of mental health care providers and of the surrounding care on offer associated with the variation in care pathways, as they may be subject to reforms to improve the quality of care.

The primary objective of the study is to identify the organizational factors and the characteristics of mental health care providers and of the surrounding care provision associated with the variability in care pathways for mental health care and in their quality.

The secondary objectives are:

• to define a method to elaborate patients' care pathways;
• to assess the respective quality of each possible pathway;
• to build a typology of care pathways with similar characteristics;
• to describe the variability in care pathways and in their associated quality;
• to identify other factors associated with the variability in care pathways and in their quality (characteristics of the patients and their environment).

The primary endpoint is the organizational factors associated with the different typology of care pathways and with the different levels of quality of care (in a multivariate modelling allowing an adjustment on the other factors).

Secondary endpoints will include:

• Identification of individualized care pathways
• Level of quality of each care pathways in ordered classes
• Typology of similar care pathways, and for each category: proportion of the different care pathways within this category, proportion in each quality level, intensity, frequency and diversity of care
• Clinical and socioeconomic characteristics of patients and of their environment significantly associated with the different typology of care pathways and with the different levels of quality of care (in a multivariate modelling allowing an adjustment to the other factors).

In order to do so, a cohort of patients will be constituted using French administrative databases. Mental health care consumption will be obtained via data from the ambulatory claims database (SNIIRAM) along with hospital discharge databases for psychiatric care (RIM-P) and non-psychiatric acute care (PMSI-MCO). These databases will be supplemented by other linkable databases in order to describe the clinical and socioeconomic characteristics of the patients and environment, as well as the characteristics of care provision and available health and social care.

All adult patients with a full-time or part-time admission to one of seven public psychiatric hospitals participating to the study or admitted to a private psychiatric hospital located on their health territories will be included in the study. Patients not residing in the health territory of those hospitals included in the study as well as patients treated exclusively in ambulatory care settings will be excluded.

Care pathways will first be constructed with an analysis of state sequences. They will then be given a quality level: a group of experts will be appointed to define specific criteria to assess quality. The typology of care pathways will then be constructed via an ascending hierarchical classification based on a distance calculation taking into account the temporal ordering of care and descriptive analysis of the groups obtained. Finally, we will identify factors associated with the variability in care pathways and in their quality using multinomial logistic regressions.

The identification of factors associated with the variability in care pathways and in their quality, in particular factors linked to the organization of mental health care and to the complementary health and health and social care provision, will lead to recommendations on how to improve the quality of care and the efficiency of the health care system. Ultimately, this will contribute to the reduction of geographic and social health inequalities.