Leader: Simone Ghislandi (BOCCONI); Other collaborator(s): Giovanni Fattore (BOCCONI), Aleksandra Torbica (BOCCONI), Irene Torrni (BOCCONI)
Task 4 will analyze the needs of the elderly population and, based on data-platforms constructed in Task 3, will focus on the analysis of data through biostatistics and econometric models to investigate whether, and how, community care meets the need of the elderly population. The ultimate aim of this task is to analyse the impact of community care (primary and social care) on consumption of hospital services (admissions, length of stay, costs). The data platform constructed in Task 3 and the subsequent analyses in Task 4 aim to provide a unique template for future evaluations of the impact of the component 1 of the Mission 6 of the PNRR (Proximity networks, facilities and telemedicine for community health care).
Part of the activities of Task 4 will begin once Task 3 activities are completed. While waiting for the data, the working group of Task 4 has begun work on the effect of household shocks on health outcomes and utilization of health services by the elderly. The data used come from the survey Health and Retirement Study, which includes information on US individuals over 50, and the Survey of Health and Retirement, which interviews EU individuals over 50.
Brief description of the activities and of the intermediate results
The completion of one of the datasets related to Task 3 made it possible to begin analyses of the effect of social spending on outcomes related to avoidable hospitalizations of the elderly. Meanwhile, preliminary results of the analysis on the effect of family shocks on health outcomes and use of health services by the elderly were presented at the European Health Economics Association (EUHEA) conference in Vienna (20 June-3 July 2024). The contribution submitted to the conference is reported below:
Main policy, industrial and scientific implications
The findings reported above allow identifying personality traits as risk or protective factors for the development of loneliness and, in turn, poor health levels. This indicates that a personalized approach is crucial to address each individual's loneliness and reduce its adverse impact on health.
Brief description of the activities and of the intermediate results
Preliminary results of the analysis on the effect of social spending on outcomes related to avoidable hospitalizations of the elderly were presented at the Italian Health Economics Association (AIES) conference in Naples (9-10 December 2024). The contribution submitted to the conference is reported below.
Background. There is considerable evidence that factors indicating social support are associated with an increased use of health care services. This paper investigates the effect of being unmarried, often used as an index for social support, on hospital length of stay and costs. We contribute to the existing literature using nationwide administrative data and focusing on municipal social spending (MSS) as a boost force for the effects of interest.
Methods. We exploit an administrative panel of Italian hospital discharge records between 2010 and 2019 for the entire over-45 national population. We enrich our panel with information related to the LOS threshold value, the cost associated to each DRG, and the social spending resulting from the yearly balance sheets of the patient’s municipality. We estimate the effect of being unmarried on length of stay (LOS) and costs separately for acute activity, long-term activity, and rehabilitation activity. We then exploit the LOS threshold value defined for each Diagnostic Related Group (DRG) to study the effects on below-the-threshold LOS (BTT LOS) and above-the-threshold LOS (ATT LOS) separately, and finally evaluate the variation in the coefficients of interest when MSS are also controlled for. Further investigations are finally carried out by year and macro-area of residence to assess time trends and spatial patterns, respectively.
Results. Preliminary findings show that the effects of being unmarried on our outcomes of interest are positive and statistically significant. When MSS is also included in the regression, the magnitude of the coefficient decreases for all outcomes, indicating that the effect analyzed is enhanced by the lack of social services in the municipality of residence and that, for unmarried individuals, hospital care displaces social assistance.
Main policy, industrial and scientific implications
Regarding the AIES conference contribution, results show that, from a policy perspective, government increases in social spending are necessary to provide appropriate social assistance to lonely individuals, especially to those that are most at risk e.g., males, individuals affected by respiratory diseases and individuals living in the South. This is crucial to reduce LOS and hospital cost.
Building on Task 3, the working group has initiated an analysis on the effect of general medicine on individual health status and healthcare use based on administrative data (Lombardy Region).
A biostatistician (Giulia Tarantino) has been hired to focus on the dataset of the Lombardy Region, which is finally available in the required format. The effort required a new hiring because we encountered some difficulties (slow with the access to the data. Starting from January 2025, the access is complete and the first descriptive results are expected by the end of the first quarter of 2025.
The paper on social spending have been presented and hare scheduled to be presented in different occasions:
Finally, the loneliness paper has been completed and we will proceed with submission by the end of the first trimester 2025.
The project examines how general practitioners (GPs) influence elderly care, particularly in relation to avoidable hospitalizations. The structuring of the dataset and the initial analyses will help us better understand how primary care operates in the community and assess the role of GPs in elderly health outcomes. This information will be valuable both for scientific research and to support policy decisions aimed at improving healthcare delivery and reducing unnecessary hospital admissions.
We are currently developing the descriptive statistics and expect to have preliminary results available shortly.
Regarding the loneliness paper, after an initial informal review, the research team decided to change the research question.
Currently, the paper examines the dynamic effects of living alone on objective physical and mental health outcomes, with a particular focus on the mediating role of loneliness. Using a Two-Way Fixed Effects Difference- in-Differences design, we exploit the event of transition to living alone by analyzing data from the HRS and SHARE surveys, covering the over-50 population in 20 EU countries and the US. Health outcomes include the likelihood of having psychiatric diseases, the number of co-morbidities, the probability of being hospitalized as a proxy for acute shocks, and mortality. The findings reveal that living alone has generally adverse and in some cases persistent effects on the outcomes, with the impact found to be particularly severe for those who experience post-event increases in loneliness. Such adverse effects are more pronounced among widows, men, older adults, low-income individuals, and those residing in the US and Western EU countries, where living alone is more common.
The study underscores the importance of considering living alone and loneliness as global public health priorities and points out the need for policies to promote health and reduce health inequalities among those living alone, especially for the most vulnerable groups.
The paper is almost complete for submission.
Regarding the analysis on the effect of municipality social spending, the research group continued activities along this line by starting a new paper to study the effect of social spending on hospital-related outcomes of the elderly. Preliminary results were presented at the BReCHS Seminar in Milan-Bicocca University. Here the abstract.
Recently published statistics show that 50% of patients over 70 are discharged from the hospital 7 days later, due to lack of social support after hospitalization. Despite this alarming evidence, social care remains largely underfunded and the interdependency between the social and hospital systems unquantified.
In this study, we merge admission- and municipality-level national data and use a two-stage econometric approach to estimate the effect of municipal social spending on hospital-related outcomes for elderly patients. The hypothesis we test is that hospital care is a substitute for poor social services.
The main findings show that a higher level of social spending reduces the number of hospitalizations, as well as the length of stay and cost per admission. These results call for urgent policy interventions to increase post-discharge social support for the elderly and increase coordination among different areas of care.