Leader: Paolo Berta (UNIMIB); Other collaborator(s):
The elderly population is generally affected by a mix of chronic diseases. The impact on the healthcare system is increased hospitalizations and reduced hospital efficiency. Moreover, the hospitalization of elderly patients often ends in a motorial or cognitive functional loss (or both), and effective discharge planning is needed to ensure appropriate continuity of cares after the hospitalization. Based on administrative data and robust econometrical approaches, the aim of this task is to analyse the continuity of cares from the hospitalization to appropriate social service. An effective hospital discharge planning would decrease hospital length of stay and improve the patients’ outcomes.
Brief description of the activities and of the intermediate results
The development of the task is proceeding according to what was planned when the aims of the project were defined. One of the most urgent issues is the access to the administrative data from Regione Lombardia, which involves the stipulation of an agreement between Bicocca University and Regione Lombardia. Privacy issues and the system implemented by Regione Lombardia are taking longer than expected, but activities are underway, and the goal is to close this step by July. This aspect has not affected the development of another research work related to the project aims, which concerns the study of hospital discharges and the relationship between health outcomes and patient residence in an urban or rural area. This study is under development and has been accepted for presentation at the next Age-it general meeting to be held in May in Venice.
This research aims at investigating how health outcomes change based on patient residence with a specific focus on hip and knee replacement. Equal access to quality healthcare services such as knee and hip replacement across both rural and urban areas is a pillar of equitable healthcare systems. Our study analyzes gaps in access and outcomes for these surgeries between geographical settings in Italy from 2011-2016. Using Italian administrative data, knee and hip arthroplasties were identified through DRG codes. Patients were stratified by rural or urban residence. The main outcome of interest is the readmission rate. Of 807,102 knee and hip replacements, 20.57% of patients were urban residents while 79.43% were rural. Preliminary results show that compared to the patients coming from cities, those living in rural area were more frequently discharged to rehab facilities (20.29% vs 15.85%) and had lower post-surgery readmission. Next step is to identify a causal effect of the rural/urban area on the health outcomes adopting an instrumental variable approach.
Main policy, industrial and scientific implications
From a policy perspective, it is essential to understand whether location in urban or rural areas results in a difference in outcomes. Indeed, this result has a direct impact in terms of disparities in the healthcare system and the goal of reducing disparities between citizens is an essential goal for any health care system.
Preliminary results already point to this difference, and the implication is the need for a higher relationship between hospital and community health services for post-discharge patient monitoring. In addition, further healthcare infrastructure strengthening is required to support the rehabilitation process across the entire care cycle for rural populations. Policy initiatives promoting integrated care delivery for rural communities could mitigate these inequities.
Brief description of the activities and of the intermediate results
The main development in this period is the definition of the rural versus urban characteristic of patients based on a classification system established by the National Strategy for Inner Areas (Strategia Nazionale per le Aree Interne, SNAI) in Italy. This classification is designed to identify regions with limited access to essential services, particularly in rural and remote areas.
Key Definitions:
These are municipalities or clusters of municipalities that provide a comprehensive range of key services, including: Secondary education (a full range of high schools), At least one Grade 1 emergency care hospital (Dipartimento di Emergenza e Accettazione, DEA), At least one Silver category railway station.
Areas that include these key services are classified as urban.
All municipalities that do not qualify as Service Provision Centres are classified as Inner Areas, or rural areas. These are further divided based on their distance from the nearest Service Provision Centre. The classification is based on travel time in minutes to access essential services from the nearest urban center: Outlying areas: Closer to the urban centers but still beyond the main service hubs. Intermediate areas: Farther away but still accessible. Peripheral areas: Even more remote, with limited access to urban services. Ultra-peripheral areas: The most isolated regions, farthest from essential services.
Urban: Patients living in Service Provision Centres (municipalities with comprehensive access to essential services like hospitals, schools, and transportation).
Rural: Patients living in municipalities classified as Inner Areas, which are further categorized by their distance from these Service Provision Centres.
This classification allowed us to compare healthcare access, quality, and outcomes for elderly patients undergoing hip and knee surgeries between rural and urban areas of Italy.
Main policy, industrial and scientific implications
Defining rural versus urban patients using the National Strategy for Inner Areas (SNAI) framework, based on access to essential services, has several important implications.
Classifying areas based on access to key services allows us to better identify regions that are underserved and require targeted interventions. This classification system highlights not just broad rural-urban differences but also nuances within rural areas (e.g., peripheral vs. ultra-peripheral). As a result, policies can be more tailored, focusing on improving access to healthcare in the most remote and underserved areas rather than applying blanket policies across all rural regions.
In this way, the classification highlights the healthcare access gap in remote rural areas. Policymakers can use this evidence to advocate for more investment in rural healthcare facilities, especially emergency care and rehabilitation services, which are often concentrated in urban centers. Policies could aim to decentralize specialized healthcare services and provide incentives for healthcare professionals to work in rural areas.
Since the classification is based on access to essential services, it underscores the importance of geographic equity in healthcare access. This system offers a clear basis for developing policies that ensure more equitable healthcare delivery across Italy, aiming to reduce disparities between urban and rural patients. By identifying gaps in healthcare infrastructure, policymakers can design interventions to ensure that rural residents receive care comparable to that of urban residents.
By using a service-based classification rather than a simple population density or administrative boundary approach, the study provides a more nuanced and precise measure of what constitutes rural and urban areas. This allows for a clearer understanding of how access to essential services, rather than just geographic location, impacts health outcomes. It improves the ability to isolate the true effect of healthcare access on patient outcomes, controlling variations in service availability.
It also acknowledges the heterogeneity within rural areas, which is important for research. Not all rural areas face the same level of disadvantage; some might be closer to service hubs, while others are far more isolated. This classification system allows researchers to capture these differences more accurately.
In addition, by classifying rural and urban areas based on access to services, the authors can control for factors that might otherwise confound their analysis. The ability to account for differences in service availability (rather than just rural vs. urban labels) helps in producing more accurate, reliable findings about the effect of residence on healthcare outcomes. For example, differences in readmission rates or hospital stays may be due to lack of nearby services rather than inherent differences between rural and urban populations.
This approach provides a framework that can be replicated in other regions or countries. By defining rural and urban areas in terms of access to services rather than just population density, this study contributes to a scientific methodology that could be applied in studies of other regions facing similar geographic healthcare challenges. It allows for cross-comparisons between different geographic areas and contributes to a more standardized way of examining rural-urban health disparities.
Brief description of the activities and of the intermediate results
In this part of research, we focused on the causal impact of living in a rural vs urban are, adopting an Instrumental Variable (IV) approach, a statistical technique used to address the issue of endogeneity in regression models. Endogeneity arises when an explanatory variable is correlated with the error term, leading to biased and inconsistent estimates in standard regression models. This can happen due to omitted variables, measurement errors, or reverse causality. The IV approach helps in estimating the causal effect of a variable by eliminating this bias.
In studies like this, the primary variable of interest is whether a patient lives in a rural or urban area, and how this affects their health outcomes (like length of hospital stay or readmission rates). However, where people live might not be random. For example, people with better health might choose to live in urban areas, or urban areas might have better healthcare facilities, which could affect health outcomes independently of the area itself. This creates a problem of endogeneity, where urban/rural residence is correlated with unobserved factors (like personal health or socioeconomic status) that also influence the health outcomes.
An IV is a variable that is correlated with the endogenous explanatory variable (in this case, rural vs. urban residence), but uncorrelated with the error term in the outcome equation (i.e., it does not directly affect the outcome, except through its effect on the endogenous variable). The IV helps isolate the variation in the explanatory variable that is exogenous, meaning not affected by unobserved factors that also affect the outcome.
In this research we use mean property value in each municipality as the instrumental variable for rural or urban residence. The rationale is that property values are likely to be strongly correlated with whether an area is rural or urban, but they do not directly affect the health outcomes of individuals undergoing surgery, except through their influence on the type of area in which a person lives.
To deal with the IV strategy we use a classical Two-Stage Least Squares (2SLS) approach. In the first stage the endogenous variable (urban/rural residence) is regressed on the instrumental variable (mean property value) and other exogenous control variables. This step essentially isolates the variation in the endogenous variable that is due to the instrument (i.e., the variation in urban/rural residence that is driven by property values, not by health-related factors).
In the second stage the predicted values from the first stage (the part of urban/rural residence explained by the instrument) are then used in place of the original variable in the outcome equation. This ensures that the effect of urban/rural residence on health outcomes is estimated free from the bias caused by endogeneity.
Without the IV, the estimated effect of rural vs. urban residence on health outcomes could be biased due to unobserved factors (like socioeconomic status or health behaviors) that affect both where someone lives and their health outcomes. The IV approach helps to overcome this bias by focusing on the variation in residence that is unrelated to these unobserved factors, providing a more reliable estimate of the true causal effect of residence on health outcomes.
Main policy, industrial and scientific implications
The findings indicate that while rural patients have higher comorbidity rates and less access to specialized care, the comparable outcomes suggest that the Italian healthcare system has managed to mitigate some of the challenges associated with rural healthcare. However, disparities in access to specialized hospitals remain, indicating the need for targeted investments in healthcare infrastructure in rural areas to further reduce gaps in specialized care access.
Despite these access disparities, the similarity in outcomes such as hospital stay and readmission rates between rural and urban patients highlights the effectiveness of Italy’s healthcare system in managing postoperative care. Policymakers can learn from both rural and urban healthcare models to improve efficiency across the board, particularly by focusing on strategies that promote high-quality care in resource-limited settings.
As Italy faces an aging population, with significant portions of the elderly residing in rural areas, the paper emphasizes the need for context-specific policies to address the unique healthcare challenges faced by rural elderly populations. Policies that focus on enhancing both the availability and quality of healthcare services in remote regions could improve health equity across the country.
The research suggests that broader socioeconomic and environmental factors, including regional healthcare policies and infrastructure, play a crucial role in shaping health outcomes. This highlights the importance of addressing not only healthcare access but also underlying socioeconomic inequalities that may exacerbate health disparities.
In summary, the study underscores the importance of healthcare policies that ensure equitable access to specialized care, particularly for rural populations, while recognizing the success of existing healthcare delivery models in managing post-acute care for elderly patients.
Brief description of the activities and of the intermediate results
In this research, we examine the causal impact of living in rural vs. urban areas using an Instrumental Variable (IV) approach to address endogeneity in regression models. Endogeneity arises when an explanatory variable correlates with the error term, leading to biased estimates due to omitted variables, measurement errors, or reverse causality. The IV method helps eliminate this bias.
Our key variable is the patient’s residential location and its effect on health outcomes, specifically hip and knee replacement surgeries. We assess hospital length of stay and readmission rates at 30, 90, 180, and 365 days post-discharge. However, residential choice is not random—healthier individuals may prefer urban areas, or urban healthcare facilities might improve outcomes regardless of location, creating an endogeneity issue.
To tackle this, we use the municipality-level diffusion of ADSL in 2010 as an instrument for urban or rural residence. ADSL diffusion is strongly correlated with urbanization but does not directly influence post-surgery health outcomes except through its effect on residential choice.
Our IV strategy follows a Two-Stage Least Squares (2SLS) approach. In the first stage, urban/rural residence is regressed on ADSL diffusion and other exogenous controls, isolating the variation in residence driven by ADSL rather than health-related factors. In the second stage, the predicted values of residence from the first stage replace the original variable in the outcome equation.
Our findings indicate that urban patients have shorter hospital stays but higher readmission rates at all considered intervals compared to rural patients. Several factors explain this pattern. Urban patients have easier access to healthcare facilities, facilitating readmission for complications. This accessibility may also lead doctors to discharge urban patients earlier, knowing they can return if needed. Additionally, urban patients may be expected to receive outpatient rehabilitation, but adherence to such programs varies, increasing the risk of complications.
In contrast, rural patients often stay longer in hospitals, possibly due to limited outpatient services, ensuring a more complete recovery before discharge. Rural patients may also be less likely to seek hospital readmission for minor complications, contributing to lower readmission rates.
Our study highlights the importance of considering residential location in healthcare policy. The findings suggest that while urban patients benefit from shorter hospital stays, their higher readmission rates may indicate a need for better post-discharge follow-up. Conversely, rural patients’ longer hospital stays could reflect necessary adjustments to account for limited post-hospitalization care.
These insights contribute to the broader discussion on healthcare access and efficiency, emphasizing the need for policies that optimize hospital stays without increasing readmissions, particularly for urban patients. Future research should explore interventions that improve post-surgical care adherence and reduce disparities in healthcare outcomes based on residential location.
At the same time, we are working on a new paper analyzing how the hospital discharge process affects the length of stay. Specifically, compared to patients discharged directly home, those discharged to a nursing home or home with the need for additional social or healthcare services tend to remain in the hospital longer. This prolonged stay contributes to reduced system efficiency.
Main policy, industrial and scientific implications
Policy Implications
Our findings have significant implications for healthcare policy, particularly in designing strategies to improve post-surgical care and resource allocation. The higher readmission rates among urban patients suggest that hospitals may be discharging them too early due to the availability of nearby medical facilities. This calls for policies that promote better outpatient follow-up programs, ensuring patients receive adequate post-discharge care to reduce unnecessary readmissions.
For rural patients, the longer hospital stays indicate a reliance on inpatient care due to limited access to outpatient rehabilitation services. Policymakers should consider expanding telemedicine and mobile health programs to support rural patients after discharge. Investments in rural healthcare infrastructure, including rehabilitation centers and home care services, could help bridge this gap, reducing the need for prolonged hospitalization.
Industrial Implications
From an industry perspective, these findings highlight the need for healthcare providers, regulators, and medical technology firms to adapt their services to different geographic contexts. The demand for outpatient rehabilitation services and home-based care solutions is particularly high in urban settings, creating opportunities for businesses specializing in remote patient monitoring, physiotherapy services, and AI-driven healthcare management.
Medical device manufacturers could develop smart rehabilitation tools, such as wearable sensors that track recovery progress, helping urban patients adhere to rehabilitation programs and reducing complications that lead to readmission. Pharmaceutical companies might also explore targeted interventions for pain management and mobility improvement to minimize the risk of post-surgical complications.
In rural areas, telemedicine providers could play a key role in post-surgical care by offering remote consultations and digital rehabilitation programs. Hospital networks could benefit from strategic partnerships with home healthcare providers, ensuring continuity of care for rural patients once they are discharged.
Scientific Implications
This study contributes to the ongoing debate on the role of residential location in healthcare outcomes. It demonstrates how using an Instrumental Variable (IV) approach can help address endogeneity issues in health economics research, setting a precedent for future studies examining causal relationships in medical and policy contexts.
The findings raise important questions about how access to healthcare influences medical decision-making and patient behavior. Future research could explore how differences in healthcare quality, patient demographics, and socioeconomic factors interact with residential location to influence surgical outcomes.
Additionally, our study suggests the need for further investigation into optimal discharge policies. Randomized controlled trials (RCTs) could test the effectiveness of enhanced post-discharge care programs in reducing readmission rates. Longitudinal studies could examine the long-term impact of early discharge on patient recovery, providing valuable insights for refining clinical guidelines.
Finally, from the new paper described above we expect relevant policy suggestions.
Hospitals should implement structured discharge planning protocols to identify patients who will require post-hospital care (e.g., nursing home placement or home-based healthcare services) as early as possible.
Policies promoting better collaboration between hospitals, social services, and long-term care facilities can prevent discharge delays. Creating dedicated hospital teams to coordinate transitions could help streamline the process.
A shortage of nursing home beds or inadequate home-care services may delay discharges. Policymakers should consider increasing capacity, funding more home-care providers, and incentivizing investments in long-term care infrastructure.
Strengthening telehealth and home-monitoring services could allow more patients to be safely discharged home with remote medical supervision, reducing the reliance on prolonged hospital stays.
A more integrated health information system that allows real-time sharing of patient data between hospitals, home-care providers, and nursing homes could accelerate discharge decisions.
By addressing the systemic inefficiencies in hospital discharge processes, policies can enhance healthcare system performance, reduce costs, and improve patient outcomes. Implementing better discharge planning, investing in post-hospital care infrastructure, reforming payment models, and leveraging technology can all contribute to a more efficient and patient-centered healthcare system.
We are currently completing the revision of a first paper, and we are preparing a first working paper before submitting this work for publication in a relevant international journal. (30/09/2024)
The paper will be presented at the American-European Health Economics Study Group - IX Edition, a workshop that will be held in Oxford (UK) at the end of May 2025. In addition, we are finalizing the paper with the aim of submitting this work for publication in a relevant international journal. (31/12/2024)