http://orcid.org/0000-0003-0154-4942
![Sample our Urban Studies journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110826/TOC-Subject-Sample-2021-Urban-Studies.jpg"})
ABSTRACT
Interregional patient mobility in a decentralized healthcare system. Regional Studies. Interregional patient mobility, measured as origin–destination patient flows between any two regions, is analysed within a dynamic spatial panel data framework using 2001–10 data on Italian hospital discharges. The aim is to assess the effects of the main determinants of patient flows, distinguishing between the impacts of regional health policies and those exerted by exogenous factors (geography, size, neighbouring regions, national policies). Empirical results indicate that the main drivers of mobility are regional income, hospital capacity, organizational structure, performance and technology. Moreover, neighbouring regions’ supply factors, specialization and performance largely affect mobility by generating significant local externalities.
摘要
去中心化的健康照护系统中的跨区域病患流动。Regional Studies. 本文运用2001年至2010年的意大利出院数据，在动态空间面板数据架构中分析以任两区域之间来源 – 目的病患流动进行测量的跨区域病患流动。本文目标在于评估病患流动的主要决定因素之影响，并区分区域健康政策的影响与外在因素（地理、规模、邻近区域、国家政策）发挥的影响。经验结果指出，驱动流动的主要因素是区域所得、医院容量、组织结构、表现与技术。此外，邻近区域的供给因素、专殊化与表现，透过生产显着的地方外部性，大幅影响了流动。
RÉSUMÉ
Mobilité interrégionale de patients dans un système de santé décentralisé. Regional Studies. La mobilité interrégionale des patients, mesurée comme flux de patients origine – destination entre deux régions quelconques, est analysée au sein d’un cadre de données de panel spatial dynamique, en faisant usage de données sur des sorties d’hôpital en Italie, au cours de la période 2001–2010. L’objet est l’évaluation des effets des principaux déterminants des flux de patients, en faisant une distinction entre les impacts des différentes politiques régionales en matière de santé et ceux qui sont exercés par des facteurs exogènes (géographie, taille, régions avoisinantes, politiques nationales). Des résultats empiriques indiquent que les principaux facteurs de la mobilité sont les revenus régionaux, la capacité des hôpitaux, la structure organisationnelle, les performances et la technologie. De plus, les facteurs d’approvisionnement, la spécialisation, et les performances des régions avoisinantes influent dans une grande mesure sur la mobilité, en générant des externalités locales significatives.
ZUSAMMENFASSUNG
Interregionale Mobilität von Patienten in einem dezentralisierten Gesundheitssystem. Regional Studies. In diesem Beitrag analysieren wir die interregionale Mobilität von Patienten – gemessen als Patientenströme vom Ursprung zum Ziel zwischen zwei beliebigen Regionen – innerhalb eines dynamischen räumlichen Paneldatenrahmens mithilfe der Daten von Krankenhausentlassungen in Italien im Zeitraum von 2001 bis 2010. Ziel ist eine Untersuchung der Auswirkungen der wichtigsten Determinanten der Patientenströme unter Unterscheidung zwischen den Auswirkungen der regionalen Gesundheitspolitiken und den Auswirkungen von exogenen Faktoren (Geografie, Größe, Nachbarregionen, nationale Politiken). Aus den empirischen Ergebnissen geht hervor, dass die wichtigsten Faktoren für Mobilität das regionale Einkommen, die Kapazität der Krankenhäuser, die Organisationsstruktur, die Leistung und die Technik sind. Darüber hinaus wirken sich die Angebotsfaktoren, Spezialisierung und Leistung von Nachbarregionen stark auf die Mobilität aus, indem sie signifikante lokale Externalitäten erzeugen.
RESUMEN
Movilidad interregional de pacientes en un sistema sanitario descentralizado. Regional Studies. En este estudio analizamos la movilidad interregional de los pacientes – medida como flujos de pacientes de origen a destino entre dos regiones cualquiera – en un marco de datos de panel espacial dinámico con ayuda de datos de altas hospitalarias en Italia entre 2001 y 2010. La finalidad de este trabajo es evaluar los efectos de los principales determinantes de los flujos de pacientes diferenciando entre las repercusiones de las políticas sanitarias de ámbito regional y las ejercidas por factores exógenos (geografía, tamaño, regiones vecinas, políticas nacionales). Los resultados empíricos indican que los principales factores de la movilidad son los ingresos regionales, la capacidad hospitalaria, la estructura organizativa, el rendimiento y la tecnología. Además, los factores de oferta, la especialización y el rendimiento de las regiones vecinas afectan en gran medida a la movilidad porque se generan externalidades significativas de ámbito local.
ACKNOWLEDGMENTS
The authors thank the Direzione Generale della Programmazione Sanitaria, Ufficio VI, at the Italian Ministry of Health, for kindly providing them with the SDO database – Archivio Nazionale Schede di Dimissione Ospedaliera (2001–10). They thank Daniela Moro for valuable assistance in preparing the database. They are also grateful for the comments made by the participants at the conferences of the International Health Economics Association in Dublin (2014), the meeting of the Italian Health Economics Association in Venice (2014), the Italian Congress of Econometrics and Empirical Economics in Salerno (2015), the Health Economics Workshop in Cambridge (2015), the International Association for Applied Econometrics Conference in Thessaloniki (2015), the Italian Statistical Society Conference in Salerno (2016), the European Regional Science Association Conference in Vienna (2016), and the Italian Economic Association in Milan (2016).
DISCLOSURE STATEMENT
No potential conflict of interest was reported by the authors.
SUPPLEMENTAL DATA
Supplemental data for this article can be accessed at https://doi.org/10.1080/00343404.2017.1307954.
ORCiD
Silvia Balia http://orcid.org/0000-0003-0154-4942
Rinaldo Brau http://orcid.org/0000-0002-6032-0773
Emanuela Marrocu http://orcid.org/0000-0002-7436-0055
Notes
1. The hypothesis of exogenous differences in efficiency levels is consistent with the evidence of the heterogeneous performance of Italian local health authorities, which follow the traditional North–South divide (Baldi & Vannoni, Citation2017).
2. More precisely, the model yields this result when considering the medium- and low-income regions together. Considering them separately, the only region that certainly reduces quality is the medium-income one, whereas for the low-income region the effect is indeterminate.
3. The Poisson or negative binomial estimation procedure has not yet been developed for the spatial autoregressive model (LeSage & Thomas-Agnan, Citation2015). Moreover, the spatial error model specification is not considered because it rules out spillover by construction.
4. On the basis of a preliminary analysis, it was found that additional lags were not significant.
5. When the lagged dependent variable is included, the strict exogeneity assumption no longer holds; in this case, it is necessary to resort to sequential exogeneity (Wooldridge, Citation2010).
6. Non-deferrable mobility is due to the accidental presence of an individual in a region different from that of residence, or as the outcome of central planning on the location of some highly specialized treatments, such as transplants. See Appendix A in the supplemental data online for a detailed account of the excluded admissions.
7. Because population is already included in the regression model, the absolute number of beds is used in place of the beds-to-population indicator. Furthermore, because targets have changed repeatedly over time, it is not possible to build an indicator based on the distance between the observed number of beds and the national target for each of the years considered.
8. The devices considered are those reported in the yearbooks of the Italian NHS: automated immunochemistry analyser, linear accelerator in radiotherapy, immunoassay analyser, anaesthesia machine, ultrasound imaging system, haemodialysis delivery system, computerized gamma camera, differential haematology analyser, analogue X-ray system, surgical light, monitor, mobile X-ray system, computerized axial tomography (CT), magnetic resonance imaging (MRI), medical imaging table, continuous ventilator system, digital angiography systems, hyperbaric chamber, mammogram, positron emission tomography (PET), integrated PET-CT, operating table, and two types of panoramic radiography machines.
9. In Italy there are eight types of RHS-financed hospital care providers: public hospitals, autonomous public enterprises, scientific institutes for research, hospitalization and healthcare, medical school hospitals, private licensed hospitals, research centres, classified hospitals, and LHA-qualified institutes.
10. Some degree of homogeneity in the organizational structure of hospital care might entail some advantages, e.g., in terms of higher efficiency due to the exploitation of economies of scale and more effective financial planning. However, these effects are not expected to offset the benefits arising from higher variety.
11. The reported dynamic CCRE specification, which specifies the individual pair terms as a function of the averages of both time-varying origin and destination characteristics and region-pairs regressors outperforms, in terms of the LR test, the two more parsimonious specifications which only include one set of average terms at a time.
12. Because the model is dynamic, interpretation focuses on short-run effects.