Inequality in geographical distribution of hospitals and hospital beds in densely populated metropolitan cities of Iran

The concern for equitable healthcare access to people is a fundamental element of many health policies. Governments have been committed to ensuring access to healthcare, and success is highly dependent on the nature of existing inequality and in the healthcare needs of people [1]. Thus, access to healthcare is a multidimensional concept that is subject to the influences of both the geographical, socio-demographic [2], information, and economic factors. The service availability is another influencing factor for the geographical accessibility of healthcare [3].

Studies reported differences in the geographical distribution of healthcare facilities based on the socioeconomic status [4, 5] and location of residential areas in a city where suburban residents had lower accessibility to healthcare facilities than residents in the central city [6]. Evidence from a recent study in China indicates extremely high inequalities in the geographical distribution of health facilities, health workforce, and hospital beds [7]. Regardless of the socioeconomic status (SES) and other distinguishing parameters of people, providing an opportunity of equal access to basic healthcare services is essential for ensuring justice in healthcare [8, 9]. However, this is not an easy task, especially in the under-resourced countries where the inequalities in the distribution of healthcare infrastructure and health workforce are common challenges [10].

The provision of healthcare often follows an inverse care law. This law completely operates where medical care is exposed to market forces. In areas with fewer healthcare facilities, more diseases and deaths are likely to happen. The phenomena are more pertinent among the low SES communities [11]. Geographical access to healthcare services contribute to the minimization of the health inequalities [12], and there existed an inverse relationship between geographical access to a healthcare facility and healthcare utilization [13, 14]. Therefore, measuring the locations of hospitals relative to the residents in different SES can provide a clear picture of the geographical inequalities in hospital distributions [15‐17].

Iran’s healthcare system is organized into national, provincial, and city levels [18], and is primarily an insurance-based [19]. The focus of the healthcare system has been on primary care with a clearly defined vision and strong government commitment that demonstrated success in realizing the universal access to health care [20]. However, the issue of marginalization and increasing slum areas due to urbanization and high rural-urban migration challenged the equitable access to healthcare [21]. Despite the reports of few studies on healthcare and distributions of health facilities in the urban and rural settings in Iran [22, 23], the issues of inequalities in hospitals and hospital bed distributions in the highly populated cities received less attention.

A recent study analyzed inequality in the distribution of hospitals and assessed the inverse care law hypothesis. The findings revealed high inequalities in the geographical distributions of hospitals and hospital beds in favor of the relatively affluent areas [24]. The study provided a new way of viewing and measuring inequalities in metropolitan cities. The current study is a continuation of the previous one on a wider scale to determine the extent of the inequalities in the hospitals and hospital beds with respect to the SES of residents in five metropolitan cities in Iran. Measuring socio-economic inequalities in hospital and hospital beds help informed policy decisions to ensure equity in healthcare in the urban settings of Iran and perhaps other similar contexts.

This study analyzed the relationship between residential floor area per capita and locations of hospitals using QGIS, the extent of Gini inequality in hospital beds distributions, and the extent of socio-economic inequality in hospital beds distribution in five metropolitan cities in Iran. Out of the total 306 hospitals included in the analysis, the public, private and other hospitals accounted for 128 (41.8%), 101 (33.0%), and 77 (25.2%) hospitals, respectively. Sixty (46.9%) of the public hospitals, 62(61.4%) of the private hospitals, and 40 (51.9%) other ownership hospitals found in Tehran. The rest hospitals were in the remaining four cities. The aggregated hospital bed to population ratio of the five cities was 1:32 (that is, 32 beds per 10,000 population). The average hospital beds to population ratios were 1:35 in Tehran, 1:42 in Shiraz, 1:37 in Tabriz, 1:30 in Esfahan, and 1:26 in Mashhad, respectively (Table 1).

The hospitals distributions are in favor of districts with larger residential floor area per capita (Fig. 1). For example, 66% (107/162) of the hospitals in Tehran and 51.3% (20/39) in Mashhad are located in the districts ranked in two highest socioeconomic classes. Similarly, the rates for Tabriz, Shiraz and Esfahan were 37.9% (11/29), 57.5% (23/40) and 69.4% (25/36), respectively. The GIS revealed that 216 (70.6%) hospitals are located in the two highest socioeconomic classes in the cities. Only 29 (9.5%) hospitals are located in the lowest class. The SES of the residents is presented in categories ranging from poor (deep red color) to rich (light-blue color) as shown in Fig. 1.

Regardless of ownership, the Gini indices (GIs) for the hospital beds in all the cities were above 0.55 (Table 2). The concentration indices for the private hospital beds in Mashhad and Tehran, and for the public hospital beds in Esfahan and Tehran were significantly high (p < 0.05). The aggregated CI value for the private hospital beds per capita (CI = 0.49) was insignificant (p = 0.059), while for the public hospitals beds (CI = 0.33) and the total hospital beds (CI = 0.29) per capita were significant (p < 0.05) (Table 3).

About 25% of the residents in the lowest rank of the residential floor area per capita have almost no access to the private hospital beds, while 20% of people in the highest rank of the residential floor area per capita have access to 60% of the private hospital beds (Fig. 2a). Furthermore, around 20% of people in the lowest rank of the residential floor area per capita have access to about 5% of the public hospital beds, whereas 20% of residents in the highest rank of the floor area per capita have access to about 50% of the hospital beds (Fig. 2b).

The geographical distribution of the public, private and total hospitals beds in Tehran revealed moderately strong (p < 0.05) positive correlation with the SES of the residents. In Esfahan, the public hospital beds and the overall hospitals beds distributions had strong positive correlation (p < 0.05), while the private hospitals beds distributions in Mashhad and Shiraz were strongly (p < 0.05) positively correlated (Table 4).

This study analyzed the geographical distribution of hospitals and hospital beds in relation to socioeconomic status of districts in five metropolitan cities in Iran. Different approaches were used to measure and characterize the inequalities in hospital beds distributions. The GIS indicated that the majority of the hospitals are located in districts with relatively higher socioeconomic status. Other studies also reported inequitable distribution of hospitals in urban settings [2, 38, 45] and a significant difference between public and private health facilities distributions [46]. This might imply an overlooked issue of inequality in healthcare and the financial and non-financial consequences it might have especially in highly populated cities [10, 47].

A study found inequalities in the distribution of health facilities based on socioeconomic status of people [45]. The establishment of hospitals in areas where the socioeconomically privileged people live like found in our study could have contributed to a negative influence in healthcare utilization and outcomes. This is because not only the geographical distance of the location of the hospitals but also in the highly populace cities that the travel-time to seek healthcare is influenced by high traffic [48]. This situation can lead to delay in medical diagnosis, especially among the socioeconomically disadvantaged population groups [49] and to a high cost of travel to seek healthcare among the service users.

The Gini and concentration indices also revealed high inequalities in the distribution of hospital beds in the cities. The overall inequalities for hospital beds in our study are higher than those reported in another study [50]. However, the Gini values alone are not sufficient to inform policy about the actual inequality in healthcare [51]. Thus, the inequality in private hospital beds with respect to the floor area per capita of residents was significantly in favor of the rich in two (Tehran, and Mashhad) of the five main cities. This might imply that the private hospitals are being established in the districts with the richer people as they are for profit or the owners might have preferred areas that will increase their utility. Such differences might contribute to the worsening of the maldistribution of medical resources [11]. The inequalities in the public hospital beds were significantly in favor of the rich in only two (Tehran and Esfahan) of the five cities analyzed. The concentration curves of the hospital beds also revealed the tendencies to be in favor of the rich. Previous studies in Tehran and elsewhere also found hospital beds distributions and hospital visit per capita in favor of the rich [24, 52, 53].

Our study indicated a disproportionately higher disparity in the distribution of hospitals and hospital beds in the districts of the cities in favor of the rich. A study in Japan identified scarcity of healthcare resources in low socioeconomic population groups, an imbalance in healthcare services provision, and a positive correlation between hospital bed density and inpatient flow ratio [54]. An association between highway and major arterial roads, number of subway entrances, and row house areas with hospital distributions in a metropolitan city was also reported [55]. The different analytical approaches applied in our study clearly demonstrated the inequality in hospital and hospital beds concentrations. The inequalities in geographical distribution of the healthcare infrastructure, particularly the hospitals and hospital beds in metropolitan cities are usually mediated by a disproportionately operating multiple factors including the social, economic, infrastructure and other dimensions. Our study uncovered the extent of existing disparities in the distributions of hospitals and hospital beds in the metropolitan cities in Iran. The findings highlight not only the policy and level of effort required to address the identified gap in access to healthcare in the metropolitan cities but also inform the strategies to be followed to ensure fair access of hospital services and prevent the occurrence of similar issue in the growing urban settings in Iran.

The study has some important limitations. The use of residential floor area per capita is a valid and relevant measure of SES, however, it is not a direct indicator of SES of the residents. Identifying other determinants of access to healthcare such as road traffic, travel time, cost, etc. in the metropolitan cities could better inform policy decisions. This study was limited to few number of metropolitan cities in Iran, and the analysis did not consider the populations living proximal to the study cities, while it could have some effect to the measured inequalities.

This study uncovered the presence of clear inequalities among the main cities in the distribution of hospitals and hospital beds in favor of people in the higher socioeconomic status. The inequalities were particularly high for the distributions of the private hospitals and hospital beds, and  indicates an overlooked but growing concern of access to healthcare in rapidly urbanizing countries like Iran. This implies when need arises for the establishment or opening of new healthcare facilities, regardless of ownership, decision-makers should emphasize to the disadvantaged urban districts in order to ensure fairness in healthcare. Our findings indicated the allocation of hospitals towards more affluent areas in the geographically representative metropolitan cities at national level. As such it is important that hospital establishment plans should not only consider the distributions among different cities, but also within geographical localities of major urban settings. This recommendation should be qualified with the expected limitations related to economies of scale and scope in hospital planning to ensure access to health care and efficiency considerations.