An empirical analysis of Unit costs of public hospitals in Jordan CURRENT STATUS:

Background: Public providers in low-middle income countries face enormous pressures to meet complex needs of its population. This study aimed to benchmark unit costs of hospital services in public providers in Jordan to provide useful insight and readings of healthcare delivery and sustainability. Methods: Unit costs of hospital services per admission, inpatient days, outpatient visits, emergency visits and surgical operations were estimated using standard average costing method. Unit costs per inpatient days was estimated relevant to nine specialities, staff and selected medical services too. Unit costs were estimated in Jordanian Dinars (exchange rate= US$1.41). Results: Average unit cost per admission was 782.3 JDs (U$1101.8), per inpatient day was 236.6 JDs (U$ 333.2), per bed day was 172.9 JDs (U$244.9), per outpatient visit was 58.4 JDs (U$ 82.3), per operation was 449.6 JDs (U$ 633.2) and per emergency room visit was 31.8 JDs (U$44.8). ICU/CCU and OB/GYN presented the highest unit costs per inpatient days across providers; U$532.9 and U$510.7 respectively. Unit costs varied depending on service utilization, type of service and organizational outset. Conclusions: Countries in areas of political conflicts face enormous fiscal constraints, health policy makers, managers and stakeholders ought to understand these challenges and foster use of value based budgeting and performance analysis.

3 expenditures are considerably high in comparison to countries in the surrounding region (4,5,6).
Healthcare system in Jordan is facing rapid changes over the past years due to human crisis in the neighboring countries leading to increase service demand and changes in demographics and disease patterns (7). The crises in Syria, Iraq, Libya, Yemen and the economic slowdown across the area of the Gulf Cooperation increased economic constraints. Jordanians have welcomed asylum seekers and refugees starting from Palestinians in 1970s up to the recent events in the area. In 2015, the total number of refugees in Jordan was estimated at 2.5 million (from Iraq, Libya, Yemen and Syria).
Syrians account for almost half of the non-Jordanians (1.4 million) and 53% of the refugees are reported under the age of 18 years (4). For four years after the break of civil war in Syria, refuges received free health care in Jordan. This placed huge burdens on the health sector. Due to the lack of international support, health services currently provided only to Syrian refugees live inside campus who are estimated at 0.66 million. Those receive health care at the same rate of insured Jordanians by the Ministry of Health (MoH) facilities (4,8). Whilst, refugees lives outside campus pay fully out of pocket for healthcare (4,9). A recent report by the Central Bank of Ireland outlined that Jordan rely on creditors financial assistance to provide a significant portion of its public services. External aid fund in health are estimated at 6.7%, which leaves Jordan economy very sensitive to any changes in the level or terms of support or fund)4,8(. Therefore, Jordan economy is struggling to response to the complex and increased needs of its population.

Health systems in Jordan
The healthcare sector in Jordan is heavily subdivided to multiple providers including public, private and international charitable providers. The insured citizens are estimated at 68% of the total population)4,6(. The majority of healthcare fund is subsided by public providers with only 7-8% of the population has private health insurance coverage either through health insurance companies or selffunds)7(. The public sector consists of three major providers; MoH, Military Royal Medical Services (RMS) and two University affiliated hospitals (UHs).
MoH operates 32 hospitals representing 37% of hospital beds in the country. RMS provides 24% of hospital beds through 14 hospitals, whilst UHs accounts for 3% of hospital beds)6, 7(. RMS provides 4 healthcare services to 1.7 million military individuals and their dependents accounting for 25% of the population. UHs insurance program covers 2.5% of the population)4,6(. Public health provider's beneficiaries receive services with subsidized fees and only pay a margin of health costs amounting only to 20% of health costs)7(. On the other hand, direct out-of-pocket health spending by citizens account for about 33% of total health spending in the country)6(. Such mix of healthcare financing plus the growing challenges in the area beset policy makers with ethical and financial strains to ensure sustainability and equity of healthcare)4,6,7(.
In assessing health efficiency, information about unit costs is pivotal)10(. Unit costs are established tools for health evaluation and policy decisions in many countries )11-13(. A unit cost is defined as the monetary value of the resources that is consumed to create goods or services)10(. Consequently, a unit cost is the collective costs related to the material, time, equipment and capital infrastructure used to carry out a service. Unit costs often developed by direct or model based estimations)12-14(.
There are lack of estimates from developing economies due to data quality and scarcity; Jordan no exception)14-16(. Up to date, there is only one published study from Jordan relied on audit data from 2007)17(. Hammad and her colleagues estimated the direct cost of hospital services in a single public hospital affiliated to MoH in the capital Amman and therefore generalizability was not guaranteed.
Hence, a larger-scale study scoping unit costs from the perspectives of different types of public providers was warranted of value.

Methods
The aim of this study was to estimate unit costs of hospital services in Jordan to benchmark costs of healthcare services in the public sector and inform future policies and initiatives to enhance performance and efficiency.
Cost analysis from healthcare public providers' perspective in Jordan was performed. Unit costs were defined as the cost of one unit of "output" including admission, inpatient days, outpatient visits, emergency visits and surgical operations (12,18,19). Unit costs per hospital output were estimated using standard average costing; top down approach (17,18,(20)(21)(22). Total health expenditures of a provider was divided by the total output produced to determine the unit costs above.
Total health expenditures was defined as the value of health goods consumption for the production of all provider's activities (10). Total expenditures were allocated based on the proportion of resource consumed from the provider budget in the delivery and production of services within inpatient, outpatient, emergency departments and surgical units (8, 19,21). Data were collected and Unit costs of hospital services Unit costs were calculated using the average cost method; the total cost of a care provider was divided by the output units (22)(23)(24). Unit costs per inpatient days included the day of admission with exclusion of the day of discharge or death. Total inpatient days were calculated by adding the daily patient census days relative to 365 days in all departments of the study providers. A bed day was defined as a day during which a patient is confined to bed and stays overnight in the hospital. Total bed days were obtained by multiplying the total number of beds by 365 (21,24). Cost per surgical operations did not include pre or post-operative costs.
Unit costs per speciality were estimated relevant to nine specialities. Specialities require similar labor and capital inputs to treat patients with similar maladies were grouped together (17,18,22).
Total expenditures were allocated to specialities based on the proportion of inpatient admissions 6 relevant to the department. Unit cost per inpatient day per speciality was obtained using the following formula: Cost per inpatient day i = Total expenditure inpatient * [inpatient admission i /Total inpatient admission] ÷inpatient days i. (17,18,22). Inpatient days i is the number of inpatient days for a medical The unit cost per hour was estimated by dividing average salaries by working hours (13).

Unit costs of selected medical services
Charge master lists were used for unit cost estimation of procedures, imaging and laboratory tests.
Charge master lists of all procedures, lab tests, diagnostic and interventional services with its official billed charges in JDs were obtained from providers. List of prices for MoH were extracted from the civil insurance guide published in 2016 (25). For RMS and UHs the charge master lists were dated 2011 and 2012 respectively. These were the official list of charges in effective use up to the date of the study; hence, inflation of unit price was not warranted (23).
Charges for procedures, imaging and laboratory tests were analysed descriptively by the observation of minimum and maximum values. To estimate costs of providing a service to the provider; top-down costing using cost-to-charge ratio (CCR) method was employed. CCR was investigated through expert elicitation discussions and interviews with financial managers in each provider (20,24). CCR presented the ratio between a provider operating and capital costs and revenues for the fiscal year 2016-2017 (26,27). Thus, the cost of each service equalled the billing charge (unit price) multiplied by the overall CCR of a provider. Eventually, median cost and range was stipulated for a selection of medical services deemed mostly performed (26,28).
To analyize cost contributes to a medical services across providers, financial managers and clinical consultants who had current or previous administration duties relevant to allocate or mange budget in each provider were requested to complete a form independently after reviewing a set of medical services. They estimated the proportion of cost contributed by material, labor, equipment and overhead costs to the overall cost of the service. Discrepancies were discussed until agreement. Chisquare test was performed to assess significant variations between providers. P value <0.05 was considered statistically significant.

Results
Background information about the study providers are presented in Table 1 Unit costs of hospital services Table 2 presents unit costs by provider and the average unit costs of public providers in Jordan respectively.
RMS hospital services were costed the highest compared to the rest of providers except for emergency visits. UHs had the highest costs per an emergency visit. Unit costs of medical and ancillary staff Table 4

Discussion
This is the first exploratory study in the region attempted to explore unit costs of hospital services, staff, procedures, imaging and laboratory tests in public providers and pinned out financing characteristics of different type of providers. In recent report, WHO outlined that health system performance vary considerably between countries of the region but highlighted that the main challenges facing health systems is the raised health costs and inequity of health care services particularly in areas affected by acute and chronic conflicts)4(. The complexity of the health outlook is increasing in Jordan and the refugees crisis have contributed to increased demand for health services at MoH and thus referrals to other providers. Thus, it is important for local health policy makers and mangers to understand financing characteristics of the key providers in the country and how costs vary across settings. On the other hand, it is also important for international policy makers and funding organizations to recognize the complexity of healthcare delivery for when terms of financial assistance are set to support refugees and hosting communities)8(. Unit costs varied across providers depending on service utilization, type of service and provider organizational outset. This is in agreement to previous studies explored cost drivers in hospital budgeting decisions in middle to low income countries) 16,19,21,22(. Unit costs of hospital services in military health services were found highest when compared to other providers. RMS services predominantly as a tertiary provider and oriented mainly to inpatient care treating complex referrals from other providers)7(. Hence, higher costs can be potentially due to complexity of care. The cost per outpatient attendance in RMS was the highest too; where services utilization is relatively low though. Higher density of services where most expensive procedures often performed presumably contribute to higher costs (6,7). Thus, information on service utilization and type of service are essential to inform patient co-payment and reimbursement policies (14,23). In addition to the above, type of provider might be an influential factor too. Cost per operation in UHs was the lowest. UHs are teaching facilities and potentially perform greater number of operations by residents or fellows. This has been highlighted previously where labor costs in teaching hospitals found lower when compared to tertiary settings (19,21). Of note, UHs had the highest cost per emergency room visit. UHs services heavily focused on outpatient services that are not available during weekends, holidays and out of working hours. Effective referral system and standard treatment protocols across public provider should be promoted to enhance service delivery and minimized avoidable costs (4).
Unit costs of inpatient days per medical specialty showed highest unit costs for ICU/CCU, gynecology, ophthalmology and neonatal care. This is consistent with previous reports outlining these units as specialized with fewer patients and expensive equipments (11,(17)(18)(19)22). In depth evaluation is needed in future studies to understand cost drivers across departments and disease mix.
The variation across providers in cost components analysis for the same or similar medical services and CCRs was high. This outlines high heterogeneity in the delivery and performance of the public health sector in Jordan. Mostly the existing charges used for patients or payers billing are out dated and ultimately doesn't relate to overall organizational sustainability or profitability. For instance, MoH costs were extremely underestimated compared to the other providers. Despite the fact that almost 92% of MoH budget is financed by Ministry of Finance and the remaining left by donors and households (4,6), the premium paid and co-payment by patients are the main revenues source for MoH. MoH beneficiaries for instances can receive health care services by other public providers without referral or treatment algorithms. Moh eventually would reimburse other public providers on 11 fee for service base. Herein, health authorities should understand that sustainability might be jeopardized in future times without outlining the true health costs for caring for both Jordanians and non-Jordanians. Extensive review of the financing structure and reimbursement strategies across different settings is recommended. Therefore, manager should be opt to adapt evidence based assessment of performance to identify areas where cost savings and profitability can be maximized.

Future Implications
Understanding costs incurred for a patient care is valuable to health decision makers, payers as well as care providers. Adaption of national training and education programs on health economics and pharmacoeconomics is essential. Public providers in Jordan are mostly onus on providing an acceptable level of care within a defined budget rather than generating profit (7,29). Organizational profitability can produce considerable investments to human and capital development. Therefore, health mangers must equip staff with skills and knowledge on cost estimation and efficiency assessment.
Institutionalizing health policy and service research is Jordan is essential too. The Jordan response plan for 2017-2019 has specified needs to strengthen and increase the capacity of the healthcare system in Jordan [8], this can be achieved by sharing and exchange expertise on evidence based decision making between local and international experts. Political support to generate laws legitimizing the role of economic evidence is also needed. These are essential factors to promote the use of economic evidence in service delivery and international donors can help to promote political support by funding new or existing projects enforcing use of economic research in health policy planning (30)(31)(32).

Challenges to pharmacoeconomics research
The biggest challenge faced this study was willingness for data access. Being the first exploratory study that acquired monetary information across various providers, mangers were resistance to share information openly. They perceived this as uncommon practice with unpredicted consequences. The study was granted approvals but with limited access to hospital records and accounting archives.
Consequently, immense efforts were made to validate each estimate reported empirically in this 12 study matrices through triangulation and expert elicitation. Nevertheless, these challenges faced with data collection are in line with challenges reported previously highlighting similar barriers exist to economic research in health across countries (33)(34)(35).
Therefore, knowledge exchange and discussions with experts, mangers, decision makers across countries is valuable to increase understanding of policy and similarity or varices in the economic context across the world.

Strengths and Limitations
The merit of this study is to establish a benchmark for hospital unit costs in Jordan to draw lessons to share with health policy makers, mangers and researchers. It also highlights financing issues that might influence budgeting and reimbursement analysis by donors locally as well in the area.
However, the unit cost calculation of hospital services was based on the average method, which might be different form per patient or per disease estimates (23). Therefore, this study serves as a proxy benchmark and future research attempts to assess costs per patient or disease mix should be pursued.
Unit costs per specialty were calculated by dividing the total cost of each provider or a department by a proxy measure of output. This carry the risk of underestimating units with high capital investments, fewer patients and high maintenance costs (36,21). The study providers had centralized management systems for allocating resources to their individual health facilities. Facilities do not have their own budgets and had limited autonomy. They receive allocations of equipment, supplies 13 and pharmaceuticals as per their needs. This means that transactions are only recorded if they result in payments or receipts of cash.
In this study, charges were used as a proxy to estimate costs of medical services. RCC methodology is considered a less complicated method in practice, since it requires only data concerning the charges for medical services (28). Employing micro-costing method often produce most accurate estimates (23). However, it was not possible to employ this approach due to data quality and accessibility.
Future attempts to draw unit costs estimates derived from micro-costing method must be pursed if institutional and political support can be warranted.

Conclusions
With the political conflicts and the fiscal constrains faced by many low and middle income countries in unrested regions, local and international health policy makers, managers, researchers and donors should understands the complexity and challenges faced by individuals and health care managers.
Efforts and knowledge exchange is advocated to promot resources and tools to improve capacity and health care system performance.   Cost components analysis of selected services across providers

Supplementary Files
This is a list of supplementary files associated with this preprint. Click to download.