What do Cambodia’s public health services cost and why does this matter?

Introducing a routine health services costing system to improve efficiency and quality of health services  

Well-baby clinic at Tain Krosaing Health Centre

As Cambodia gears up towards Universal Health Coverage, a German-supported project aims at improving effectiveness and efficiency by determining the costs of public health services. Knowing these costs, differentiated by type of facility, will enable the Ministry of Health to realistically budget for services and to establish how much 3rd party payers should be charged for them.

Cambodia has come far since three decades of genocide and war (1969-98) that cost millions of its citizens their lives. With resolute government policies, between 2006 and 2014 the country reduced its poverty rate from 47.8% to 13.5%, and its maternal mortality rate from 472 /100,000 live births to 170. In 2015 Cambodia attained the status of Lower Middle-Income Country (LMIC).

Ensuring access of all its citizens to adequate healthcare has been a constant priority of the Royal Government of Cambodia, which launched a far-reaching health system reform in 1995. In addition to multiplying the number of public health facilities and the services provided, it introduced social health protection schemes to improve financial access. The Health Equity Fund (HEF), a social assistance scheme managed by the Ministry of Health, currently covers approximately 2.5 million people identified as living in poverty out of Cambodia’s total population of 16 million, and the Ministry of Labour’s National Social Security Fund (NSSF), a contributory social health insurance scheme, covers an estimated 1.8 million government and private sector employees.

Health budgeting is predominantly historical and not necessarily needs-based

However, budgeting and rational use of resources remain a weakness of the public health system. The major source of income for the facilities is the budget provided by the Ministry of Economy and Finance, and the budgeting process is mainly incremental: it is not necessarily based on population needs. Consequently, costs are given insufficient attention by public health providers and everyone just spends the budget provided. Since budgeting does not reflect prevailing needs, efficiency is not guaranteed, nor optimum use of human and financial resources to enhance quality – particularly important in the context of an LMIC with limited resources.

Furthermore, not knowing what each service really costs means that 3rd party payers for health protection schemes such as HEF and NSSF purchase services from providers on the basis of ‘guesstimates’. On the one hand these do not ensure that real costs are adequately covered, and on the other they do not allow for strategic purchasing to encourage quality of care. Imminent negotiations with NSSF make it all the more urgent to determine the actual costs of health provision as a realistic basis for setting prices.

Reforms to attain UHC must be grounded on a sound financial basis

Knowing what health services cost is vital for Cambodia’s aspiration to attain Universal Health Coverage (UHC), which must be based on sound health financing arrangements and a rational use of resources. Without a clear understanding of which costs need to be covered and how these funds can be raised in a sustainable manner, there can be no UHC. The Ministry of Health, with support from the German Federal Ministry of Economic Cooperation and Development (BMZ) through the Deutsche Gesellschaft für Internationale Zusammenarbeit (GIZ), therefore assessed in 61 health facilities in seven provinces how much specific services cost the health system – differentiated by the level of facility at which they are provided. The researchers also examined if such health service costing could be set up as a routine system rather than as an occasional exercise.

The findings of this study were meant to:

  1. guide the public health system in negotiating realistic prices with the 3rd party social health protection payers NSSF and HEF; 
  2. support the Ministry of Health and health facility managers in making rational budget decisions; and 
  3. provide a sound basis for policy decisions on health system organisation and functioning.

How was the costing assessment conducted?

Data entry at Kampong Thom Provincial Referral Hospital

Based on service data from fiscal year 2016, a cost assessment was conducted in 2017 in the three provinces where the GIZ-supported Social Health Protection Project is operational – Kampong Thom, Kampot and Kep – in a total of 20 health centres and five hospitals. The assessment was then extended in collaboration with the National Institute of Public Health (NIPH), and with additional financial support from USAID and the Czech Republic, to another four provinces with 24 health centres and 12 hospitals.

As a first step the researchers determined the type of data they would need in order to calculate unit costs for each service on each level of health facility. On this basis they developed a data collection tool to gather information on 1) expenditures such as costs for labour and supplies; 2) facilities’ income from various sources such as government, patient fees and 3rd party reimbursements; and 3) basic service provision statistics from each facility.

The costing information was collected with great attention to detail. None of the selected facilities was able to provide 100% of the required information, and much of it was not available in electronic form. Complementary sources included information gathered on district or provincial level, and in selected departments and programmes of the Ministry of Health. For each health facility the researchers conducted specific investigations on staff remuneration and costs of drugs, laboratory and imaging, as well as calculating the current value of equipment, vehicles and buildings, based on purchase cost and date and subsequent depreciation. When the data were only available on paper, the results of one or two months were extrapolated to the entire year.

Provider costs were calculated using a ‘step-down allocation’, attributing indirect (overhead) costs from support services such as laundry, pharmacy or laboratory to the cost centres of departments where curative and preventive services to patients are rendered. Finally the total amount of these costs for each of these cost centres was divided by the number of provider-patient contacts in order to determine average cost for the respective costing unit, e.g. for health centres cost per curative service, per delivery, or per preventive service, and for hospitals cost per outpatient visit, per inpatient day or per inpatient stay.

What are the preliminary findings of the costing assessment?

The first phase of the assessment conducted in the three GIZ-supported provinces has provided initial findings that will be completed when the data from the other four provinces and the national hospital have been processed, to give a national representative picture.

In these 25 facilities, on average 88% of income came from the Cambodian government, provided mostly in kind in the form of salaries (36%) and drugs and medical materials (45%). Reflecting the government’s policy of strongly subsidised public health services, user fees made up only 12% of facilities’ income. Of these 12%, 7% were from out-of-pocket payments of user fees, as recorded in the health facilities’ logbooks and monthly financial reports to the district level, while the remaining 5% were social health protection scheme payments on behalf of their members/beneficiaries.

This very low proportion of 7% for out-of-pocket payments for user fees in the public health facilities contrasts dramatically with the figures for total health expenditure in Cambodia, of which 60% is attributed to out-of-pocket payments by users. In addition to expenses such as for transport or meals for accompanying family members, this high proportion reflects priority recourse to self-medication and the private sector, a major challenge for Cambodia’s health system. Paradoxically, it is in part the rationality of the public sector that makes it less attractive to patients: The commercial private sector is ‘customer-oriented’ with appealing surroundings, welcoming personnel and short waiting times. Above all, private clinics give patients easy access to popular – and expensive – treatments such as injectables and intravenous infusions, whereas public services are required to follow medical protocol, including referring patients to a distant hospital for more advanced treatment. Using nearby private facilities reduces transport costs, and many allow their patients to pay for their treatment in instalments.

The unit costs calculated for each type of service showed striking variation between facilities. For instance, on health centre level the average unit cost per facility for an outpatient visit varied from 1.30 USD to nearly 7 USD. When correlated with the quality scores of the Ministry’s annual Quality Assessment exercise, the researchers were in for a surprise: Higher unit cost did not correlate with better quality, but rather with lower use. This makes sense: If two health facilities have the same budget, staff and supplies, but one treats 100 patients per day, and the other only 50, the latter’s resources are being used less efficiently than the former, and each patient-provider contact costs the double of the more frequented facility. The correlation between quality and number of contacts could be an interesting additional dimension to investigate.

Private clinics are perceived as a factor, not only in patients’ paying the lion’s share in health costs overall, but in high unit costs and low efficiency of nearby public facilities due to underuse. Preventive services, on the other hand, are well frequented in the public facilities, since they present little financial incentive for private clinics to provide them.

Health centres with beds for hospitalisation are significantly costlier than those without because of the larger staff required for round-the-clock service and their generally low bed occupancy – some such centres had no inpatients at all in 2016.

Such hard figures finally provide a rational basis for policy-making. They can have important implications for potentially reorganising Cambodia’s health services, e.g. by shifting human and financial resources from underutilised facilities to ones with greater demand. At the same time, equity must be kept in mind: Rather than just basing new planning on average unit costs, the great variation revealed by the assessment may reflect a need to invest more in poor and isolated areas to maintain equitable access for all.

What are the next steps for Cambodia’s public health sector?

The NIPH appreciates the costing assessment, as it is a first effort to get a large enough sample size for a number of purposes: first of all, setting prices for 3rd party payment by NSSF and HEF; secondly, more rational budgeting, for example of personnel and drugs; and finally, making policy recommendations on health system reorganisation, for instance rather than maintaining costly beds in certain health centres and lower-level hospitals, referring patients directly to higher-level, more fully equipped district and provincial hospitals.

The results of the assessment in the initial three provinces also confirm the need to invest in the quality of public health centres so that patients use them as their entry into the health system (since it is cheapest to treat at the lowest appropriate level), rather than presenting directly on hospital level.

...and for health services costing?

There is a consensus that this costing assessment should be the basis for a long-term, standardised and reliable routine costing system to guide future health service planning and price negotiation with 3rd party payers. The requirements for any such system, to more conveniently obtain and process the data, would include a basic standardised accounting system, as well as checks and balances for accuracy throughout the reporting. The data must not only be correct, but also systematically communicated upwards to NIPH for analysis.

The approach espoused by NIPH would repeat the same calculations at one- or two-year intervals in the same 61 health facilities. This would provide the government with national representative cost figures on a routine basis to make high-level decisions.

In two GIZ-supported provinces, such periodic assessments will be complemented by efforts to strengthen decentralised cost-based decision-making. The data collection tools that were developed for the national-level assessment will be used to continuously generate service cost data from approximately 70 health facilities, allowing provincial, district and facility managers to optimise their services on this basis. To implement this approach, senior staff members of all participating facilities and associated district directorates will be trained in accounting and use of costing data.

The results and perspectives of the costing assessment will be presented to the Ministry of Health and other relevant stakeholders, including health partners, the General Secretariat to the National Social Protection Council, and NSSF. A national workshop will present the costing exercise to representatives of the provinces and hospitals. Costing will be incorporated into NIPH’s Masters’ courses on hospital management and on general management of health services, and research will continue with Masters’ students, to create a critical mass of national consultants and researchers familiar with the topic of health services costing.

If applied systematically, costing assessment has the potential to revolutionise the health system. Once restructuring starts, based on the costing assessment, health providers and managers will become aware of the importance of knowing and monitoring the cost of services. GIZ will continue to support Cambodia both on national and decentralised level in mastering routine costing of health services and in translating its findings into optimisation of the health system.

April 2019
Dr Mary White-Kaba, Dr Bart Jacobs

BMZ glossary

Close window


Share page