Cost Analysis (Part 2) of Essential Health Services in Cambodia
Cost Analysis (Part 2) of Essential Health Services in CambodiaMOH/WHO Health Sector Reform Phase III Project
Steve Fabricant, Sok Kanha, and Khout Thavary
Final Draft 11 December 2003
Executive Summary:
In order to complete the costing of the MPA and CPA, this study examined the costs of National Health Programs, National Hospitals, and the cost of Provincial Health Departments and Operational Districts. This information complements the previously-reported cost data from MOH district and provincial hospitals and health centers.
This report corrects an error found in the first study due the use of incorrect denominators (provincial populations) in calculating per capita costs for provincial hospitals. The corrected per capita costs of MPA and CPA at provincial health facilities based on a sample of pilot districts are (costs in parentheses include depreciation):
Health Centers (non-bedded): $0.96 ($1.09)
District Hospitals $0.73 ($0.88)
Provincial Hospitals $0.46 ($0.52)
The additional per capita costs of MPA and CPA found in this part of the study are:
Operational District Offices $0.15 ($0.18) (only non-pilot districts)
Provincial Health Department $0.24 ($0.28)
National-level hospitals and 10 national programs add the following per capita costs:
4 National Referral Hospitals $0.32 ($0.42) (3 referral hospitals missing)
10 National Health Programs $0.76 ($0.80) (NCHADS and others missing)
NCHADS (from separate evaluation) $0.53 (National budget and donor funding)
A nominal total cost of MPA/CPA is $3.62 per capita ($4.17 with depreciation), or $4.15 excluding depreciation but including NCHADS. The latter four cost categories above are only suggestive of true costs because: a) the ODO costs are only for non-pilot districts; pilot districts are much more costly; b) the costs of 3 non-sampled general referral hospitals are not included; and c) costs direct service delivery by NGOs in HIV/AIDS and reproductive health are not included, and some national centers are also not included.
As a general observation, national referral hospitals and programs are operating with a very low ratio of staff costs. Unit (bed-day) costs are calculated for the National Referral Hospitals, ranging from $10.00 to $28.00, with costs of surgical bed-days between $33.00 and $48.00. Cost recovery is effective in the referral hospitals, with approximately 20% of total costs from user fees. About 50% of the costs of the national programs comes from external donors.
There are major discrepancies between costs of the national program reported here and the estimates and projections made in the 2004-2006 Public Investment Program.
19
Cost Analysis (Part 2) of Essential Health Services in Cambodia
Cost Analysis (Part 2) of Essential Health Services in CambodiaMOH/WHO Health Sector Reform Phase III Project
Steve Fabricant (WHO Consultant), Sok Kanha (MOH), and Khout Thavary (MOH)
December 2003
I. Study Objectives and Methods Used
The main objective of the Cost Analysis of Essential Health Services in Cambodia study described here is to determine the full costs of the Minimum Package of Activities (MPA) and the Complementary Package of Activities (CPA). The initial part of the study (disseminated in 2002) examined in detail the full and marginal operational costs of health centers, district hospitals, and provincial hospitals, including a sample of health facilities using contracting. This second part adds the costs of the national programs which support the MPA and CPA technically and logistically, the tertiary referral hospitals which also support the basic services, and the costs of managing the provincial and district health facilities through the Provincial Health Departments (PHDs) and the Operational Districts (ODs). The current and projected outputs of the national programs, and the outputs of the hospitals were also determined so that future financial requirements can be projected.
Although the MOH budget and the Priority Action Plan (PAP) give information on the national government contribution to the cost of all the programs, this field study was designed to capture additional funding sources, including user fees and external donors. In addition, the data collected reflects actual expenditures from the previous year, rather than the nominal budgets.
This part of the study was financed by WHO/Cambodia, with field data collection was managed by the Ministry of Health. This report presents an analysis of data collected for this study. Information on costs and outputs was collected from a sample of ten national health programs and four referral hospitals, and from three representative provincial health departments that included five operational districts.
The authors appreciate the support of Dr. Char Meng Chuor, Director of the Ministry of Health Department of Planning and Health Information, Mr. Chea Kim Long, Director of the Ministry of Health Department of Budget and Finance, and Dr. Aye Aye Thwin, Sector-Wide Management Advisor, WHO Cambodia. The data collection team from MOH did excellent work and also contributed many useful ideas.
Sampling details
The sampled National Referral Hospitals were the Preah Bat Norodom Sihanouk Hospital, (a large general hospital with 365 official beds, surgical cases, and outpatient clinic), The National Tuberculosis Hospital (150 beds, outpatient clinic, but no surgery), the National Pediatric Hospital (94 beds, surgery, and outpatient clinic), and the National Maternal and Child Center (142 beds, surgery, and outpatient clinic)[1]. The MCH Center is administratively a part of the National MCH Program, and the Tuberculosis Hospital is related to the National TB and Leprosy Program.
The National Programs sampled were the National TB/Leprosy Control Program, the National Malaria Control Program (including malaria/filariasis, dengue hemorraghic fever and schistosomiasis as separately budgeted programs), the National Maternal and Child Health Program (including Reproductive Health, Immunization, Nutrition, and Cholera as separately budgeted programs), and the National Center for Public Health (including the clinical laboratory), and the National Center for Health Promotion.[2]
The sampled Provincial Health Departments were chosen to represent the wide range of populations and operational districts which are found among the 24 Cambodian provinces and municipalities. The sampled provinces are Kampong Cham (10 ODs, 128 Health Centers), Takeo (5 ODs, 70 Health Centers), and Kampong Chhnang (2 ODs, 34 Health Centers). One operational district in each of Takeo and Kampong Chhnang provinces was selected as a sample, and two ODs in Kampong Cham province, one of which was a “contracting-out” pilot district.
Specific costing methods used:
The costing methodology employed was a simplified form of that used in the first part of the study, allowing valid comparisons of the costs of hospital services. MOH staff members received one week of training and then collected data during August of 2003. Data on staff costs were obtained from the individual hospital, program, or PHD/OD accounting offices, which incorporate MOH salary data, NGO records of bonuses and incentive payments, and user fee reports. The cost of drugs and supplies was derived from internal accounting sources, as were the other operational costs. Other information needed such as lab and x-ray outputs, was available from the hospitals.
Stepdown of Hospital Costs:
“Stepdown” allocation of hospital indirect and paraclinical costs was used to estimate the costs of major hospital services (inpatients, surgical patients, and outpatients). The stepdown followed the following procedure:
a. For each of the four major hospital services or cost centers under consideration (general inpatients, surgical patients, outpatients, and paraclinical services), the total direct costs was calculated by adding labor costs to the cost of drugs and supplies.
b. Total indirect costs was calculated as the sum of administrative staff labor, operating costs, maintenance, and “other operating costs”.
c. Total indirect costs were allocated among the four cost centers on the basis of the relative share of direct labor costs for each service.
d. The total cost of paraclinical services was calculated by adding direct paraclinical staff costs, costs of paraclinical supplies (reagents, films, chemicals, blood bank) and the paraclinical proportion of indirect costs.
e. Total paraclinical costs were allocated to general inpatients, surgical inpatients, and outpatients according to the proportion of tests performed for each of the three services. (No attempt was made to calculate average costs separately for x-ray and laboratory tests. Except for the TB Hospital, the ratio of the two categories was nearly the same, so little error is introduced by weighting them equally.)
f. Total costs for general inpatients, surgical inpatients, and outpatients were calculated as the sum of their direct costs (staff labor and supplies), the allocated indirect costs, and allocated paraclinical costs.
g. Depreciation was allocated proportionally to each cost center’s total of direct and indirect costs.
Hospital unit costs
Unit costs for the sampled hospitals (cost per admission or bed-day) were calculated in two ways: on the basis of actual utilization (cost per admission), and on a basis of assumed 100% utilization (cost per bed-day). Although the former is more often used as an indicator of hospital costs, it does not permit a true comparison of costs or efficiency between hospitals since those unit costs depend on occupancy rate. On the other hand, comparison of costs at assumed full occupancy does allow a fair comparison between hospitals’ cost structures, but does not take into account the effect of actual utilization rates.
Depreciation
Capital depreciation costs are an indicator of the rate which capital assets are being used up and must be replaced. This was an emphasis of the first part of the MPA/CPA costing, and has been carried on in the second part. Depreciation was calculated conventionally, with some exceptions. For vehicles, annual depreciation was calculated by the recorded purchase cost of cars, pickups, and ambulances which were less than ten years old, divided by the expected 10-year lifespan. For motorbikes, a five-year expected life was used, and only the purchase cost of those less than five years old was used.
For major equipment, the total cost obtained from asset records was divided by ten years. Clearly, this method does not take into account the depreciated or expected remaining life of assets, but to do this would have been very complex and would have added little useful information. It is believed that the approach used provides an adequate estimate of recurrent capital costs for the purpose of budgeting for maintenance and replacement of vehicles and equipment.
Estimating the depreciation costs of buildings was more problematical since the asset values of most buildings are carried in the MOH asset books at their original cost, which is much lower than a realistic estimate of current replacement costs. The cost of some buildings was not known, so an estimate of $200 per square meter has been used, proportional to the estimated $180 per square meter for concrete buildings in rural areas used in the first part of this study. Some national hospital and program buildings were provided by donors, and are carried on the asset books at the cost of construction paid by the donor to a foreign contractor, many times higher than local replacement cost.
In general, building depreciation is a grey area in this study. For the national programs (with the exception of the TB program), the influence of the building capital depreciation costs on total costs is rather small, but it is significantly large for the national hospitals studied, especially the MCH Hospital. The costing results are reported (as in the first part of the costing study) both including and excluding depreciation.
Non-governmental budget contribution to hospital expenditures: The percentage of labor costs and total costs contributed by user fees was also calculated, and the percentage of total costs contributed by external donors.
National Programs
No stepdown was done for program costs because the cost-accounting data available from programs did not support calculation of the cost of individual program activities. The basic cost categories used followed those used in the recent budget exercise: operational cost categories included staff costs, IEC, supplies and drugs, and other operating costs[3]; capital costs included training, equipment, new construction, and other investment.[4]
The percentage of total costs contributed by external donors.was also calculated, and also for user fees in the case of one national center. Costs are reported both including and excluding depreciation.
It was thought initially that future budgetary requirements for the national programs could be extrapolated from the year 2002 costs and reported progress indicators. Since most of the programs studied have multiple activities with different degrees of progress, and since these individual activities could not be costed, this proved to be an unfeasible method. Options for estimating the costs of fully-functioning health services at higher rates of utilization are presented in Annexe 1.
Provincial Health Departments and Operational District Offices
Again, no stepdown was necessary, and the costs of the sampled PHDs and ODs are reported directly according to the same budget items mentioned above.[5] Costs are reported both including and excluding depreciation.
I. Study Results:[6]
A. Hospital costs:
Table 1 presents the basic financial and depreciation costs for the sampled hospitals. Cost patterns differ widely among the four hospitals. This reflects differences in the hospitals’ mission and management, which could be further clarified by a focused study of costs.
Table 1: Basic annual cost structure of sampled national hospitals
Preah Bat Sihanouk Hospital / National TB Hospital / National MCH Hospital / National Pediatric Hospital / Means and PercentRecurrent cost / Total inpatient beds / 365 / 150 / 142 / 94 / 188
Staff Labor / Inpatient wards / $152,085 / $25,846 / $137,542 / $8,774
Surgery block+ ICU* / 16,845 / $0 / 58,948 / 24,066
Outpatient / 28,654 / $5,364 / 40,429 / 29,956
Paraclinical / 13,226 / $11,098 / 31,201 / 18,400
Administrative / 12,794 / $0 / 117,609 / 130,721
Total labor / $223,604 / $42,308 / $385,728 / $211,917 / 21.5%
Drugs and Supplies / Inpatient wards / $98,286 / $420,430 / $185,196 / $138,762
Inpatient surgery & ICU / 98,892 / 0 / 13,042 / 117,465
Outpatient / 29,512 / 285,037 / 52,168 / 2,469
Ancillary (paraclinical) / 3,865 / 7,126 / 10,434 / 14,655
Total drugs and supplies / $230,556 / $712,593 / $260,839 / $273,351 / 36.8%
Indirect costs / Operating costs / $619,950 / $96,241 / $493,571 / $250,462
Maintenance / 5,243 / 100 / 2,375 / 13,018
Other / 66,228 / 30,818 / 0 / 99,709
Total indirect costs / $691,421 / $127,159 / $495,945 / $363,189 / 41.7%
(Line a) Total all accounting costs / $1,145,581 / $882,060 / $1,142,513 / $848,457 / 100.0%
Depreciation / Buildings / $330,044 / $14,517 / $677,324 / $25,141
Vehicles / 9,285 / 0 / 3,008 / 12,345
Equipment / 19,572 / 33,889 / 117,889 / 23,286
Total depreciation / $358,901 / $62,923 / $798,222 / $60,771 / 24.2% of below
(Line b) Total cost incl. depreciation / $1,504,482 / $944,983 / $1,940,735 / $909,228 / 100.0%
The building depreciation amounts introduce some upward bias into the totals in the tables and may be disregarded when making overall comparisons. The high building depreciation for Sihanouk Hospital is the result of the large building size, while for the NMCH Hospital it is a result of the high cost of construction as maintained on the asset register. Mean depreciation for the four hospitals is 24.2% of total costs, compared to about 14% for provincial health facilities.