KENYA DYNAMIC HEALTH SERVICE COSTING MODEL

Transcription

1 KENYA DYNAMIC HEALTH SERVICE COSTING MODEL USER MANUAL G I Z

2 Contents INTRODUCTION TO THE COSTING MODEL... 1 GETTING STARTED... 1 MODEL ASSUMPTIONS... 6 RESOURCE REQUIREMENTS FOR KEPH CONDITIONS UTILISATION SCENARIOS OUTPUT ii

3 LIST FIGURES Figure 1: Set up new user... 1 Figure 2: Login existing user... 2 Figure 3: Naming a new dataset, selecting and existing dataset... 2 Figure 4: Main menu... 3 Figure 5 Recalculate instruction after changes in assumptions... 5 Figure 6: Information at bottom of screen... 5 Figure 7: Assumptions sub-menu... 6 Figure 8: Add or delete condition menu... 6 Figure 9: Update medical supplies and drugs... 7 Figure 10: Update salaries... 9 Figure 11: Assumptions inflation assumptions... 9 Figure 12: Assumptions medical supplies mark-ups Figure 13: Assumptions facility productivity Figure 14: Assumptions building costs Figure 15: Assumptions depreciation Figure 16: Assumptions cost drivers by department type Figure 17: Assumptions capacity thresholds Figure 18: Distribution of KEPH patients Figure 19: Human resource requirements by condition Figure 20: Outpatient medicine requirements by condition Figure 21: Inpatient human resource requirements by condition Figure 22: Inpatient medicines requirements by condition Figure 23: Distribution scenario menu Figure 24: KEPH allocation by provider Figure 25: Non-KEPH utilisation Figure 26: Scenario facility menu Figure 27: Menu for running unit costs Figure 28: Menu for summary assumption Figure 29: Menu for summary costs Figure 30: Menu for costs by economic classification Figure 31: Menu for costs by function classification Figure 32: Menu for costs by broad service categories Figure 33: Menu for details of costing analysis Figure 34: Menu for unit cost iii

4 LIST OF ACRONYMS AND ABBREVIATIONS ALOS DCM FBO KEPH NGO Average Length of Stay Dynamic Costing Model Faith Based Organization Kenya Essential Packages of Health Non-governmental Organization iv

5 INTRODUCTION TO THE COSTING MODEL The Kenya Health System Costing Model is a Web based framework for estimating the current and future costs of services. The model was originally developed in Excel and Access to operate on individual machines. Estimates are performed in two principle ways. The general costs of health services are estimated for the base year and for a user-specified scenario year for inpatient and outpatient care. In addition, the costs of the Kenyan Essential Package of Health Care (KEPH) can be estimated in detail for a range of 60+ conditions. Both sections of the model can provide total and unit costs of services. The costing for health conditions used both activity-based costing and step down cost. The activity based costing provides the cost of drugs/ supplies and direct labour (medical officer or doctor, clinical officer, and nurse) used to provide a service to one patient or client in the outpatient department or inpatient department. The step down provides other recurrent cost, equipment cost, infrastructure cost. These costs are given by outpatient visit and by inpatient day. The model is menu driven and the user does not need to interact or even see the main tables upon which the model is based. GETTING STARTED The model is web-accessible at the following URL: The model works in most browsers but is configured for Microsoft Explorer and functions best in this environment. On first login, the user will be asked to set up an account with choice of user name and password (Figure 1). Figure 1: Set up new user 1

6 Once this is done the user is presented with a login screen (Figure 2). Figure 2: Login in existing user Once the user has logged in, a screen for selecting a data-set is presented. This screen will include previously created datasets and the option to create a new dataset. The user can either select a preexisting data set or create a new dataset (Figure 3Figure 3). Figure 3: Naming a new dataset, selecting and existing dataset 2

7 For new datasets, the user chooses a name for this dataset. If this is the first time the user has used the model, the only option will be to create a new dataset. New data sets all make use of the same baseline data-set with standard options and assumptions, These can be modified by the user and are automatically saved. The user can later revisit the modified data. Once a data set is selected the user will be taken to the main menu page (Figure 4). Figure 4: Main menu The main menu is has 7 items. The first 4 items provide options for manipulating the data in order to develop scenarios for cost. These are the basic model assumptions that are will be applied to scenario projections ( Model Assumptions ), options to inspect or modify the assumptions about the way KEPH conditions are treated ( Resource requirements for KEPH conditions ), Targets for scenarios ( Utilisation Scenarios ) and options on facility size optimisation ( Facility optimisation options ). The option to recalculate costs should be used once the user is satisfied with changes. A warning on the front page will be given when a recalculation is necessary ( 3

8 Figure 5). 4

9 Figure 5 Recalculate instruction after changes in assumptions The final two items provide outputs. The first of these ( Summary Costs ) provide detailed costs generated for the baseline and scenarios. The final option ( KEPH Unit Costs ) provides estimates of the unit costs of different KEPH conditions at different levels of the health system. There is a [Home] button on each page that takes the user back to the home page. Figure 6: Information at bottom of screen The information at the bottom of this and subsequent screens provides basic information on the session including: name of current user logged in, name of the current data set, Scenario year selected, scenario target, capacity optimisation and the current version of the software (Figure 6). 5

10 MODEL ASSUMPTIONS The model assumption menu has four main options for specifying assumptions used in computation of the baseline and scenario costing: condition list, medical supplies list, salary levels, and general assumptions (Figure 7). Figure 7: Assumptions sub-menu Add or Delete Condition allows the user to add or delete a KEPH condition. The user firstly selects the main and sub group for the condition. A new condition can be added by typing a name in the New Condition text box, selecting the Target Group from a drop-down list that corresponds to the target group of the new condition, and clicking on the Add icon (Figure 8). To delete a single condition, the user selects a condition from Condition drop-down list and click on the Delete icon. Figure 8: Add or delete condition menu 6

11 Medical Supplies List allows the user to inspect and modify medicine and supplies (Figure 9). The options for each item include currency, pack price, units per pack, type of unit, and dose per unit. By default all items are specified in Kenyan Shillings (Ksh). Please note that the price applies to the base year. Prices in scenario year will automatically be adjusted for inflation specified in the general assumptions. The additional items can be added into medical supply category table as an administrator option. Figure 9: Update medical supplies and drugs Medical Staff Salary and Allowance at all levels of the system can be specified here ( 7

12 Figure 10). These are the salaries and allowances (if applicable) that are used in the base and scenarios. The current values reflect, for public facilities, the salary scales prevailing in the base year (2012). In non-government facilities they reflect reported salaries by FBO/NGO and private facilities. For the scenario year salaries will automatically be increased for staff inflation (see General Assumptions). It will mostly not be necessary to change the salaries. Any changes applied to staff will to both base and with the addition of inflation the scenario year. 8

13 Figure 10: Update salaries General Assumptions reveal a series of seven menu options, accessed by tabs, used to specify how costs are calculated by the model. Price Increases (Figure 11) specify percentage annual increases in the main economic cost categories: variable costs including medical supplies, staffing and overheads (recurrent and capital). Figure 11: Assumptions inflation assumptions 9

14 Medical Supplies Mark-ups take account the cost of storage, wastage and distribution in the estimation of the price of medical supplies (Figure 12). Figure 12: Assumptions medical supplies mark-ups Facility Productivity makes assumptions about how hard to work facility resources, including target bed occupancy, the relative resource requirements of an inpatient day to an outpatient visit 1, working days per year (out of 365 days) and hours per day for staff (after allowances for breaks, training and administration) (Figure 13). Figure 13: Assumptions facility productivity 1 This allows computation of an adjusted bed-day. This is included because it provides a crude comparator between all types of facility and output (inpatient and outpatient) by converting outpatient visits into resource equivalent bed-days. This is on the basis that the resources required to provide an inpatient day is roughly equivalent to the 5 (by default) outpatient visits. So adjusted bed-days = bed-days + outpatient visits/5. 10

15 Building Cost presents the costs per square metre of new building at different levels of the facility. It includes basic building, utilities, fixtures, and fittings but excludes medical equipment (Figure 14). Figure 14: Assumptions building costs Depreciation indicates the expected life of different capital items (Figure 15). A discount rate can also be specified. A positive value implies that the simple straight-line annualisation of capital, that is item value/expected life, will be adjusted for the opportunity cost of capital according to a net present value (NPV) formula 2. Figure 15: Assumptions depreciation Scenario Cost Drivers specify the determinants of an increase in each of the main cost categories for each of the main categories of facility department (Figure 16). Costs are divided into three categories as follows: 2 To confirm the formula: present value of asset and summing a geometric series 11

16 Variable costs that are expected to rise in direct proportion to workload e.g. drugs and medical supplies Semi-Variable costs that increase with workload but not proportionately e.g. salaries. These costs usually increase in a stepped way, with a given level of staff able to absorb and increase in workload up to a threshold. Once this threshold is exceeded more staff are required. Fixed costs costs that increase not directly with workload but with the overall size of the facility or not at all e.g. building maintenance, utilities. These proportions are used in the calculation of costs in the scenarios. Workload refers to numbers of bed-days and outpatient visits (in patient departments) or the activity of support departments such as laboratory tests. Beds in service departments refer to the number of beds in that department. ALL beds refer to the total bed stock of the facility, a proxy measure of the physical capacity of a facility. The proportion of costs that are not assigned to these drivers will remain unchanged with changes in activity of the facility. They will, however, change if a smaller or larger number of facilities are specified. Figure 16: Assumptions cost drivers by department type Capacity Thresholds specify the maximum ratio of capacity ( 12

17 Figure 17). The model will automatically warn the user during scenario building when it estimates that the capacity of facilities exceeds this ratio. This is based on the bed occupancy target and also threshold increases for inpatients and outpatients. Once the baseline activity plus the threshold is exceeded a warning is given. 13

18 Figure 17: Assumptions capacity thresholds RESOURCE REQUIREMENTS FOR KEPH CONDITIONS The normative costs menu permits the user to enter or modify resource information on the way KEPH services should be provided. These data are used to compute normative costs for both baseline and scenarios. The user firstly selects the main and sub group for the condition and the condition itself. Treatment for acute conditions is entered for the complete episode of that condition. For chronic conditions, treatment is for an entire year e.g. sufficient ART drugs are included to mitigate the effects of the disease for a year. Please note, the main idea here is to reflect the resources that are required to deliver treatment for the condition at a good level of quality that can be feasibly financed by the government or insurance funds. It is not intended to include all resources that are required for the most expensive medical care that, from a medical perspective, would be classified as redundant. Where possible the resources recorded should reflect standard treatment guidelines or protocols. Patient Distribution specifies each condition for which the user then enters the proportion of patients requiring hospitalisation and the target group for treatment ( 14

19 Figure 18). Users then specify information on drugs and medical supplies by age group and staff input. Data entry for outpatients is available for all conditions. For inpatients data entry is only available if the user has specified a proportion requiring hospitalisation. 15

20 Figure 18: Distribution of KEPH patients Outpatient Care: treatment and staffing specify information on members of staff directly looking after the patients: clinical officer, medical officer, nursing officer, and senior medical officer (Figure 19). Other staff members are incorporated into the indirect costs of patient treatment computed by the facility allocation section of the model. For each outpatient visit the proportion requiring the staff contact and numbers of minutes of contact time is entered. Figure 19: Human resource requirements by condition Outpatient Care: medicine and test specify drugs and medical supplies by age groups (Figure 20). Age may determine both the type of medicine given and the dosage rate. For each medicine selected the 16

21 user enters the proportion of patients requiring the item, dosage per day and number of visits the medicine is required for. Figure 20: Outpatient medicine requirements by condition Inpatient treatment and staffing and Inpatient medicines and test specify similar information for inpatients for the duration of the patient (Figure 21 & Figure 22). In addition, the average number of days in hospital is added plus minutes of operating time if required. Figure 21: Inpatient human resource requirements by condition Figure 22: Inpatient medicines requirements by condition 17

22 UTILISATION SCENARIOS This section is to develop scenarios that examine how much it would cost to deliver services to an increasing proportion of the core target groups. A rough comparison of the KEPH patients recorded in the survey with the expected number of patients based on full coverage 3 of likely illness suggests that services covers just under 40% of population needs for the KEPH services. A scenario year is selected first. Then, the model is loaded with census projections to the scenario year so will adjust the size of the target groups according to these projections. KEPH Targets specify the target population the expected incidence/prevalence of the disease in target population, the proportion requiring treatment and inpatient admission rate for each KEPH condition (Figure 23). The user also asked to specify a target level of coverage by the scenario year. 100 percent will ensure that all the needs of the target group are covered but this may turn out to be unaffordable or considered unrealistic in terms of service take-up. The model cannot make these judgements but will show users the implications in cost terms. Two sliders are provided. The first changes the target level of coverage for the current KEPH condition shown above. After selecting, the user should select the Update button at the bottom of the page. 3 Full coverage implies that all patients requiring treatment in target groups associated with each condition are provided with services. 18

23 The second slider updates all conditions simultaneously to one desired level of coverage. The user should select Update All to update all conditions. Figure 23: Distribution scenario menu The Percent of Group with Condition is defined as the proportion of the target group with the condition in any year. It is noted that for some condition this may exceed 100%. So, for example, for childhood diarrhoea children under the age of five would be expected to have at least four episodes in a year, implying a probability of 400%. At the same time not all episodes require treatment. Many of these episodes will be self-limiting or require simple oral rehydration at home. The Percent with Condition Requiring Treatment specifies what proportion of these episodes will require formal medical intervention. KEPH Allocation by Provider and Non-KEPH Utilisation allow the user to consider how the workload should be spread between public, NGO/FBO, and private providers for inpatient and outpatient care ( 19

24 Figure 24 & Figure 25). The default is the current pattern of services. At baseline this indicates that 75.6% of outpatients and 77.6% of inpatients are treated in public providers. Around 39.3% of outpatients and 10.6% of inpatients are treated in health centres. For scenario modelling the user is able to vary these assumptions to assess the impact on costs. Use the Update button once the desired changes have been made. 20

25 Figure 24: KEPH allocation by provider Figure 25: Non-KEPH utilisation 21

26 Facility Optimisation Options are used to consider how the model optimises the use of facility capacity across the country (Figure 26). There are three options. The default is to keep the same number of facilities present at the baseline. If the model has previously been set to optimise, then the number of facilities can be re-set to the baseline by clicking the Reset button ( Reset Facilities to Baseline ). The disadvantage of this scenario is that it could lead to over-crowded facilities or conversely some facilities that are barely used. The second option, is to increase the facilities for types where there is already overcrowding ( Increase Facilities where necessary ). It does not assume that there is a reduction in the number of facilities of any type. This might be considered a politically manageable option since it requires no down-sizing. The final option is a full optimisation scenario where facility numbers are increased or reduced where necessary to achieve the target level of utilisation. It gives an idea of maximum technical productivity but could be hard to implement. It is also not necessarily efficient since it could necessitate patients being forced to travel long distances to get to the nearest facility. Figure 26: Scenario facility menu Once the user is happy with the scenarios, the unit costs need to be generated by selecting the option Calculate Unit Costs (Figure 27). Computation takes around 2 minutes but depends on the speed of the connection to the server. Please note that although summary results will be available before this option is selected results are likely to be unreliable. Output based on the baseline and scenario calculations can now be inspected. Progress with the re-calculation is indicated. Figure 27: Menu for running unit costs 22

27 OUTPUT The model produces the actual costs of the baseline year, the normative cost of the same year and prediction for actual and normative costs for future years. The model distinguishes between all levels of health care services and all trustees. The user has access to the data base in order to simulate scenarios, such as changes in the work load, share of conditions, staffing levels, salaries, and prices in other input factors (e.g. drugs). Based on this data base the updating studying generates outputs on actual costs of health care services in Kenya of the financial year from 2013 to The results of the analysis are provided in terms of summary, detailed output and condition unit cost. Summary output provides overall Summary Costs for baseline and scenario (including normative when available), Economic code, Function (outpatient, inpatient, and national programmes), and KEPH Detail as well as Summary Assumption (Figure 28). The projected cost of scenarios depends on many factors described above, such as the target population, incidence/prevalence of disease, need for treatment, target coverage (proportion of target population actually covered by KEPH), and etc. Figure 28: Menu for summary assumption Actual Costs of the KEPH accounts for the current institutional variable costs (drugs, medical supplies and other supplies) and the fixed costs (utilities, maintenance, etc.) incurred by each of three costs centres: overhead, support, clinical support for varying levels of care and types of ownership in the health system using a type of step-down cost allocation process. Normative Costs of the KEPH allows the generation of cost calculation on the basis of a defined set of standards on how KEPH services could be provided with alternative resource inputs. In contrast to the actual cost calculation, which captures costs of services as it is currently given, the normative cost calculation sets the resource information according to national protocols and according to possible changes in system needs. 23

28 As a preliminary step an example scenario was undertaken for the year 2020 (Figure 29, Figure 30, 24

29 Figure 31 & Figure 32). This year is chosen to illustrate the effect of coverage since much of the cost increase is associated with the increase in targets rather than population or inflation. Figure 29: Menu for summary costs It is also assumed that non-keph utilisation holds constant at rates per 1,000 population and the distribution of workload between sectors is same as the baseline. An estimate of total actual costs is KSh 285 billion KES or about $ 78 USD per capita. Figure 30: Menu for costs by economic classification 25

30 Figure 31: Menu for costs by function classification Figure 32: Menu for costs by broad service categories In addition to these outputs, Detailed Output by health sector and per condition or disease are generated ( 26

31 Figure 33). By clicking on each below, the detailed costing analysis table is provided. 27

32 Figure 33: Menu for details of costing analysis Unit Costs of KEPH Conditions provide an immediate calculation of the unit costs of individual KEPH conditions. This incorporates both the direct (supplies and direct staff) and indirect element (Figure 34). The latter is determined by the level of facility, length of stay and outpatient visits (user specified). Figure 34: Menu for unit cost 28