Concept: Costing Methods: An Overview of Costing Health Services in Manitoba
Last Updated: 2017-10-12
1.1 - Sources of Expenditure Data
Health care in Canada is a public funded service and expenditure data is available from many different sources. These sources usually break down expenditures into various categories that can be used in determining the total cost of health care services. MCHP research has used the following data sources:
- Management Information System (MIS) Data - MCHP glossary term;
- Manitoba Health Annual Statistics (see the Manitoba Health web site for more information);
- National Health Expenditure (NHEX) Report (1975-2007) - NOTE: This report is no longer available on-line - see the National Health Expenditure Database Metadata on the CIHI web site for more information).
Earlier sources of expenditure information for costing research projects include:
- Maryland (1991/1992) financial data;
- financial and statistical data from the Manitoba Hospital Statistics Part I (HS-1) forms (these forms are no longer in use); and
- supplementary sources of statistical and financial data.
1.2 - Types of Cost Data
Depending on the source data and the specific objectives of the study, different types of expenditure data can be included in the total cost of providing health services. The types of costs to consider include direct costs, indirect costs, and full costs.
- Direct Costs - are costs that are directly attributable to patient care. Examples of direct costs include: nursing services, drugs, medical supplies, diagnostic imaging, rehabilitation and food services.
- Indirect Costs - are costs that are not directly related to patient care. Examples of indirect costs include: general administration, health records, information technology, physical plant and maintenance, human resources, volunteer services, capital expenses, and other regional services.
- Full Costs - are all the costs involved in providing health care services. Simply put, this is direct costs plus indirect costs.
In some cases the source of data will determine the ability as to what types of costs can be included in the costing method. The data should be investigated and costs identified prior to the inclusion/exclusion of specific cost elements in the total cost of health services.1.3 - Approaches to Costing Health Services
Two approaches to costing health services can be broadly categorized as micro-costing (bottom-up) or standard (top-down) costing.
1.3.1 - Micro-Costing (Bottom-Up)
This approach requires the researcher to identify and specify all of the resources that are used by individual patients. This type of costing requires detailed knowledge of the treatment and services provided to individual patients. All of the actual costs of the individual's treatment and services are assigned to that one case. For example, calculating the micro-costs for an intensive care unit (ICU) stay requires identifying dollar values for drugs and supplies used in treating the patient, identification and calculation of the nursing costs involved, cost of physician services provided, cost of equipment, and any other costs that can be contributed to the individual ICU stay. In this approach, the cost of similar cases may vary because of a small difference in the treatment and/or services provided.
In Manitoba, a micro-costing approach can be applied to physician and drug data because this data is collected on an individual basis. However, micro-costing has not been implemented for hospital, home care or personal home care data because the accounting systems for these services do not collect cost data for individual patients. For more information about the use of the micro-costing method see Jacobs et al. (1999).1.3.2 - Standard Costing (Top Down)
This approach is commonly referred to as "average" costing because the method takes total health care expenditures and divides it by a measure of total services provided (the output) to determine a cost per patient (Jacobs et al, 1999). A simple formula illustrates this method:
Average cost = total expenditure ÷ total output
For example, if a hospital spends $2,000,000 treating all patients and 2,000 patients are treated, then the average cost per case is $1,000.Two different methods of standard costing have been used at MCHP:
- Per Diem Costs - Per diem costs can be calculated by dividing the total expenditure for services by the total number of days of service to give an average cost per day. In general, per diem costs are considered a poor estimate of individual resource utilization because no adjustments are made for differences in patient or provider characteristics, which are likely to affect resource utilization. However, certain economic analyses, such as a population-based perspective, may appropriately use per diem costs, particularly when differences in user characteristics are not important or more suitable data are not available. This approach has been used for costing hospital, personal care home and home care services. See Wall et al. (1994) and Finlayson et al. (2010) for examples using per diem costing methods.
- Case Mix Costing - With case mix costing, patients are divided into clinically meaningful groups that are expected to use similar amounts of hospital resources. The case mix system assigns a "relative" weight to patient cases and assumes a standard consumption of resources among similar cases. In some methods, adjustments are made for atypical cases, the complexity/severity of the case, and the age range of different cases.
Two different case mix systems have been used over time in the Manitoba data: Case Mix Groups (CMGs™) and Diagnosis Related Groups (DRGs™) or Refined Diagnosis Related Groups (RDRGs™). More recent work uses the CMG™ case mix system. Detailed information about the CMG™ and DRG™ case mix systems can be found in the following concepts:
- Case Mix Groups (CMGs™) - Overview
- Diagnosis Related Groups (DRGs™) - Overview
- Case Mix Groups (CMGs™) versus Diagnosis Related Groups (RDRGs™)
The following steps outline a general approach to determining the "relative" cost of hospital inpatient and day procedure services:
Case mix costing methods have been widely used for costing hospital services in MCHP research (Shanahan et al. (1994), (1996), and (1997); Mustard & Derksen (1997); Jacobs et al. (1999); and Finlayson et al. (2007), (2009)). Information related to the methods used in these research projects is presented in the section titled Costing Hospital Services.
- Identify the relative weight (RW) of individual cases, and sum the total weight for all cases. For CMGs™ and Day Procedure Groups (DPGs™), the relative weight is referred to as the Resource Intensity Weight (RIW™). For RDRGs™, the relative weight is referred to as the Relative Case Weight (RCW).
- Identify the total expenditure associated with the services, using either a micro-costing or standard costing approach. Decide whether the direct costs, indirect costs, or full costs will be included, based on the source of data and what the research requires.
- Calculate the average cost per weighted case (CWC) using the following formula:
CWC = Total Expenditure ÷ Sum of All Weights (inpatient and day procedures)
- Identify the cost of a individual case using the following formula:
Cost of Case = CWC * RW of the case1.4 - Which Method is Appropriate ?
The simple answer to this is that it depends on the research project objectives and the type of cost data that is available to the project.
If timely and accurate data for different services can be collected and analyzed to place the expenditures into different categories, then the researcher will have a choice in using direct costs, indirect costs or full costs. While there is no consensus on how to best allocate indirect costs, if the research demands these costs be included, defensible methods may be applied. Not including these costs clearly underestimates the real cost of patient care - services could not be provided without incurring the indirect costs, so any research question requiring the use of "full costs" should include these categories.
Research to primarily investigate direct patient care would most likely adopt the method that includes only direct costs and exclude any of the indirect costs. Using full costs for research for these purposes would overstate the direct patient care costs which can be directly affected by health care processes. When precision is not required, particularly when relative rather than actual costs are important, a full cost method is appropriate. This method may also be considered when financial, categorical data are limited or are not available in a timely manner.
2.1 - Costing Hospital Services
Hospital services include both inpatient and hospital day procedure (day surgery) cases that are abstracted and collected in the Hospital Abstracts Data.
Over time, MCHP has applied many different methods to calculate the cost of hospital services, based on the availability of data and the purpose of the research project. Most of these approaches involve determining the "relative" cost of services based on a case-mix classification system. The case mix costing approach is described above.
Two MCHP research projects developed a detailed cost list for hospital services. More recent MCHP research uses a Cost Per Weighted Case (CPWC) / Cost of a Standard Hospital Stay (CSHS) value to calculate hospital costs. Other research describes developing the cost using different sources of cost data. One research project developed an index measure in order to adjust the cost of hospital services on an annual basis using several years of data and the CWC from the cost list project. Indexes were developed to mitigate the hours of work required to develop detailed costing methods and allows the measurement of annual hospital service costs more easily. These research projects are identified below.2.1.1 - Cost Lists
Over time, MCHP has developed two cost lists that provide detailed cost information on hospital inpatient and day surgery services. The most recent was developed in 2009 using the CMG™ case mix system. The first cost list was developed in 1999 using the RDRG™ case mix system. The lists identify a standard cost for a specific patient case mix classification, which provides an average cost for services, and does not necessarily represent the actual cost of the services received by an individual. Adjustments to these standard costs can be made for atypical cases, severity of cases, and for the age of the patient depending on the type of hospital abstracts data that is available for use. The development of these costs lists provides valuable information that can be used for economic evaluations throughout Canada.
The 2009 Cost ListIn Finlayson et al. (2009), the 2009 cost list was developed using MIS financial and statistical data from the 2005/06 fiscal year and hospital abstracts data from the 2005/06 and 2006/07 fiscal years. The abstracts data uses the Case Mix Groups (CMG™) case mix data, including complexity, provided by CIHI. The cost list focuses on the direct cost for inpatients and hospital day surgery cases and identifies the "average" cost for over 550 different categories/groups of CMGs™ and DPGs™. Information presented in the inpatient cost list includes: the total number of cases, average length of stay (LOS), and the average cost per case for typical cases for each of three different age categories and by complexity level, for atypical cases and for all similar cases combined. Another cost list is presented for DPGs™ that includes the number of cases, cost per case, and the total cost for each DPG™. The research also presents a cost per weighted case (CPWC) for three different types of hospitals (teaching, urban community, and other) in Manitoba.
For more detailed information, see the Hospital Costing: Using the 2009 Cost List for Manitoba Hospital Services concept, or review the full report titled The Direct Cost of Hospitalizations in Manitoba, 2005/06 by Finlayson et al., (2009).The 1999 Cost List
In Jacobs et al. (1999), the 1999 cost list was developed using hospital data from fiscal years 1993/94 and 1994/1995 and Maryland cost data from 1991/1992. This list reports standard costs for inpatient hospital care by type of case and type of hospital (e.g., chronic and long term care hospitals) and for hospital outpatient/day surgery procedures using the Refined Diagnosis Related Groups (RDRG™) case mix classification system.
For more detailed information, see the Costing Using the 1999 Cost List for Manitoba Health Services concept, or review the full report titled Cost List for Manitoba Health Services by Jacobs et al., (1999).2.1.2 - Using Cost Per Weighted Case (CPWC) / Cost of a Standard Hospital Stay (CSHS) Values
Several recent MCHP research projects use the CPWC / CSHS values to calculate costs for hospital services. For more information on these methods, see the Calculating Hospital Costs Using Cost Per Weighted Case (CPWC) / Cost of a Standard Hospital Stay (CSHS) Values concept.
2.1.3 - Other Hospital Costing Research
Several early MCHP research projects use the RDRG™ case mix system because this is the type of data available in the hospital abstracts data at that time. Three of these projects (Shannahan et al. (1994), Shannahan et al. (1996), and Shannahan et al. (1997)), as well as the cost list developed by Jacobs et al. (1999) are described and compared in the Hospital Costing in the 1990's concept. Other concepts describing research that use the RDRG™ case mix system include: Adjusted Clinical Group (ACG) - Relative Cost Weights , Costing - Low Birth Weight Infants Project , and the Hospital Care Cost Index concepts.
More recent MCHP hospital costing research uses the CMG™ case mix system. Additional concepts documenting the use of the CMG case mix system include: Hospital Costing in the 1990's - see Mustard et al. (1997), Calculating Hospital Costs Using Cost Per Weighted Case (CPWC) / Cost of a Standard Hospital Stay (CSHS) Values , and Hospital Costing Using the National Health Expenditure (NHEX) Database.2.2 - Costing Physician Services
Physician cost information is collected in the Medical Services (Physician Claims) data. However, not all physician services are recorded in this database. For example, not all physicians in Manitoba are paid on a fee-for-service basis. Some physicians are paid by salary, session rate, or contract. Physicians on these alternative payment plans are only required to submit claims for administrative purposes as a record of service provided, and in some cases, these may not be recorded in the Medical Services database. In addition to this, the medical services database contains mostly ambulatory (out-of hospital) services, and does not collect all in-hospital physician services.
For physician services in the Medical Services database, the cost of a physician service is equal to the professional fee, or tariff, paid to a physician for the service provided.
NOTE: For the cost of services related to emergency departments (ER) and laboratory testing, if present in the Medical Services database, these only contain the physician portion of the service and do not include other costs associated with the ER or laboratory (i.e.: nursing hours or tech time). In addition, in-hospital diagnostic services are funded through the hospital's global budget and may be reported separately in the source data. These costs will not appear in the physician services data.
For a more detailed description on physician service costs and data related to physician services, see the following concepts in the MCHP Concept Dictionary:
Another method investigated at MCHP related to the cost of physician services is the development of a Physician Price Index. This index was developed to be able to measure physician costs in constant dollars (dollar values equivalent to the same year) and allow the comparison of changes in costs over a number of years. For more detailed information, see the Physician Price Index concept in the MCHP Concept Dictionary.2.3 - Costing Home Care Services
The Home Care data in the Manitoba Population Research Data Repository contains a record of each person receiving home care. This database contains information on the number of days a person had an open file (i.e.: was an active home care case), but does not contain information about the specific services a person received while receiving home care services.
In Finlayson et al. (2010), a per diem method is used to estimate home care costs:Per diem cost of home care = total expenditures of home care program ÷ total number of open file daysIn this research, total expenditures of the Home Care program were taken from the Manitoba Health Annual Report, 2005-2006. The annual report information is available on the Manitoba Health web site at: http://www.gov.mb.ca/health/ann/index.html.
It is important to note the limitations of the per diem costing method in this situation. For example, without service data, we cannot identify people who are receiving different levels of services, such as homemaking versus nursing care, and therefore the actual cost of services cannot be attributed to an individual patient. This is why a per diem rate is used.
See the following concepts for more detailed Home Care information:
2.4 - Costing Personal Care Home (PCH) Services
The Long Term Care (LTC) Utilization History Database in the Manitoba Population Research Data Repository contains service information for LTC/Personal Care Home (PCH) utilization in Manitoba. There are two types of PCH's in Manitoba - proprietary (for profit) and non-proprietary. Non-proprietary homes include those sponsored by ethnic and religious organizations as well as non-affiliated homes. These homes are block funded through the Regional Health Authority (RHA) in which the PCH is located. Proprietary PCHs in Winnipeg are funded through the Winnipeg Regional Health Authority (WRHA) on the basis of a negotiated agreement.
The cost of personal care homes is shared between the province (Manitoba Health) and the person who requires the services. See Manitoba Health - Personal Care Services - Questions and Answers About Personal Care Services and Charges web site at: http://www.gov.mb.ca/health/pcs/qanda.html for more information.
PCH residents pay a per diem based on net family income. As of August 1, 2009 this ranges from $30.60 to $71.80. Manitoba Health pays the remainder of the cost for each day of care through funding provided to the RHA. The shared funding between Manitoba Health and the resident covers the cost of the following services:
- meals (including meals for special diets);
- assistance with daily living activities like bathing, getting dressed and using the bathroom;
- necessary nursing care;
- routine medical and surgical supplies;
- prescription drugs eligible under Manitoba's Personal Care Home Program;
- physiotherapy and occupational therapy, if the facility is approved to provide these services; and
- routine laundry and linen services.
In Finlayson et al. (2010), a per diem method is used to estimate both proprietary and non-proprietary personal care home costs:Per diem cost of PCH = total expenditures of PCH program ÷ total number of resident daysIn this research, total expenditures of the PCH program were taken from the Manitoba Health Annual Report, 2005-2006. The annual report information is available on the Manitoba Health web site at: http://www.gov.mb.ca/health/ann/index.html . The LTC Utilization History database at MCHP is used to determine the number of resident days.
Note: The weighted number of resident days in a PCH have been used as a component to calculate the cost of PCH use in other MCHP research (Shanahan et al., 1997) and (Jacobs et al., 1999). The weighted days were estimated using the number of nursing hours dedicated to a resident; the more nursing hours required the higher the weight. Weighted days have been recommended as an appropriate method for costing and have been calculated using the Repository at MCHP. However, work by Finlayson et al. (2007) suggests that there is very little difference in per diem costs based on weighted and un-weighted days.
See the following concepts for more detailed information on PCHs:
2.5 - Costing Pharmaceuticals/Prescription Drugs
The Drug Program Information Network (DPIN) Data in the Manitoba Population Research Data Repository contains information on pharmaceutical/prescription drug utilization in Manitoba. DPIN is an on-line, electronic prescription drug database that connects all pharmacies in Manitoba to a central database maintained by Manitoba Health. Information regarding pharmaceutical dispensations outside the hospital is captured in real time for all Manitoba residents (including Registered First Nations), regardless of insurance coverage or final payer.
DPIN collects data on prescriptions issued to Manitobans from different sources/carriers, including Pharmacare (C1), Personal Care Homes (PCH) (C2), Employment/Income Assistance (C3), Palliative care (C4), and non-adjudicated sources (third party/client pay). The type of data collected includes: a patient identifier, the type of drug, the Drug Identification Number (DIN), dosage, prescription date, ingredient cost and professional (dispensing) fee. Using the DIN, the Anatomical Therapeutic Classification (ATC) code can also be identified. See the Drug Program Information Network (DPIN) Data Description for more detailed information.
NOTE: The DPIN system does not collect data from hospital pharmacies, nursing stations, ward stock and outpatient visits at CancerCare Manitoba. Researchers working with pharmaceutical data should keep in mind that the "relative" costs of hospital inpatient drug utilization are included in the Case Mix Group (CMG™) and Day Procedure Group (DPG™) weights.
For costing purposes, the total cost of a prescription is the sum of the drug ingredient cost plus the professional (dispensing) fee. In some cases, the values may need to be imputed. For more detailed information on pharmaceutical/drug costs, see the Calculating Costs/Expenditures for Pharmaceuticals concept in the MCHP Concept Dictionary.