Question

The use of ratio analysis is a common technique in financial management for interpreting values on financial statements and putting them into some context about how an organization is performing. The same approach is extremely valuable in operations management for understanding how efficiently an organization is producing services for patients. A health care provider organization operates as a production function with inputs and outputs, just as a factory does that produces goods for sale. In the case of a health care organization, the inputs are varied and examples include labor, supplies, use of outside service vendors, and capital equipment to produce multiple types of output, including a patient day, surgical procedure, diagnostic test, meals for patients or visitors, or claim for reimbursement. As with any other production function (P), operations management seeks to maximize the volume of output (O) for a given amount of input (I) using the ratio: img Depending on the manager’s perspective, the definition of productivity, the inputs, and the outputs used in this ratio may vary. In some cases, the manager may define productivity as total cost per unit of output—one particular cost element per unit of output, such as salaries. Other perspectives may evaluate productivity as the number of inputs (such as labor hours) per unit of output. Health care management operational metrics use this same production function approach where the ratio of inputs per unit of output is measured. In this chapter, the common operational metrics and their derivation will be presented. INPUT MEASURES FOR OPERATING METRICS There are different ways of determining an input used in the production function just described. Depending on the organization’s goal or the particular problem being addressed, multiple input measures may be useful in establishing solid operational controls. Generally, an input can be measured based on the cost of resources devoted to the production of patient care or the number of individual units of a particular resource used. Costs of resources are often used in conjunction with evaluation of results in the organization’s income statement. For example, the cost per unit of output may be used to determine the organization’s performance against income statement goals when identifying if there are “good” or “bad” results in a given accounting period. Using this perspective, a good result would be defined as cost per unit of output being below a target value. Conversely, a bad result would be determined if the cost per unit of output were above that target value. Using cost per unit of output as a measure of operational effectiveness has some benefit in that it is easily derived from the organization’s normally produced financial statements; therefore, data for operational analysis is readily available. Such data is also commonly understood among managers in the health care setting. However, regular variations in operating cost, such as normal inflation, changes in sources for inputs, or changes in the mix or quality of inputs, can all create routine variations from the assumptions made in determining operating cost per unit of output benchmark. Some of these issues may be beyond the operations manager’s control and may limit how meaningful the cost per unit of output approach is to determining operational effectiveness. Therefore, it may be helpful to consider observed results using a cost per unit of output approach, along with the number of units of input used to generate a given level of output. Units of input may provide a more objective view on evaluating operational performance. Examples of such units of input are labor hours; number of supply items, such as syringes or exam gloves; or the number of medications used to produce a unit of output. Units of input do not have the same price variations that a cost per unit of output would have: A labor hour used to produce a lab test may change in value if an employee gets a pay adjustment, but the time used to produce that output remains constant in its measure. Therefore, in operations management it is valuable to know the number of units used to produce patient care outputs to avoid the challenge of weighing the reasonableness of results in terms of changes in the prices of inputs. However, those items are typically not readily obtained from financial statements and require some additional work by the operations manager to gain access to the organization’s statistical reports in order to track the number of units used in a given reporting period. There are multiple reports generated by the organization—often to support preparation of financial statements—that can be used to obtain counts of production inputs in a health care setting. Examples of these sources will be discussed in the next section. When considering units of input in the hospital setting, labor units are significant because labor costs make up more than half of the hospital’s operating expenses. Measuring the productivity of labor in particular can be valuable in understanding variations in the organization’s financial performance. Labor productivity is typically measured by the amount of output per employee or per labor hour. The definition of an employee in the operations management field is usually based on the full-time equivalent (FTE) employee measure. Many operating metrics in the health care field use the FTE per unit of output to evaluate labor productivity. The definition of a full-time employee is 40 hours of productive work in 1 week. Because hospitals operate 24 hours a day, 7 days a week, this means 40 hours of production spread across 7 days in a calendar week. In addition, some employees may not work a full 40-hour week, yet together they equal the amount of time worked by one full-time employee. For example, if two employees both work 25 hours in a week, together they have worked 1.25 FTE (25 hours per week for employee #1 + 25 hours per week for employee #2) ÷ 40 hours for one full-time employee per week = 1.25 FTE). Additional details on other FTE calculations can be found in the Productivity and Performance Management chapter. SOURCES OF DATA FOR OPERATIONAL METRICS Usable data needed to calculate the various operational metrics can be obtained from reports that are routinely prepared within today’s health care organization, including an income statement and statistical compilations. The income statement is used by the organization for external reporting and internal management purposes, along with the balance sheet and statement of cash flows. Many organizations will even prepare income statements at a department level to assist individual department managers in running a specific department. If managers are particularly interested in the costs per unit of output, then the income statement is likely the most valuable data source, especially for comparison of actual operating results with budget targets. This is particularly true when evaluating operational results within departments or subunits of an organization. It is important for operations managers to remember that comparison to budgeted cost targets have some limitations, depending on the changes in the price of inputs used and the mix of different inputs used to generate observed results. Cost comparisons have great value in operations management due to the ease in which data can be obtained from common financial statements, but they must be used with caution to consider any changes in the price or mix of inputs that differ from the assumptions used in setting budget targets. Routine financial statements are often supplemented with at least a basic description of the operating statistics for the organization to provide some context to the reader of the level of activity described in financial statement results. In some cases, such as the filing of government-required annual reports, certain operating statistics, such as patient days, discharges, and employee data, are mandatory. Perhaps the most common example of such mandatory reports is the Medicare Cost Report submitted to the Centers for Medicare and Medicaid Services (CMS) by hospitals, skilled nursing facilities, and other institutional providers that participate in the federal Medicare program. As a result, financial managers are likely already collecting a wide array of statistical data that is used to prepare required reports to external parties. These data can provide valuable insight into operations management in measuring the volume of inputs used to generate organizational output. The departments within an institutional provider such as a hospital often collect operating statistics to measure activity levels for use with internal management reporting or to document patient care rendered during a given time period. These data may include manual patient logs that can be summarized or a compilation of daily transaction logs in a department. Another excellent source of data for operational inputs to the production of patient care outputs is the accounting records used to generate financial reports, such as payroll journals or inventory control reports. The labor distribution usually categorizes paid labor hours as being productive, overtime, vacation, sick, or other classifications and can be valuable for identifying productive FTE for operational analysis. Inventory control reports can describe the units of supply issued to a department for use during a specified time period and can be associated with output volumes to evaluate supply inputs to production. A list of commonly used operational data sources is presented in Table 7–1. Generally speaking, health care organizations are considered to be data-intensive enterprises and so have a wealth of statistical data that often goes unused in operations management. The challenge for the operations manager is to understand which data are collected in the organization, how the data are collected, how the data can relate to the organization’s operational performance, and how to obtain the data with a minimum of disruption to normal production functions. Table 7–2 provides an example income statement with basic operational statistics for a small community hospital. Data from this table will be used to calculate the operating metric examples to follow. Table 7–1 Examples of Sources of Operational Data Input/Output Source(s) Labor Cost Organization or departmental income statements Supply Cost Organization or departmental income statements Labor Hours Payroll journals, labor distributions Supply Units Inventory management journals Emergency Room (ER) Patients Served Department volume logs, patient accounting records with patients having ER services, medical record counts of ER patients Tests Performed Department volume logs, patient accounting records of tests charged Surgical Procedures Performed Department patient logs, medical record counts of surgical procedures Table 7–2 Sample Income Statement and Summary Operating Statistics Example Community Hospital Summary of Financial & Operational Data for the Year Ended 12/31/20XX Inpatient Revenues $66,179,014 Outpatient Revenues 24,966,033 Total Revenues $91,145,047 Allowances & Discounts $35,820,003 Bad Debt 1,066,397 Total Revenue Deductions $36,886,400 Net Revenue $54,258,647 Salaries & Wages $27,621,506 Contract Labor 1,287,162 Benefits 7,374,942 Supplies 9,392,171 Repairs & Maintenance 1,268,733 Purchased Services 980,245 Other Operating Expenses 732,612 Total Operating Expenses $48,657,371 Operating Margin $5,601,276 Depreciation & Amortization $6,169,524 Total Margin ($568,248) Investments $1,252,376 Donations 309,893 Total Non-Operating Income $1,562,269 Net Income $994,021 Beds in Operation 76 Patient Days 14,543 Discharges 2,796 Outpatient Visits 36,877 Productive Labor Hours 644,890 Non-Productive Labor Hours 77,387 Total Paid Hours 722,277 OUTPUT MEASURES The common output measures in an institutional health care provider organization such as a hospital relate to one of two types of service: inpatient or outpatient. Inpatient volume measures have been the traditional index of output for a hospital since the history of hospital care in the United States until the mid-1980s and centered on care to patients who stayed in the hospital for a period of more than 1 day. Since then, the traditional inpatient volume measures have evolved to take into account services provided to patients who visit the hospital for care but do not stay overnight—the outpatient. The patient day has been the most common measure of output for a hospital over time and represents one patient staying in the hospital’s inpatient care units at midnight on a given day. The count of patient days in a hospital is based on the hospital’s daily midnight census. For example, if Hometown Hospital has 63 patients in beds in its inpatient care units at midnight on March 3, then it has produced 63 patient days of care. Patient days are typically reported on a monthly, quarterly, or yearly basis and commonly accompany the income statement for a hospital. Because a hospital can compile patient days on a daily basis for a time period greater than 1 day, managers often look to an average number of patient days to gauge the level of inpatient activity for a certain period, or average daily census (ADC). If Hometown Hospital recorded 2,105 patient days during the month of March, then its ADC for March is 67.9 (2,105 patient days during the month ÷ 31 days in March = 67.9 ADC). When considering inpatient volumes over a period of time, either the patient day or ADC is an appropriate measure of hospital output. When a patient enters the hospital for an inpatient stay, that event is counted as an admission and is a common operating statistic in hospitals. Since inception of prospective payment by Medicare in the mid-1980s, hospital payments have been based on when the patient leaves the hospital, also called a discharge. Because discharges represent the complete occasion of care for a patient (whereas an admission represents only the start of an inpatient hospitalization), operations management uses the discharge as a measure of the number of inpatients served in a given time period. If Hometown Hospital sent five patients home after an overnight stay in the hospital on September 23, then it has recorded five discharges for that day. As with patient days, discharges are usually totaled during monthly, quarterly, and yearly time periods. As mentioned earlier, hospitals have moved away from a focus on care to patients who stay overnight in the hospital and toward services to outpatients. However, outpatient units of service can be counted in a myriad of ways—tests performed, procedures completed, or treatments performed. Further, a simple test in the laboratory (such as a routine urinalysis) may be less sophisticated than an outpatient MRI scan or an outpatient orthopedic surgery. It can be difficult to identify one meaningful measure of output for an organization with multiple outputs of varying sophistication or focus. Thus, the adjusted patient day is used as an index of a hospital’s total output. It takes the inpatient days produced in the hospital for a given time period and inflates them to account for an estimate of the relative value of outpatient services provided during the same interval. The adjusted patient day is calculated using the formula: img An illustration of this calculation uses data from Table 7–2. Example Community Hospital recorded 14,543 patient days, $66,179,014 in inpatient revenues, and $24,966,033 in outpatient revenues. Using these data, the adjusted patient days during the year for Example Community are calculated as: img The same adjustment can be applied to the hospital’s count of discharges to express inpatient discharges in terms of the hospital’s overall inpatient and outpatient outputs. This measure is called the adjusted discharge and is calculated as: img Using the same data from Example Hospital yields the following calculation: img Considering the multiple types of output produced in a hospital organization, these aggregate measures of output are the most common for assessment of hospital operations. If the focus of an operational assessment is a particular department or subunit of the hospital, the department’s specific output, such as tests, examinations, treatments, or procedures, may be used. Because a specific department’s output will generally be the same for an inpatient or an outpatient, it is not necessary to adjust for inpatient or outpatient volumes when looking at the department’s operational performance. For example, if the radiology department produced 12,000 tests for inpatients and another 3,500 for outpatients, the 15,500 total tests represent the total output for this department. If the department uses a relative value unit measure such as the College of American Pathologists (CAP) unit, the same approach would apply. If the hospital lab produced inpatient tests totaling 162,500 CAP units and outpatient tests that total another 44,000 CAP units, the lab’s output can be expressed as 206,500 CAP units. COMMON OPERATING METRICS There are several common operating metrics used in today’s hospitals. The following section will define these metrics and provide an example using data from Table 7–2. As mentioned earlier, total patient days for a period are typically assessed using an average over a specified period of time (month, quarter, or year) and expressed as average daily census (ADC) or average occupied beds (OB). The same holds true for adjusted patient days, and a common metric to determine adjusted patient day volumes is adjusted average daily census (AADC) or adjusted occupied bed (AOB). The AADC metric calculated using example data for the past year as well as the adjusted patient day calculation shown earlier is completed as follows: img Comparing the AADC calculated here with the 39.84 inpatient ADC (14,543 inpatient days ÷ 365 days in a year = 39.84) suggests that Example Hospital produced about 37.75% of the output for outpatients that it did for inpatients during the past year. The number of patient days can be a valuable measure of how efficiently a hospital completes treatment of a patient’s condition. Given that today’s hospitals are typically paid a fixed prospective amount per discharge from Medicare and many managed care plans, the incentive is to minimize the number of days a patient stays before discharge. This metric is known as the average length of stay. An example calculation using data from Table 7–2 yields the following result: img This calculation tells the manager at Example Hospital that, on average, an inpatient stayed in the hospital 5.29 days before discharge. Comparing this value to a benchmark length of stay can tell the operations manager if patients are staying longer than perhaps they should, based on the experience of other facilities, and may identify a potential area of improvement for the hospital. Management makes decisions on how much capacity to make available in a hospital, usually expressed by the number of beds available for patients to occupy. Knowing the extent to which that capacity is being used can help determine if the organization is supporting unused capacity or is operating at a high level of utilization that could result in turning business away. This metric is called the occupancy percentage and is calculated using data from Example Hospital as follows: img This calculation indicates that Example Hospital is operating at about 52% of its available capacity and may have the opportunity to attract additional business or perhaps reduce the available number of beds to decrease the resources expended to support unused capacity. Labor is one of the largest resource inputs used in a hospital to produce patient care services, and the costs of labor can ruin the organization’s financial results. While labor costs are important to hospital management, the management of actual labor hours can be the key to effectively controlling labor costs that appear on financial statements. This can be measured using the ratio FTE employees per occupied bed (FTE/OB). The data for Example Hospital presented here is for a 1-year period, where a full-time employee would work 2,080 hours (40 hours × 52 weeks in a year). Using an annual FTE hours basis, the FTE/OB value is calculated to be: img This result shows that Example Hospital uses an average of 7.78 FTE for every inpatient served in the hospital each day. The FTE/OB metric does not take into account the volume of outputs produced for services to outpatients. If a hospital provides a significant volume of services to outpatients, the FTE/OB metric may not fully account for a hospital’s workload. To address this concern, measurement of FTE employees per adjusted occupied bed (FTE/AOB) may better express the ratio of labor inputs per unit of total output for the hospital. Using the AADC value for Example Hospital calculated earlier, the FTE/AOB for the past year is: img Thus, Example Hospital used an average of 5.65 FTE in the production of 1 adjusted patient day during the past year. If a manager wishes to evaluate the operational efficiency of a specific department, then the same relationship described in the FTE/OB or FTE/AOB metrics for productive labor hours per unit of output can be used to calculate productive hours per unit in a specific department. If the radiology department of Example Hospital recorded 11,463 productive hours in the past year to produce 16,772 procedures in the past year, the hours per unit are calculated as: img This calculated result tells the radiology manager at Example Hospital that it takes about 41 minutes (0.68 hours per procedure × 60 minutes in an hour = 41) of employee labor to produce one test for a patient. Conversely, the department manager may want to know how many procedures per employee are produced per year. Using data from the radiology department at Example Hospital, the number of procedures per employee is: img Because the productive hours in the radiology department for the year translate to 5.51 FTE, and those labor hours resulted in production of 16,772 tests, then on average one full-time employee produced 3,043 tests. The metrics described so far look at units of output per unit of input. However, the operations manager should still review operating expenses per unit of output to evaluate the total mix of resources used in producing a unit of output. It is not feasible to calculate the different units of measure for the multiple inputs used in producing patient care services in a hospital—labor hours, units of supply, dollars of purchased services, or lease of equipment. As a result, operating cost per unit of output is the most reasonable approach to measuring the value of all inputs to producing a unit of patient care. Total operating expense per occupied bed, operating expense per adjusted occupied bed, operating expense per discharge, or operating expense per adjusted discharge are all examples of ratios used to evaluate the costs per unit of production based on the different units of production described earlier. The following example calculates operating expense per adjusted occupied bed; this same formula can also be used to calculate different units of measure: img The operating expense per adjusted occupied bed for Example Hospital is calculated by using the following values from Table 7–2 img The value for operating expense per discharge is: img Operating expense per adjusted discharge amounts to: img Operating expense per occupied bed equals: img Another perspective on the unit of output in a health care organization is to address the multiple services provided by a hospital in revenues rather than in units such as discharges or patient days. If the hospital units of output are widely varied in terms of sophistication or type of delivery (such as in a hospital that has inpatient services but also operates a skilled nursing unit or an ambulance service, then revenues may be a more appropriate overall measure of output), calculating net revenue per FTE can tell an operations manager the amount of net revenue that was created on average by each employee in the organization. Using values from the operating statement for Example Hospital in Table 7–2, the net revenue per FTE is calculated as: img On average, each employee at Example Hospital in the past year completed work that resulted in net revenues for the organization of $175,003.44. There are several other relationships that an operations manager can evaluate in assessment of operational productivity in a hospital or other health care facility. However, a critical part of using operational metrics is to compare those calculated values to industry benchmarks or trending calculated values of these metrics over time to determine if changes over time show improvement or decline in operational performance. The metrics described here can be compared to benchmarks established by health care industry organizations such as the American Hospital Association or the Healthcare Financial Management Association. Use of industry benchmarks can be valuable for measuring how an organization compares with other organizations, but they should be used with caution. Each hospital will vary based on local labor markets, availability of resources, the payment resources of patients in the service area, and the general priorities and values of the organization’s management and governing body. OTHER OPERATIONAL METRICS The operational metrics described so far focus primarily on the production efficiency of a health care organization, evaluating the number of inputs per output produced. Depending on the organization’s strategic objectives, other metrics not described here may be considered as—or more—important. A key step in monitoring the correct operational metrics for an organization is to establish organizational goals and then link the metrics to outcomes that support such goals. For example, an organization may be performing poorly on clinical goals used to determine payment rates (such as the Value-Based Purchasing Program or “pay for performance” under the Patient Protection and Affordable Care Act of 2010). To improve performance toward those clinical care quality objectives, the organization may establish a patient safety goal and management may decide to adopt monitoring of medication transcription accuracy as a way to reduce patient medication errors. Another example would be comparison of patient treatment records against an established care plan to determine compliance with evidence-based treatment guidelines. Such an approach may be useful in organizations that incur financial losses on patients whose care is reimbursed on a prospective payment basis. Establishing a baseline treatment plan under which the organization can keep costs below reimbursed amounts, as well as monitoring compliance with the plan, can help lead the organization to improved financial results. In this way, the organization has set an overall objective and identified a metric or multiple metrics to measure performance that supports achievement of the objective. Such metrics may not have published industry standard benchmarks, but nonetheless they have value in driving organizational performance improvement. In this situation, the organization must develop its own benchmarks or at least a baseline level for use in monitoring performance. Developing a baseline level of performance for operational metric evaluation is a multistep process: 1. Define the measurement and specific data elements to be used in calculation of the metric. 2. Establish any inclusion or exclusion criteria for data used in developing a baseline (such as excluding patients with a low hemoglobin value from counting compliance with an antithrombolytic medication guideline for patients seen in the emergency room with a suspected heart attack). 3. Define the data-gathering methodology not only for the baseline but for ongoing monitoring (such as manual chart reviews or ad hoc data queries from an electronic medical record database). 4. Establish the desired outcome to be measured by the metric, such as improving accuracy in assessing a patient’s medical history to establish the presence of a community-acquired infection (which can defend against an insurer denial of payment for a suspected hospital-acquired infection). Once these baseline development guides are established and affected parties have had the chance to buy in to the use of selected metrics in managing operational performance, the organization must gather historical data and calculate a baseline level for the metric of interest. A helpful guideline in terms of the amount of historical data to use is at least the number of months in a typical operating cycle for the organization so that seasonal variations in volume, resource availability, or other external influences on performance can be taken into account. This usually translates into a minimum of 3 to 6 months but could be as long as a year if necessary to fully account for seasonal variation (as in areas where patient census fluctuates widely due to normal phenomena such as seasonal migration of retirees). If the data have not previously been collected in the organization, it is essential that someone other than the primary data gatherer validate the data to assure accuracy in and relevance to the baseline establishment process. Once a baseline is established, it can be used by the organization to evaluate ongoing performance with the metric. However, the baseline should be validated after a few months of use to be sure that it is relevant to actual practice in the organization (usually after 3 to 6 months) and then on a routine basis thereafter. A part of the validation process must be to track not only the chosen metric but also performance against the organizational goal to verify that the association between the selected metric and goal achievement remains reasonable. It would not make sense for an organization to track performance on an operational metric that did not lead to the desired overall result. Assuming that the metric tracks with desired organizational outcomes, it must be integrated into the organization’s routine management reporting structure and managers responsible for performance on that metric identified. Managers whose performance is measured using a new operational metric must be able to participate in development of the metric, calculate the baseline measurement, and, above all, influence performance on the metric. It is an ineffective use of organizational resources to measure performance on a metric that managers cannot influence. Moreover, holding managers accountable for performance on a metric that they cannot influence will lead to frustration, burnout, and loss of management talent to the organization. USING OPERATIONAL METRICS Multiple levels of managers within a health care organization can use operational metrics. However, the perspective on which metrics will be used is based on the manager’s role and responsibility within the organization. In fact, the adoption of operational metrics represents a strategic decision for the organization, which must consider how the metrics will be used to manage the organization. Also, the availability of data for calculating these metrics should be considered before a management approach using these operational metrics is adopted. Finally, the organization’s priorities toward financial performance, operational efficiency, or measurement of quality outcomes must be considered in developing the operational metrics used by management. The number of metrics used by the organization should be manageable without the devotion of significant additional resources to calculating metrics or preparing routine reports on them (Ronen & Pliskin, 2006). The use of operational metrics should improve efficiency in the organization, not create a need for additional resources that do not add to the production of patient care services. Therefore, managers must balance the need for detailed evaluation of operational performance and the available resources to report on and assist in monitoring the metrics. One guideline is to use between 5 and 12 metrics in an average-size organization and no more than 20 in a large organization or multifacility system. Reporting on the metrics should occur as frequently as is practical considering the caveat about devoting additional resources to reporting on operational metrics. Again, additional resources that are not devoted to the production of patient care outputs should be weighed against the value of detailed monitoring and reporting of operational metrics. Generally speaking, reports on operational metrics should be prepared with the same frequency as routine financial reports in the organization (Langabeer, 2009). If reports are presented to management on monthly, quarterly, and yearly bases, then reports on operational metrics should be prepared in the same time frames. The only exception to this rule would be if management felt it necessary to monitor certain high-risk or high-priority metrics on a daily basis during a time of challenging financial results. Examples of such daily monitoring metrics would be average length of stay, occupancy percentage, average daily census, adjusted average daily census, and FTE/AOB. These metrics provide a valuable overview of the organization’s production efficiency, and improvement on these metrics over time should lead to enhanced financial performance.

What are some sources for benchmarks of the operational metrics described in this chapter? Why are they important, and what are the limitations to their use?

Answer 1

The some sources of operational benchmarks can be the following:-

• An industry standard for operational metrics.
• Choosing the most desired value or the best value among all the competitors for the particular benchmark as published in annual reports. For example, if we have the annual reports of say 5 competitors and the cost per unit is revealed in those annual reports, then it is useful to consider a benchmark for the cost per unit which is the lowest among all the competitors.
• Setting best performance of the organization across the years as a benchmark.
• An aspirational target for a parameter which can be used to drive improvement can be set as a benchmark.

The benchmarks are important as they provide an information about the current state of business and operational effectiveness. A variance between the actual value of the metric and benchmark provide an indicator towards the extent of improvement that is required to achieve a desired state of operation. It also provides a quantitative target to aspire toward and achieve for maintaining organizational growth.

The limitations to the use of benchmarks are following:

• It is only a quantitative number and does not reveal qualitative issues.
• Setting benchmark at an appropriate level is very important. If benchmark is kept lower, the target is easily achievable and desired improvement is not achieved. If the benchmark is unrealistic, then the workforce may get demotivated and the desired state is not achieved.
• Industry data may not be available all the time and circumstances vary from company to company so a common benchmark may not work all the time.