Designing Meaningful Industry Metrics for Clinical Productivity for Anesthesiology Departments

Abstract

The development of meaningful measures for the comparison of clinical productivity is important for the strategic planning and business operations of both private and academic anesthesiology groups. For academic groups, university administration (i.e., deans and hospital administrators) provides an additional impetus to create a system of metrics by increasingly demanding that all departments, including anesthesiology departments, be able to measure productivity and compare it with industry standards, a process often called “benchmarking”(1). Although educational, research, and administrative productivity may be measured similarly in anesthesiology and in nonanesthesiology specialties, measuring clinical productivity in the discipline of anesthesiology, either for individuals or departments, poses unique challenges because of the different billing system (i.e., base units plus time), the need to staff operating rooms (OR sites) independent of workload, differences in concurrency, and nonanesthesia factors (e.g., duration of surgery, type of surgery) (2). Further complicating the quantification of clinical productivity is the fact that individual and departmental productivity may require different measures. Because of the multiple anesthesiologist-dependent and anesthesiologist-independent factors, we reported in a previous study that “clinical days worked” is the most meaningful measurement of individual productivity (2). For departmental productivity, measurements other than clinical days worked may be more informative. Some anesthesiology groups may have already developed group-specific productivity measurements (i.e., key indicators of business operations), but these systems may not easily be applied to other anesthesiology groups. Currently, the only industry measurements available are “units,” or total American Society of Anesthesiologists (ASA) units (tASA) per full-time equivalent (FTE) (3,4). However, because of concurrency differences between groups (i.e., anesthesia care delivered personally by anesthesiologists versus direction of variable numbers of residents, nurse anesthetists, or anesthesia assistants), measuring units per FTE is insufficient to permit meaningful comparisons. The purpose of this article is to demonstrate the collection of data and calculation of clinical productivity measurements to compare two different anesthesiology groups—an academic department using an anesthesia care team model and a private, medical doctor-only group providing personally performed anesthesia care. In addition to the current industry standard (total units per FTE), new measurements based on “per OR site” and “per case” were evaluated. Information about clinical activity and billing data was collected from January 1, 1999, to December 31, 1999, for one academic anesthesiology department (Group A) and one private practice anesthesiology group (Group P). Total ASA units billed, time units (TU) billed, number of cases, average daily number of anesthesiologists who staffed the OR (OR FTE), and average daily number of OR sites staffed were collected (Table 1). All anesthesia care billed with ASA units was included except for obstetric care. Pain management and critical care were excluded because these services are billed with relative value resource-based system units. Any clinical activity not billed (e.g., preoperative outpatient assessment clinic) was also excluded. The productivity measures that were examined are listed and defined in Table 2.Table 1: Data Provided for Productivity MeasuresTable 2: Departmental Productivity MeasuresResults Group A is an academic department that uses the care team model to provide anesthesia care; therefore, the number of OR FTEs is less than the number of OR sites (Table 3). In contrast, Group P is a medical doctor-only group that provides only personally performed anesthesia care; therefore, the number of OR FTEs equals the number of OR sites. This difference is apparent in the quantification of concurrency for each group (1.57 for Group A and 1.00 for Group P) (Table 4).Table 3: Data Collected for MeasurementsTable 4: Productivity MeasuresBy using tASA/OR FTE as the measure of productivity, Group P was only 88% as productive as Group A (Table 4). However, when the difference in concurrency is accounted for in the measurement of tASA/OR site, Group P was more productive (139% of Group A). Similarly, TU/OR FTE comparison shows that Group A billed for more hours than Group P, but, after correcting for concurrency, Group A and Group P had almost identical TU/OR site values (i.e., they worked similar numbers of actual hours). Measurement of cases per OR site illustrates that Group P provided anesthesia for 78% more cases in the same amount of time as Group A. Furthermore, for each case, Group P’s surgical duration was only slightly more than one-half that of Group A (6.44 TU per case for Group P and 11.26 TU per case for Group A). The shorter surgical times account for a larger tASA per hour measurement (Group P is 136% more than Group A). This is probably not attributable to more difficult cases being performed in academic settings, because Group P’s base units per case are slightly larger than Group A’s, suggesting a similar complexity of surgical procedures. Discussion In this report we illustrate the collection, calculation, and use of clinical productivity for comparison of two different anesthesiology groups. We found that “per OR site” measurements removed the confounding factors of different concurrencies. Additional measurements (e.g., tASA/h) allowed for a more informative comparison in OR management issues. Our goal was to propose clinical productivity measurements that met two conditions: (a) use of raw data that would be available without requiring an individual anesthesiology group to invest heavily in personnel or other resources and (b) facilitation of meaningful and accurate comparisons between different anesthesiology groups. To meet the first condition, we focused on billing—cases, charges and revenue, base units, and TUs—because this information should be available to each anesthesiology group. We then discarded charges and revenue because these reflect factors that are not dependent on the efforts of individual clinicians. We then chose numbers of cases, tASA, and TU as possible numerators of measurements. For denominators, we chose average daily FTEs and OR sites. To obtain those numbers, we developed a method of surveying one day a month to estimate the daily numbers. By choosing these types of data to collect, we excluded all clinical services (specifically critical care and pain management) that are billed without using ASA units and other activities (such as preoperative evaluation and contract hospitals) for which patient-specific services are not billed. This was done because, although academic departments usually provide these services, many private practice groups do not. If included, the comparisons might otherwise not be accurate. In addition, on the basis of previous experience, we excluded obstetric anesthesia, which is difficult to compare with OR care because of a difference in the methods of billing TU and the highly variable approaches used to provide coverage of obstetric anesthesia (2). We also excluded from analysis FTEs who performed obstetric anesthesia billings and services that were not billed by using ASA units. To further simplify data collection, we accepted two additional limitations. First, we chose to not include placement of central venous and arterial catheters because these activities are billed by using relative value units. To include these low-volume activities in the comparison, we would have had to convert ASA units to relative value units. Although methods for that conversion have been published (5), we considered the effort involved unjustified for the purpose of comparing different anesthesiology groups. Perhaps the effort would be justified if the purpose were to compare the productivity of anesthesiologists with that of other specialists. Second, we did not include on-call (night or weekend) FTEs in the estimate of daily FTEs because of the differences between groups in staffing after-hours cases. We therefore included all ASA units as if they were billed during the weekday schedule. This approach is further justified because the resultant increase in the TU/OR site measurement recognizes the effort of performing cases after hours, on weekends, or both. The second condition for productivity measurements was to allow comparison between different anesthesiology groups. For comparisons between groups or for establishing benchmarks against which groups can compare their performance, measurements different from those used by groups to monitor the productivity of individual anesthesiologists may be useful. Two major factors that differ between anesthesiology groups—concurrency and duration of surgery—influence billed ASA units. Although Posner and Freund (6) have shown that tASAs increase as concurrency increases, the only industry-wide measure currently available (at the time of this study) is “total units per FTE,” a metric that does not correct for concurrency (3,4). To correct for the confounding effect of concurrency, we considered two options: either to correct “per OR FTE” measurements to a concurrency of 1:2 (one anesthesiologist to two OR sites) or to correct “per OR FTE” measurements to a concurrency of 1:1 (one anesthesiologist to one OR site). Arithmetically, the first correction simply equals twice the second correction. The second correction is the same as performing calculations “per OR site,” i.e., dividing the productivity data by the number of OR sites. We chose to use the “per OR site” calculation, because it seemed to be more useful for OR management evaluation. In reviewing our results, we found that the current industry standard, “total units per FTE,” did not meet the second condition of allowing meaningful comparison between different groups. We found that a set of measurements based on “per OR site” and “per case” provided more precise comparisons between different groups. In addition to data quantifying tASA units, data describing TU and cases performed allowed for additional measurements that provided information about the duration of surgery and the length of the workday. Base units per case were used to determine whether the productivity differences (in particular, the presence of shorter surgeries) were more likely related to faster or less complicated surgeries. Because the base units were actually larger in the private practice group, this measurement suggests that the larger number of cases per OR site and tASA per OR site for Group P, despite use of the same amount of anesthesia time (TU per OR site), is explained by faster surgeries rather than by less complicated surgeries. Because Group P provides anesthesia care for private practice surgeons, the finding that the private-practice surgeons are faster than faculty-supervised surgical residents is not surprising. However, the results illustrate how the duration of surgery is an important determinant of anesthesia productivity and billable charges. The tASA per hour of care measurement illustrates this difference. For every hour of anesthesia care, Group P’s bills were 36% more than those of Group A. The next step after designing these measures is to test them in a larger number of groups with the purpose of determining whether the measures continue to provide meaningful comparisons. We have already begun this process, and, if it is successful, the next step will be to consider these productivity methodologies and metrics for use as an industry-wide standard for clinical productivity in anesthesiology. This would permit an industry-wide survey and reporting to allow anesthesiology groups and departments to benchmark their clinical activities.