Technical Staffing Ratios: A Summary of 3 College of American Pathologists Q-Probes Studies

As with earlier Q-Probes technical staffing ratio studies,^7,8  College of American Pathologists Q-Probes program subscribers submitted, in the standard Q-Probes study format, data on the levels of staffing and test volumes in their laboratories for studies performed in 2014, 2016, and 2019.³  From these data, we calculated productivity staffing ratios, defined as testing volumes per full-time equivalent (FTE), and the degree of managerial oversight, defined as the ratio of nonmanagement to management FTEs for various laboratory divisions. In order to investigate conditions that might influence productivity, participants completed a general questionnaire surveying their hospital and laboratory demographics and practices. Within each study year we performed statistical analyses to determine which demographic and relevant practice characteristics were significantly associated with technical staffing ratios.

With the exceptions of the laboratory division of molecular pathology, which we excluded from the 2019 study, and “overall laboratory,” which we included only in the 2019 study, the 3 studies examined staffing ratios in all laboratory divisions listed below. To ensure comparability of results, we adopted a number of standard definitions as follows.

We defined individual sections or divisions of the laboratory for each study. The studies included the following sections in at least 1 of the 3 study years.

A section of the laboratory that performed one or more of the following: gynecologic cytopathology, nongynecologic cytopathology, surgical pathology, histochemistry, immunohistochemistry, or autopsy pathology. The studies did not include test counts or staff related to electron microscopy, flow cytometry, peripheral blood smear evaluation, or pathology assistant staff time.

A section (sections) of the laboratory that included all routine and special testing in chemistry, hematology, coagulation, urinalysis, serology, and immunology, whether performed in the central (main) laboratory or an off-site laboratory. We included blood gas testing and emergency department stat laboratory testing only if performed by laboratory staff. We also included in this section tests and staff associated with central specimen processing areas or send-out areas. We did not include staff time devoted to athletic drug testing, heavy metal analysis, quantitative sweat analysis, unbilled manual smear reviews, toxicologic confirmatory testing, or any calculated tests.

A section of the laboratory that performed one or more of the following: aerobic and anaerobic cultures, antimicrobial susceptibility testing, mycobacterial cultures, mycology cultures, and parasitology. Food and Drug Administration–approved tests for DNA detection of microorganisms such as bacteria, viruses, fungi, and tuberculi were included in the microbiology section. The studies did not include serologic testing, virus cultures, or rapid tests for viruses.

A section of the laboratory that performed one or more of the following: blood type/ABO, type and screens, crossmatch, atypical antibody identification, or holding and dispensing of blood products. The studies did not include the following in this section: test counts or staff related to the collection of autologous blood, collection of homologous blood, and any testing associated with unit collection such as viral marker testing, tissue bank functions (bone, skin, etc), or on-site pheresis (therapeutic or other).

A section or part of another section of the laboratory that performed one or more of the following: in situ hybridization, fluorescence in situ hybridization, nucleic acid sequencing, polymerase chain reaction, nucleic acid hybridization, RNA amplification, and similar technologies. The studies only included Food and Drug Administration–approved tests that had a Current Procedural Terminology (CPT) code. If a molecular test was performed in another section, participants were instructed to count staff associated with the molecular testing in this section. This included molecular testing performed in the anatomic pathology section on paraffin-embedded material and cytologic material. The studies did not include test counts or staff related to inherited disease, karyotyping, human leukocyte antigen typing, forensics or parentage applications, or staff used to perform laboratory-developed tests. The molecular pathology section was not studied in 2019.

Refers to the sum of all laboratory departments within an institution, in addition to the 5 sections defined above.

Staff who spent at least 50% of time supervising activities of others or performing other managerial duties.

Staff who spent less than 50% of time supervising others, including technologists or technicians, cytotechnologists, or histotechnologists primarily performing bench testing or procedures in addition to some nontechnical tasks.

All pathologist and nonpathologist physicians and individuals with PhD or other doctorate degrees who may or may not supervise activities of others. These studies included the tabulation of time spent by doctoral level staff supervising activities of others. We excluded time spent in research, academic tasks, or in developing esoteric tests.

Laboratory staff FTE categories considered subsets of the 3 major laboratory cost center categories (management, nonmanagement, and doctoral staff).

Laboratorians whose work was dedicated to quality assurance activities, which may have included monitoring of laboratory processes, equipment, and materials to meet required standards; review of data quality and performance; and ensuring safety and compliance with rules and regulations.

Staff who spent their time preparing samples, accompanying paperwork, and computer entries for testing done at an off-site laboratory (eg, regional, specialty, or centralized health care laboratories), and may also have received and recorded results from these off-site laboratories.

We tabulated all staff (ie, physicians, technologist, etc) hours as FTEs, where 1 FTE = 40 h/wk (2080 h/y) of paid activity assigned to a laboratory cost center. If an institution considered 1 FTE = 37.5 h/wk, we had participants normalize their time to 40 h/wk. We did not tabulate paid disability time. When staff members served in both management and technical or nontechnical roles, we had participants divide their time between the applicable categories. For example, if a full-time chemistry supervisor spent 25% of the time performing laboratory tests, we had participants report the supervisor's time as 0.75 management FTE and 0.25 nonmanagement FTE in the chemistry/hematology/immunology section. We excluded from our tabulations time that staff were engaged in externally funded research and training for medical technologists, histotechnologists, or cytotechnologists. If staff worked in more than one laboratory section (ie, their work hours were split between anatomic pathology, chemistry/hematology, blood bank, and/or microbiology), we tabulated their time according to the amount they spent in each laboratory section. We assigned to one of the laboratory sections all of the activity of shared staff on all shifts (except for time related to externally funded research, time teaching in a training program, time spent in nontesting sections, and time that was specifically excluded from section totals per instructions). This study did not include staff time associated with working with laboratory computer systems, phlebotomy (blood collection), point-of-care testing, send-out tests; in nontechnical tasks such as couriers, clerical, transcription, marketing, or billing operations; in academic research; in developing esoteric tests; and in training programs for medical technologists/technicians, histotechnologists, or cytotechnologists.

Instructions for counting tests differed for the various laboratory sections. We counted all billable tests on all shifts attributable to a laboratory section unless there were specific instructions to exclude a type of testing or to consider that testing to be part of another section of the laboratory. We included tests performed by trainees under supervision for the benefit of patients. Regardless of the laboratory section, we excluded send-out tests (tests referred for analysis outside an individual's institution); point-of-care testing; and billable tests associated with phlebotomy, cytogenetics, histocompatibility testing, virology tests, academic research, and training programs for technologists/technicians, histotechnologists, or cytotechnologists.

In all 3 studies, we had participants record the numbers of cytology accessions and histology blocks for each most recently completed fiscal year. Billable tests in histology for all 3 studies included surgical pathology, histochemistry, immunohistochemistry, and autopsy pathology test counts. We used in our calculations of current billable tests for each year the CPT codes listed in the kit instructions.

We instructed participants to unbundle chemistry profiles when counting tests in the chemistry/hematology/immunology section. For instance, an electrolyte panel consisting of 5 analytes was counted as 5 tests. We included in this section test counts associated with serology and urinalysis even if the institutions normally considered these sections part of microbiology or a different section of the laboratory, tests associated with laboratory central accessioning and/or processing areas or send-out areas, and tests associated with off-site or ambulatory care laboratories.

We included all routine bacteriology, mycobacteriology, mycology, and parasitology billable tests.

Participants recorded the number of crossmatches and type and screens for the most recently complete fiscal year, including electronic crossmatches (CPT code 86923), atypical antibody detection workups (CPT codes 86860, 86870, and 86880), and transfusion of blood product units (packed red blood cells, fresh frozen plasma, platelets, and cryoprecipitate).

Participants included CPT codes for billable molecular tests 87470 through 87999, 81200 through 81355, 88365, and 88120 in 2016, and 87470 through 87801, 81206 through 81208, 81210, 81235, 81270, 81275, 81292 through 81301, 88120 through 88121, and 88365 through 88368 in 2014. We instructed participants to count staff performing molecular pathology in this category even if in their institutions they performed molecular pathology testing in another department.

We instructed participants to record the total number of billable tests performed by the entire laboratory comprising all divisions, excluding send-out tests.

We included on-site and remote (limited function) laboratories and assigned testing their testing volumes and staff (except for phlebotomy staff) to one of the laboratory sections defined above. We did not include point-of-care testing as a testing location.

From these data, we calculated productivity for each section. We defined productivity for chemistry/hematology/immunology, microbiology, and the overall laboratory as the ratio of the number of billable tests per nonmanagement FTE. We calculated 2 labor productivity ratios for the anatomic pathology section: the number of tissue blocks per histology nonmanagement FTE and the number of cytology accessions per cytology nonmanagement FTE. We calculated 2 productivity ratios for the transfusion medicine section: the number of blood product units transfused per nonmanagement FTE and the number of crossmatches and/or type and screens performed per nonmanagement FTE. We replaced the term management span of control used in previous technical staffing ratio studies with degree of managerial oversight, which we defined as the ratio of nonmanagement FTE to management FTE for a given section or the overall laboratory. We determined the degree of managerial oversight for histology and cytology combined. We defined nonacute care patient testing as laboratory testing for patients who were not admitted to an acute care hospital or emergency department.

In each study, we established peer groups to allow study participants to compare their performance in the 5 testing departments with participating institutions with similar workloads. We established these peer groups by determining testing volume breakpoints, which we computed using an average linkage clustering algorithm that defines groups that minimize variances within clusters using the average distance between pairs of observations. We then defined testing volume peer group breakpoints by applying test volume thresholds that separated the clusters while maintaining adequate peer group sample sizes. We summarized results by peer group only if we detected statistically significant associations between test volumes and their corresponding labor productivity measurements.

In order to investigate conditions that might influence productivity, participants completed a general questionnaire surveying their hospital and laboratory demographics and practices. The list of inquiries concerning practice characteristics included laboratory testing hours, scope of hospital services (eg, transplants, intensive care services, teaching hospital), the contribution of test volumes from geographically separate testing facilities, and approximations of the ratio of nonacute to acute care patients.

Within each study year we performed statistical analyses to determine which independent demographic and relevant practice characteristics were significantly associated with technical staffing ratios using nonparametric Kruskal-Wallis tests. We considered a P value less than .05 to be statistically significant. We used Kruskal-Wallis tests to test for overall differences in staffing ratio distributions across the 2014, 2016, and 2019 studies. We applied a Bonferroni adjustment when reporting significant associations between staffing ratios and general practice questions that allowed more than 4 multiple responses. We performed all analyses with SAS 9.4 (SAS Institute, Cary, North Carolina).

In 2019, a total of 67 institutions submitted data during the study period. This total did not include 2 participants that we excluded from staffing ratio analyses because they submitted insufficient data. In 2016, a total of 82 participants submitted sufficient data during the study period. In 2014, a total of 79 participants submitted data during the study period. These totals represent the number of institutions that submitted data for at least one laboratory section studied in each year. Because not all study participants submitted complete data for every laboratory section, laboratory demographic variable, or practice characteristic variable, the sample sizes among the frequency distribution summaries differed from the total numbers of study participants for each year.

Table 1 summarizes the key institution demographic characteristics among participants of the 3 studies. In the 2019 study, 59 of 69 total participants (85.5%) were from the United States. The remaining participants were from Saudi Arabia (4), Bahrain (1), Brazil (1), Canada (1), Israel (1), Mexico (1) and the United Arab Emirates (1). Most participating institutions from all 3 study years were nongovernmental institutions (2019, 52 of 64 [81.3%]; 2016, 72 of 82 [87.8%]; 2014, 64 of 78 [82.1%]). The licensed bed size of most institutions was fewer than 300 beds (2019, 35 of 59 [59.3%]; 2016, 51 of 77 [66.2%]; 2014, 50 of 76 [65.8%]). Tables 2 and 3 summarize the key practice characteristics among participants of the 3 studies.

Table 4 summarizes the institution productivity and managerial oversight ratios percentile distributions for the 3 studies. Cytology accessions per cytology nonmanagement FTE productivity ratio distributions decreased significantly from 2014 to 2019 (P = .03), but not between 2016 and 2019 alone (P = .14).

In each study, we observed the degree of managerial oversight to vary among laboratory sections, but these did not vary significantly during the 3 years during which these studies were performed. Other than cytology, no productivity or managerial oversight ratios trended significantly across the 3 studies.

Table 5 summarizes the unique overall laboratory staffing ratios that we studied in 2019, including overall managerial oversight ratios, the ratios of quality assurance FTEs per 100 laboratory staff FTEs, and the numbers of FTEs preparing send-out tests per 100 overall laboratory nonmanagement FTEs. Ten percent or more of participants (10th percentile) indicated that they employed no staff designated exclusively to perform quality assurance or send-out functions. Among the 62 participants, the ratio of overall laboratory managerial oversight spanned almost fourfold between the 10th and 90th percentiles (5.1 versus 19.6 for 10th and 90th percentiles, respectively).

The Figure, A through E, displays grouped box plots of 2019 productivity staffing ratios by test volume peer group. In most laboratory departments in all 3 studies, median productivity increased with increasing testing volumes. The one exception was for the microbiology section studied in 2019. We were unable to create volume-based peer groups for microbiology in 2019 because the variability observed within microbiology billable tests was too high. There was no relationship between peer groups for the different laboratory sections; for example, peer group 1 for cytology did not necessarily contain the same laboratories as peer group 1 for chemistry/hematology/immunology.

Table 6 summarizes the practice characteristics that we examined in the 2019 study and that were significantly associated with overall laboratory staffing ratios. Institutions comprising fewer geographically separate hospitals or that operated off-campus laboratories that submitted specimens to their main laboratories for testing tended to have higher ratios of FTEs preparing send-out tests per 100 overall laboratory nonmanagement FTEs (P = .004) than did institutions with more additional laboratories off campus. In addition, institutions operating dedicated molecular pathology sections tended to have lower ratios of FTEs preparing send-out tests per 100 overall laboratories nonmanagement FTEs (P = .04) than institutions that did not operate dedicated molecular pathology sections. Lastly, institutions that used the method of budget allotment to determine the number of staff needed in their laboratories tended to have higher overall laboratory QA FTEs per 100 laboratory staff FTEs compared with institutions that did not use budget allotments to determine staff needs (P = .005).

Table 7 summarizes the demographic and practice characteristics that were significantly associated with laboratory department productivity or managerial oversight ratios across the 3 study years. In all 3 studies, institutions offering transplant services transfused significantly more units of blood per nonmanagement FTE (ie, showed greater productivity) than institutions that did not support a facility that performs transplants. Productivity associations were otherwise inconsistent across the 3 studies. Demographic and practice characteristics about which we inquired that are not listed in this table did not have significant associations with staffing and managerial oversight ratios.

These 3 laboratory technical staffing ratio Q-Probes studies performed during 5 years were designed to measure the normative rates for both laboratory productivity and the degree of laboratory managerial oversight in various laboratory testing and laboratory support departments.

Productivity varied widely among the various laboratory departments within each institution and among different institutions participating in this study. In all testing departments, productivity was driven by scale: the greater the testing volume, the higher the productivity ratio or number of tests performed by a single FTE in a given year. Demographic characteristics that were associated with greater productivity—urban locations, large bed capacities, support of transplant services and teaching programs—may have been markers of greater institutional complexity requiring greater magnitudes of laboratory testing. It is unlikely that laboratory managers can alter institutional demographics to improve their productivity. However, they may be able to alter practice characteristics such as maintaining dedicated molecular diagnostics to improve productivity (see Table 6).

The degree of managerial oversight varied among departments and institutions. The few associations between degree of managerial oversight and productivity were not consistent across the 3 years. For instance, in 2014 the degree of managerial oversight tended to be lower in anatomic pathology departments supporting teaching hospitals compared with anatomic pathology departments not supporting teaching hospitals. However, microbiology departments supporting teaching hospitals showed a higher degree of managerial oversight among laboratories supporting a teaching hospital compared with those that did not. Clearly, there were other conditions influencing managerial oversight, but that we did not investigate.

With the exception of cytology productivity ratios, the ratios of productivity and managerial oversight were unchanged in the 5 years during which we conducted these studies. The lowering of cytology productivity over time may reflect the need for cytotechnologists to support the expanding role of rapid on-site evaluations and of human papillomavirus testing in clinical diagnostics.¹²  We cannot comment on the lack of change in the other laboratory testing departments. Q-Probes Studies are cross-sectional and essentially snapshots of practice at specific times and as such do not collect data on the effects that evolving testing conditions, technologies, demographics, economic conditions, and pandemics might have on efficiency within individual laboratories.

Productivity is not synonymous with quality. For instance, use of budget allotments to determine laboratory quality assurance staffing requirements in some institutions was associated with fewer employed quality assurance FTEs per 100 total laboratory FTEs than nonuse of budget allotments (Table 6). In the 2019 study, at least 10% of participating institutions employed no FTEs to oversee quality assurance (Table 5). We did not collect data to determine whether using budget allotments to ascertain staffing needs or the absence of employees performing quality control functions improved or undermined laboratory quality indicators.

Similarly, having pathologists participate directly in staff hiring decisions was associated with lower productivity ratios in the anatomic pathology departments in 2019. Pathologist medical directors, who are required by regulatory bodies to ensure laboratory quality,⁵  may determine that staff must be added to their operations in order to maintain or improve the level of quality. Adding staff to a laboratory may be at the expense of productivity, as shown by this study. Again, we did not collect data to determine the effects of pathologists' involvement in staffing decisions on laboratory quality indicators.

The percentile distributions of productivity measures we are reporting—staffing ratios—are not benchmarks; they are normative rates. Previous authors noted that because of the variability among institutional and laboratory environments, formulaic approaches to laboratory staffing are inadequate.⁷  Characteristics such as the diversity of patient needs, technologies, degree of laboratory automation, turnaround time requirements, demands of the medical communities, availability of labor, and testing requirements preclude comparison of efficiencies among different testing divisions in the same laboratory or among the same testing divisions in different laboratories. Perhaps that is why we were unable to associate most of the actionable practice characteristics about which we inquired with productivity ratios.

Laboratories may find value in defining their own measures of staffing ratios and using them as starting points from which to track their success in implementing activities that improve productivity. However, efforts to optimize staffing ratios should never end, regardless of where a laboratory's ratio measurement resides as a normative rate. Further, we believe that laboratory managers should not engage in activities to improve productivity without simultaneously evaluating the effects those activities have on laboratory quality, examples of which have been provided elsewhere.¹³