Context.—Publicly available information concerning laboratory staffing benchmarks is scarce. One of the few publications on this topic summarized the findings of a Q-Probes study performed in 2004. This publication reports a similar survey with data collected in 2010.

Objective.—To assess the relationship between staffing levels in specified laboratory sections and test volumes in these sections and quantify management span of control.

Design.—The study defined 4 laboratory sections: anatomic pathology (including cytology), chemistry/hematology/immunology, microbiology, and transfusion medicine. It divided staff into 3 categories: management, nonmanagement (operational or bench staff), and doctoral (MD, PhD) supervisory staff. People in these categories were tabulated as full-time equivalents and exclusions specified. Tests were counted in uniform formats, specified for each laboratory section, according to Medicare rules for the bundling and unbundling of tests.

Results.—Ninety-eight participating institutions provided data that showed significant associations between test volumes and staffing for all 4 sections. There was wide variation in productivity based on volume. There was no relationship between testing volume per laboratory section and management span of control. Higher productivity in chemistry/hematology/immunology was associated with a higher fraction of tests coming from nonacute care patients. In both the 2004 and 2010 studies, productivity was inseparably linked to test volume.

Conclusions.—Higher test volume was associated with higher productivity ratios in chemistry/hematology/immunology and transfusion medicine sections. The impact of various testing services on productivity is section-specific.

Despite employee pay and benefits representing more than half of laboratory direct costs, publicly available information on staffing benchmarks is hard to find. In 2005, Valenstein et al1 published a summary of a 2004 Q-Probes study of clinical laboratory staffing in 151 institutions. That study divided the laboratory into 4 sections: anatomic pathology (with histology and cytology separately evaluated), microbiology, transfusion medicine (blood bank), and a section that combined chemistry, immunology, and hematology. The study assessed labor productivity and management span of control within the laboratory sections in relation to laboratory test volumes and institutional characteristics.

In 2011, data of this type, valuable to laboratory managers analyzing and planning staff support, remain scarce; therefore, from a similar survey done in the altered circumstances of 2010, we report a study to compare and contrast with the 2004 results. Both studies have the strength of examining both anatomic and clinical pathology laboratory sections in comparable formats. Both focus on the laboratories' basic operations that support clinical medicine: anatomic pathology, chemistry/immunology and hematology, microbiology, and transfusion medicine (blood bank). Both exclude potentially distracting specialty (boutique) laboratory operations, such as athlete drug testing and forensic toxicology, assays for heavy metals, and quantitative sweat analyses. The 2 Q-Probes studies also excluded testing in which nonlaboratory personnel figured as test operators, most prominently in point-of-care testing. The objective in doing so is to get to the analysis of staff support for clinical laboratories' central clinical mission.

The study that we report here, and compare and contrast to its predecessor, benefits from experience accrued in another College of American Pathologists (CAP) program that collected information on laboratories' economic characteristics. The Laboratory Management Improvement Program (LMIP) developed practical definitions of comparable laboratory testing within sections that are key to developing comparable information about various staff positions (line, management, technical, and professional).2 The LMIP also worked out practical, yet highly reproducible, definitions of tests themselves that could be applied across different institutions. On this basis, we could count staff and the tests that they perform in a remarkably uniform way to ensure comparability of results.

MATERIALS AND METHODS

As with the 2004 study,1 CAP Q-Probes program subscribers submitted data in the standard Q-Probes study format for the 2010 study.3 For the recent study, CAP Q-Probes subscribers submitted data in early 2010.

Definitions, inclusions, and exclusions were as follows.

  • a.

    Laboratory sections: 4 sections of the laboratory were defined for this study.

    1. Anatomic pathology. A section of the laboratory that performs 1 or more of the following activities: gynecologic cytopathology, nongynecologic cytopathology, surgical pathology, histochemistry, immunohistochemistry, or autopsy pathology. Not included were test counts or staff related to electron microscopy, flow cytometry, or peripheral blood smear evaluation.

    2. Chemistry/hematology/immunology. A section of the laboratory that includes routine and special testing in chemistry, hematology, coagulation, urinalysis, serology, and immunology, whether done in the central (main) laboratory or an off-site laboratory. Staff members associated with a central specimen processing area or send-out area were included in this section. Send-out test counts and staff time devoted to phlebotomy were not included in the study. Staff or test counts related to the following were also not included: athlete drug testing, heavy metal analysis, quantitative sweat analysis, point-of-care tests, manual smear reviews (unless these reviews were billed separately), toxicologic confirmatory testing, and any calculated tests (results produced from ratios of other, directly measured tests).

    3. Microbiology. A section of the laboratory that performs 1 or more of the following tests: aerobic and anaerobic cultures, antimicrobial susceptibility testing, mycobacterial culture, mycology cultures, and parasitology. Not included were test counts or staff time devoted to serologic testing, virus cultures, or rapid tests for viruses.

    4. Transfusion medicine. A section of the laboratory that performs one or more of the following procedures: blood type/ABO, atypical antibody identification, or holding and dispensing of blood products. Not included were test counts or staff related to collection of autologous blood, collection of heterologous blood, any testing associated with unit collection, such as viral marker testing, tissue bank functions (bone, skin, etc), or therapeutic or other on-site apheresis.

  • b.

    Staff: Staff were counted in a uniform manner to ensure comparability of results. Each employee assigned to the laboratory cost center was assigned to 1 of the following 3 categories.

    1. Management staff. Staff members who spend more than 50% of time supervising activities of others.

    2. Nonmanagement staff. Staff members who do not spend at least 50% of time supervising others; this category includes medical technologists or technicians and cytotechnologists or histotechnologists who perform primarily “bench” tests or procedures, but may also perform some nontechnical tasks.

    3. Doctoral staff. Staff physicians or other individuals with PhD or other doctorate degree who may or may not supervise activities of others. Time spent by doctoral-level staff supervising activities of others was tabulated as part of this study. Time spent in research, academic time, or time spent developing esoteric tests was not included in this study.

  • c.

    Staff tabulation: Staffing was measured as full-time equivalents (FTEs). One FTE equals 40 hours/week (2080 hours/year) of paid activity assigned to a laboratory cost center. Paid disability time was not tabulated. When staff members serve in both a management and technical or nontechnical role, their time was divided between the applicable categories. The laboratory participants determined the divisions regarding personnel serving in both management and nonmanagement roles. For example, if a supervisor spent 25% of the time performing laboratory tests, the participants were instructed to report that individual as 0.25 nonmanagement FTE and 0.75 management FTE. If staff members were engaged in externally funded research or a training program for medical technologists, histotechnologists, or cytotechnologists, their externally funded research time or time spent in training was not included. When staff members worked in more than 1 laboratory section (ie, shared by anatomic pathology, chemistry/hematology, blood bank, and/or microbiology), their time was distributed according to the amount of time spent in each laboratory section. All of the activity of shared staff (except for time that is specifically excluded from the section totals per instructions) was accounted for and assigned to 1 of the 4 laboratory testing sections. Staff from all shifts was included. Staff devoted to autopsy activities was tabulated along with histology staff.

    Exclusions from staff tabulation: This study did not include staff time associated with working with a laboratory computer system, phlebotomy, point-of-care testing, or nontechnical tasks such as courier, clerical, transcription, marketing, or billing operations. Staff members associated with a central specimen processing section were included. Staff and tests associated with point-of-care/bedside testing were excluded. Nonmanagerial FTEs associated with blood collection (autologous or otherwise) and therapeutic apheresis were excluded from the total number of nonmanagerial FTE counts.

  • d.

    Counting tests: Instructions for counting tests differed for the 4 laboratory sections. All billable tests attributable to a laboratory section were counted, unless there were specific instructions to exclude a type of testing or to consider it part of another section of the laboratory. Regardless of the laboratory section, send-out tests (tests referred for analysis outside an individual's institution) were not counted.

    • 1.

      Total the surgical pathology, histochemistry, immunohistochemistry, and autopsy pathology test counts and record as the total number of billable tests in histology. These tests are included in the following current procedural terminology (CPT) codes: 88300, 88302–88305, 88307, 88309, 88311–88319, 88329, 88331–88332, 88342–88347, and 88355–88399.

    • 2.

      For counting tests in the chemistry/hematology/immunology section, chemistry profiles were unbundled, in conformance with current Medicare billing requirements. Thus, an electrolyte panel consisting of 5 analytes is to be counted as 5 tests. However, components of a hemogram (hemoglobin, white blood cell count, platelet count, etc) were not unbundled, because this test has 1 CPT code and Medicare requires that it be billed only once for each hemogram. The total number of manual differential tests for the most recently completed fiscal year was recorded. Test counts and staff associated with testing performed at off-site laboratories were included. Test counts and staff associated with a serology and urinalysis section were included, even if the institution normally considers this section part of microbiology or some other section of the laboratory. If the laboratory had a central specimen processing section, no separate activity (billable tests) was associated with central specimen processing.

    • 2.

      As stipulated above, send-out tests were not included in the test counts. Billable tests associated with phlebotomy were also excluded.

    • 3.

      Microbiology tests included all routine bacteriology, mycobacteriology, mycology, and parasitology billable tests and staff in this section. The following were not included: virus cultures or rapid tests for viruses, molecular amplification tests, and serology (include serology in the chemistry/hematology/immunology section). Urine culture billable tests in the most recently completed fiscal year were included (CPT codes 87086 and 87088).

    • 4.

      Transfusion medicine testing included the number of crossmatches, type and screens, and atypical antibody detection workups (CPT codes 86860, 86870, and 86880) for the most recently completed fiscal year, but did not include viral marker testing associated with unit testing.

    • 4.

      The total number of packed red blood cells, fresh frozen plasma, platelets, and cryoprecipitate units transfused in the most recently completed fiscal year was tabulated.

  • e.

    Testing locations: On-site and remote (limited function) laboratories were included in this study. On-site and remote testing volume and staff (except for phlebotomy staff) were assigned to 1 of the 4 defined laboratory sections. Point-of-care testing was not counted as a testing location.

Labor productivity was determined for each section. Productivity for chemistry/hematology/immunology and microbiology is defined as the ratio of the number of billable tests per nonmanagement FTE. Two labor productivity ratios were calculated for the anatomic pathology section. The number of tissue blocks per histology nonmanagement FTE and the number of cytology accessions per cytology nonmanagement FTE were calculated. Two productivity ratios were calculated for the transfusion medicine section: (1) the number of blood product units (red blood cells, fresh frozen plasma, platelets, and cryoprecipitate) transfused per nonmanagement FTE; and (2) the number of crossmatches or type and screens performed per nonmanagement FTE. Management span of control is defined as the ratio of nonmanagement FTE per management FTE. Management span of control was determined for histology and cytology combined. Nonacute care patient testing refers to laboratory tests for patients not in an acute care hospital or emergency department.

Statistical Analysis

Labor productivity ratios were calculated as tests per nonmanagement technical staff person, and management span of control was calculated as ratios of nonmanagement FTE per management FTE. Both these ratios were tested for associations with institutions' test volume and demographic variables. There were 6 labor productivity variables and 4 management span of control variables. Staffing ratios were considered the dependent variables.

Each of the plausibly associated variables was tested separately for association with test volume within the section (these volumes were blocks and slides for anatomic pathology, billable tests for chemistry/hematology/immunology, and transfusions, or crossmatch and type and screen procedures, for transfusion medicine). Each variable was also tested separately for association with the demographic variables listed in Tables 1 and 2. Productivity in anatomic pathology was additionally tested for association with specimen complexity (the ratio of CPT 88307 and 88309 procedures to CPT 88305 and 88304 procedures). Productivity in microbiology was additionally tested for association with the extent of mycology, bacteriology, and parasitology testing, and the percentage of bacteriology cultures that were performed on urine. Productivity in chemistry/hematology/immunology was additionally tested for association with the percentage of billable manual peripheral smear reviews and with the extent of chemistry testing.

Table 1.

Institution Demographics, 2004 and 2010 Studies

Institution Demographics, 2004 and 2010 Studies
Institution Demographics, 2004 and 2010 Studies
Table 2.

Other Characteristics of Participating Laboratories, 2004 and 2010 Studies

Other Characteristics of Participating Laboratories, 2004 and 2010 Studies
Other Characteristics of Participating Laboratories, 2004 and 2010 Studies

Nonparametric Wilcoxon rank sum tests and Kruskal-Wallis tests were used to test associations; a P value < .05 was considered to be statistically significant. Because the testing volume of a laboratory was so closely associated with productivity, laboratory test volume was explicitly introduced as a covariate when evaluating the relationship of other variables to labor productivity and management span of control. Only variables that were independent predictors of staffing at the significance threshold (P < .05) were considered to show a significant association.

Note: Several participating institutions did not answer all of the questions on the questionnaire or operated laboratories that did not contain all 4 laboratory sections. These institutions were excluded only from tabulations and analyses that required the missing data elements.

RESULTS

Participant Characteristics

Ninety-eight institutions submitted staffing data and test counts. Most of the institutions (94%) were from the United States, with 2 from Canada, 2 from Saudi Arabia, 1 from Bermuda, and 1 from Singapore. Of participating institutions, 39.8% were teaching hospitals, and 29.9% had a pathology residency program. Within the past 2 years, the College of American Pathologists had inspected 86.6% of the laboratories, and laboratory inspections were conducted by the Joint Commission (21.6%). Tables 1 through 3 show characteristics of participating institutions in both the 2004 and the 2010 study. All but 1 participating laboratory performed testing for inpatients and outpatients, with this laboratory reporting that 100% of tests were for patients not in acute care hospitals or an emergency department.

Table 3.

Distribution of Laboratory Testing For Nonacute Care Patients and Testing Referred to Other Laboratories

Distribution of Laboratory Testing For Nonacute Care Patients and Testing Referred to Other Laboratories
Distribution of Laboratory Testing For Nonacute Care Patients and Testing Referred to Other Laboratories

Staffing Ratios

The data distributions for all labor productivity and management span of control measures are shown in Table 4 and compared to the results of the 2004 study.1 There was a wide range in these performance measures among participating laboratories, but remarkable similarities in the performance distribution of this study and that of the 2004 study.1 

Table 4.

Distribution of Staffing Ratiosa

Distribution of Staffing Ratiosa
Distribution of Staffing Ratiosa

Association of Test Volume With Labor Productivity

Higher volume sections tended to have higher productivity ratios for technical staff (Table 5). A significant association was found for labor productivity in chemistry/hematology/immunology and transfusion medicine but not in histology and microbiology: P ≤ .01 for all comparisons (cytology, P  =  .01; chemistry/hematology/immunology, P < .001; transfusion medicine–transfusions, P < .001; transfusion medicine–crossmatch or type and screen, P < .001) except histology and microbiology, with P  =  .11 for microbiology and P  =  .10 for histology.

Table 5.

Association of Test Volume With Labor Productivity

Association of Test Volume With Labor Productivity
Association of Test Volume With Labor Productivity

The study participants were aggregated into laboratory-section peer groups of similar testing volume. Peer groups were not created for microbiology and the histology subsection, because of the lack of relationship between test volume and labor productivity in these 2 laboratory sections. There is no relationship between peer groups for the different laboratory sections; for example, peer group 1 for cytology does not necessarily contain the same laboratories as peer group 1 in chemistry/hematology/immunology.

The distribution of labor productivity ratios, using peer groups of similar testing volume, is presented in Table 5. Even within peer groups, there was wide variation in productivity from laboratory to laboratory.

Lack of Association of Test Volume With Management Span of Control Except in Blood Bank

There were no relationships between the testing volume per laboratory section and management span of control in anatomic pathology, histology or cytology section, chemistry/hematology/immunology, or microbiology.

In the transfusion medicine laboratory section, testing volume, when defined as the number of transfused units, was statistically associated with greater management span of control (P < .001). Managers of larger testing volume sections tended to supervise more people. This relationship is shown in Table 6.

Table 6.

Association of Transfusion Volume With Management Span of Controla

Association of Transfusion Volume With Management Span of Controla
Association of Transfusion Volume With Management Span of Controla

Variables Associated With Labor Productivity and Management Span of Control

Several variables were found to be significantly associated with 1 or more staffing ratios.

Higher labor productivity in cytology (number of cytology accessions per cytology nonmanagement FTE) was associated with laboratories that provide neuropathology services (median cytology productivity ratio, 4336 versus 1939, P < .001). Higher management span of control in anatomic pathology was associated with laboratories that perform immunohistochemistry and/or in situ hybridization services (median nonmanagement FTEs per management FTE ratio, 10.0 versus 6.5, P  =  .01).

There were no demographic or practice variables associated with the histology labor productivity ratio or the number of histology blocks per histology nonmanagement FTE.

Higher labor productivity in chemistry/hematology/immunology (number of billable tests per nonmanagement FTE) was associated with a higher percentage of laboratory tests for nonacute care patients (Table 7). This association was not biased by the population of reference laboratories participating in the study. The number of manual peripheral smears performed by the section was not associated with lower labor productivity, even though examination of manual peripheral smears is particularly time-consuming. Lower management span of control was associated with a higher percentage of laboratory tests for nonacute care patients (Table 8).

Table 7.

Relationship Between Chemistry/Hematology/Immunology Labor Productivity Ratio and Percentage of Nonacute Testinga

Relationship Between Chemistry/Hematology/Immunology Labor Productivity Ratio and Percentage of Nonacute Testinga
Relationship Between Chemistry/Hematology/Immunology Labor Productivity Ratio and Percentage of Nonacute Testinga
Table 8.

Relationship Between Management Span of Control in Chemistry/Hematology/Immunology and Percentage of Nonacute Testinga

Relationship Between Management Span of Control in Chemistry/Hematology/Immunology and Percentage of Nonacute Testinga
Relationship Between Management Span of Control in Chemistry/Hematology/Immunology and Percentage of Nonacute Testinga

Higher labor productivity in microbiology (number of billable tests per nonmanagement FTE) was associated with laboratories that performed limited parasitology testing (no parasitology performed [9213] versus limited parasitology performed [12 985] versus identification of organisms to the extent required for clinical diagnosis and therapy selection [9686], P  =  .02). Laboratories that performed a high percentage of urine cultures did not tend to be more productive, even though urine cultures are considered by many to require less work than other types of cultures. Higher management span of control was associated with laboratories that performed fungal cultures and identification of common pathogens (9.1 versus 4.7; P  =  .03).

There were no demographic or practice variables associated with the labor productivity ratios in transfusion medicine besides the volume associations discussed earlier, or with management span of control.

There is no association between the number of separate testing facilities and the percentage of testing for nonacute care patients.

The populations of participating laboratories in the 2004 and 2010 studies were significantly different with respect to institution size. This demographic difference did not have a measurable effect on the results. It was not a significant covariate for any of the staffing ratio models. A comparison of the 2004 and 2010 staffing and management span of control ratios, using a multivariate linear regression model and controlling for test volume, found no significant differences. We also compared the 2004 and 2010 performance for the subset of 16 laboratories that participated in both studies and found no statistically significant differences in the patterns of productivity and span of control in the 2 studies.

COMMENT

Balancing laboratory efficiency with optimal service levels is always a challenge. Because of the significant costs related to labor, appropriate staffing is an important management responsibility. Many factors enter into these decisions, including an understanding of reasonable labor productivity, consideration of staff experience and motivation, availability of automation, special labor-intensive testing, and clerical and teaching responsibilities. Similarly, the utilization and distribution of managers in the table of organization requires the consideration of institutional human resources policies, technical knowledge expectations, individual experience, and additional, nonsupervisory, job expectations.

As pointed out by Valenstein et al,1 formulaic approaches to laboratory staffing seem inadequate. The inadequacy stems from variability in every institution and laboratory. This variation, which both the 2004 and the 2010 study document in similar patterns, must be taken into account for decisions regarding the appropriate balance of cost efficiency and level of service provided. Because of the significant association of productivity with testing volume in several of the laboratory sections, the breakdown of productivity data by volume-determined peer groups provides the most relevant benchmark information.

Findings of the sort presented here do point out opportunities for improvement that staffing formulas miss. Laboratories with labor productivity in the bottom quartiles should review their operational efficiency and assess whether staffing adjustments or reallocation are necessary and appropriate. Those laboratories with productivity in the top quartile should examine their service performance, as measured by frequency of errors, and throughput (turnaround time). If these indices increase, then the cost of efficiency may be too high.

Our current study has the same limitations that were noted by Valenstein et al1 in their original study:

  1. Nontechnical staff members were not counted in this study, and the variable application of such staff will influence the functional level in a given laboratory.

  2. This study did not address or examine errors or throughput times. Highly productive laboratories may also have high error rates or slow turnaround time.

  3. Subtle differences in test mix or differences in human resource policies were not considered in this study but could affect staffing needs/requirements.

  4. Data were self-reported and validation of all the data submitted was not possible, although statistical procedures to eliminate selected outliers from the data pool before analysis was performed by CAP biostatisticians.

  5. The measures selected for the productivity evaluation in this study are not the only measures that could have been selected. Use of alternative measures may have produced different results.

  6. This study represents a sample of the clinical laboratory population, which may or may not be representative of industry performance.

Despite these limitations, this study, like the previous Q-Probes study,1 provides useful reference information for laboratory directors and managers to evaluate the staff and efficiency of their laboratories.

The first positive point emphasized by the consistent findings in the 2004 and 2010 studies is that comparable productivity information can be generated, which takes into account the differences among laboratory sections. A second point is that productivity is inseparably linked to test volume, and a third point is that nonacute outpatient testing amplifies this effect, for the central chemistry/hematology/immunology section. Along the same lines, a fourth point is that lower management span of control was associated with a higher percentage of laboratory tests for such nonacute patients. A fifth point is that in the microbiology section, limited parasitology testing was associated with higher productivity, demonstrating that full-service parasitology is indeed a labor-intensive undertaking. Sixth, the fraction of urine cultures (thought by some observers to be a labor-saving opportunity) did not require less work. Seventh, to the surprise of some observers, the fraction of manual review of blood smears in hematology was not associated with a significant decrease in productivity. The studies also emphasized 2 pertinent negatives that hospital and health care system administrators tend to ignore at their peril: first, no demographic or practice variables could be linked to variations in productivity; in other words, laboratory sectional productivity tends to be structural, depending on the services offered rather than functional, depending on how those services are designed; second, the structures, in the sense that we are using, of transfusion medicine section and anatomic pathology subsections are strikingly different from those of chemistry/hematology/immunology and microbiology. Studies such as the 2 Q-Probes studies, which collect staffing and test volume data, perform an important service for administrative decision makers: they separate plausible hypotheses about laboratory productivity from evidence-based observation about the relation between these fundamental measures for the people doing testing and the number of tests that they do. Future studies of laboratory staffing, focused on the impact of laboratory automation technology, may provide valuable additional data for laboratory management.

References

1.
Valenstein
PN
,
Souers
R
,
Wilkinson
DS
.
Staffing benchmarks for clinical laboratories: a College of American Pathologists Q-Probes study of staffing at 151 institutions
.
Arch Pathol Lab Med
.
2005
;
129
(
4
):
467
473
.
2.
Valenstein
P
,
Praestgaard
A
,
Lepoff
R
.
Six-year trends in productivity and utilization of 73 clinical laboratories: a College of American Pathologists Laboratory Management Index Program study
.
Arch Pathol Lab Med
.
2001
;
125
(
9
):
1153
1161
.
3.
Howanitz
PJ
.
Quality assurance measurements in departments of pathology and laboratory medicine
.
Arch Pathol Lab Med
.
1990
;
114
(
11
):
1131
1135
.

Author notes

From the Department of Pathology and Laboratory Medicine, Henry Ford Hospital, Detroit, Michigan (Dr Jones); the Department of Pathology and Laboratory Medicine, University of Wisconsin School of Medicine and Public Health, Madison, Wisconsin (Dr Darcy); Statistics/Biostatistics, College of American Pathologists, Northfield, Illinois (Ms Souers); and Department of Pathology and Laboratory Medicine, Henry Ford Health System, Detroit, Michigan (Dr Meier).

The authors have no relevant financial interest in the products or companies described in this article.