Feature Story

How many staff does it take?
It’s no light bulb joke — Q-Probes sizes up staffing ratios

November 2004
Karen Lusky

If an airline discovered it was using half the number of mechanics as other airlines of similar size, it would have to ask why, right?

Paul Valenstein, MD, poses the question to explain a key recommendation in the new Q-Probes study that measures technical and management staffing ratios in labs that do similar types and volumes of testing. That advice: Proceed with caution if your lab is an outlier.

The CAP Q-Probes study, "Technical Staffing Ratios," doesn’t probe the "whys" of higher staff productivity as much as it provides a frame of reference for labs to gauge how they compare with their peers. And the benchmarking can offer a wake-up call to laboratories that find themselves on either end of the bell curve, so they can figure out if they need to move a bit more toward the median.

"An airline that’s using fewer mechanics to maintain its fleet wouldn’t necessarily be cutting corners or headed for an accident," says Dr. Valenstein, president of Pathology and Laboratory Management Associates, Ann Arbor, Mich., and chair of the CAP Quality Practices Committee. "Maybe the mechanics are better trained or the fleet is newer and doesn’t require as much service," he adds. "But then again, maybe not."

By the same token, he says, "a highly productive laboratory doesn’t necessarily have problems with staff morale, turnaround time, or accuracy." But prudence requires the lab to think through why it’s so productive—and whether it’s unknowingly paying a price for high productivity somewhere else, he says.

Conversely, labs that have low staff productivity compared with that of their peers should look at their lab design, workflow, instrumentation, and hiring decisions, says study coauthor David Wilkinson, MD, PhD, chair of pathology and director of laboratories for the Virginia Commonwealth University Health System, Richmond, and a member of the Quality Practices Committee.

The Q-Probes study, the results of which were released in September, isn’t intended to provide a rigid formula for staffing. Instead, it’s meant to be a framework for organizations trying to provide appropriate levels of labor, which accounts for 50 to 70 percent of direct clinical laboratory cost, say the study’s authors.

To conduct the Q-Probes study, the researchers tapped 151 institutions in early 2004 to provide staffing and test-volume data for their most recent fiscal year. The participating hospitals included 61.2 percent private nonprofits, 11 percent private for-profit, about 10 percent government-owned, 5.4 percent university, and one independent lab. Thirty-five percent were teaching hospitals, and 21 percent had a pathology residency program.

The study addresses two questions on the minds of lab directors, managers, and technical staff alike: What is a reasonable number of technologists required to staff a laboratory? And what is a reasonable number of managers required to supervise those technologists?

"The chief financial officer might be thinking that one technologist is enough to staff the laboratory, and the medical technologists already in the lab might say that 600 are necessary to do a good job," Dr. Valenstein says. "This Q-Probes study provides information about how many technologists and managers are being employed in real-world laboratories of different sizes and configurations."

As a first step in conducting the study, the researchers had to define what they were measuring in such a way that would allow apples to be compared with apples. With that in mind, Drs. Valenstein and Wilkinson decided to look at the productivity of technologists and frontline managers separately for each of four major laboratory sections: anatomic pathology; chemistry/hematology/immunology; microbiology; and transfusion medicine. (Related article: "How lab sections were defined")

The study defines technical staff productivity as the amount of work performed per full-time equivalent staff member (40 hours per week or 2,080 hours per year of paid activity assigned to the laboratory cost center). Management productivity, or "management span of control," is simply the number of technical staff supervised per full-time manager.

The study defines a manager as someone who spends more than 50 percent of his or her time supervising frontline staff. (It doesn’t include corporate-level managers who set strategy.) Nonmanagement staff spend more than half their time doing "bench" testing or procedures, or nontechnical tasks. Nonmanagement staff include medical technologists or technicians, cytotechnologists, or histotechnologists.

Employees who served in both technical (bench work) and supervisory roles had their time divided between the two categories.

The study excludes staff time spent working with a laboratory computer system, phlebotomy, point-of-care testing, or a number of nontechnical tasks, such as courier, clerical transcription, marketing, or billing operations.

The study defines output or work differently for each of the four laboratory sections. "For technologists in microbiology and chemistry/ hematology, we defined productivity as the number of billable tests produced per technologist per year," Dr. Valenstein explains.

In transfusion medicine, the researchers counted separately the number of crossmatches, type and screens, and the units transfused per technologist per year. For histology, they counted tissue blocks, and in cytology it was accessions per FTE technical staff per year.

For each laboratory section, the researchers reported a ratio of work per technologist, referred to as "labor productivity." For the AP section, they calculated separate labor productivity ratios for histology and cytology.

Managerial productivity was reported as management span of control (the number of people supervised by a manager). The premise is that "a more productive manager supervises more people," Dr. Valenstein says.

So what did the data show? Labor productivity increased with the volume of testing in every section except histology. "Larger laboratories had higher labor productivity," he says. Since managers of small labs can’t bump up their test counts over night, the researchers had to find a way to provide meaningful staffing ratios for directors of different sized operations. So they further divided each of the four laboratory sections into peer groups based solely on volume of testing. Peer group "No. 1" for cytology was smaller than peer group "No. 2," and so on. But hospitals with a peer group "No. 1" cytology lab didn’t necessarily have a peer group "No. 1" microbiology section.

For example, peer group No. 1 in chemistry/hematology/immunology labor productivity (defined as lab sections with a range of billable tests from 59,113 to 370,338) showed 11,233 billable tests per nonmanagement FTE at the 10th percentile, and 66,125 at the 90th percentile. The much higher-volume peer group No. 3 (lab sections with 1,423,725 to 6,491,344 billable tests per year) showed 27,788 billable tests per nonmanagement FTE at the 10th percentile, compared with 105,081 at the 90th percentile.

Overall, the study showed a threefold variability in labor productivity ratios from the 10th to the 90th percentile for all lab sections except histology. "But even within laboratory volume peer groups, there’s a two- to fourfold variation in productivity from the 10th to the 90th percentile," Dr. Wilkinson says. "Those variations within a peer group reflect management decisions that affect productivity."

Dr. Wilkinson believes histology stood as the lone exception to the relationship between test volume and technical staff productivity because it’s such a hands-on process. "Essentially every step in histology has to be handled by an individual, whereas in chemistry, one tech may be handling anywhere from one specimen to 1,000 an hour," he says.

(That’s not to say that histology labs can’t improve their productivity. Dr. Wilkinson points to data from his own laboratory, for example, showing dramatic improvements in productivity resulting from combining two histology laboratories that weren’t operating at their full capacity: a surgical pathology laboratory and a neuropathology laboratory.)

What was the impact of laboratory testing volume on management span of control? In AP sections (histology only) and in transfusion medicine, managers of larger testing-volume sections tended to supervise more people. But that relationship didn’t hold true for chemistry/ hematology/immunology and microbiology sections—or for the cytology subsection in anatomic pathology.

Even so, the study showed that as labs become more productive in aggregate, the span of managerial control does increase. "Ergo, one characteristic of highly productive labs may be a higher span of managerial control," Dr. Wilkinson says. That is, fewer managers per supervisee.

Management span of control is also determined to some extent by the type of testing a lab does. Chemistry had the highest median management span of control (11.6 nonmanagement staff per manager). Transfusion medicine was lowest, with a median of 5.5 staff per manager. AP sections and microbiology fell in the middle. What might explain the differences? "Chemistry is highly automated and tends to run all three shifts, so generally one tends to see more workers per manager in chemistry than in other areas," says Dr. Wilkinson.

AP labs, on the other hand, usually run only one shift in most settings, so you need a supervisor for that shift. "If the lab runs all three shifts, there may be a lead tech on other shifts but still the one supervisor," he says. And transfusion medicine faces a lot of technical and regulatory issues that might dictate management loads, Dr. Wilkinson notes.

The type of testing can also affect technologists’ productivity. For example, Dr. Valenstein says, "microbiology labs that perform extensive mycology testing produce fewer billable tests than those that perform no or minimal mycology." Interestingly, the study did not find higher productivity in laboratories that perform a high percentage of urine cultures, even though many people view urine cultures as requiring less work than other types of cultures.

Laboratory sections that find themselves in the bottom or top quartiles of labor productivity should scrutinize their staffing and ask themselves key questions.

"A director overseeing a laboratory that is significantly less productive than peer labs [matched for volume of testing] might look to see if the lab’s productivity has crept downward slowly over time. Perhaps the workload diminished slowly while staffing remained the same," suggests Dr. Valenstein. "Or the facility may have lost its edge while other similarly sized labs passed it by and increased their productivity." Perhaps this lab needs to assign technical staff to other areas, stop hiring, or do both for a while and let attrition reduce the number of technical staff, he adds.

Labs that want to track their productivity over time might want to consider enrolling in the CAP’s Laboratory Management Index Program, or LMIP. "By using the LMIP Peer Group Directory, they may be able to contact labs that are more productive and gain insights into how the more productive labs function," Dr. Wilkinson says.

What might labs do if they find themselves in the lowest quartile of management span of control? The lab could, of course, lay off some of the managers, an option that many labs try to avoid. "Another strategy is to combine a large section with a smaller one and anticipate that one of the managers may leave or go back to the bench," Dr. Valenstein says. "If that’s not feasible, it may be possible for managers to do more hands-on testing."

The study authors also recommend that, in some cases, organizations may need to streamline administrative tasks (such as number of meetings, management training sessions, and burdensome human resource policies) to allow managers to supervise more people.

The directors of labs in the top quartile of labor productivity might pat themselves on the back for their efficiency, but they should also dissect what’s driving their numbers. "For example, the laboratory may enjoy high labor productivity because it has invested more heavily in automation than other laboratories," says Dr. Valenstein. "Yet the lab may or may not be more productive overall, because it has tied up capital in the automation equipment. So what economists call ’total factor productivity’ may be average even though labor productivity is high."

Another explanation for high productivity can be found in the test mix or the setting in which services are provided. "If a lab performs relatively simple tests or serves an outpatient population where hour-by-hour turnaround is less important, the lab may be able to function with fewer staff," he says. More experienced or motivated staff are also typically more productive.

If the lab can’t readily explain why it’s more productive, however, it might analyze key outcomes to see if its efficiency is taking a hidden toll on patient safety, staff morale, or customer satisfaction.

For example, does the lab have an unacceptable level of errors or near misses? "You can measure errors, or error surrogates, in a number of ways," says Dr. Wilkinson. One way is to look at the lab’s rate of amended or corrected lab reports. Say the lab ran a serum potassium and releases the value but then realizes that the quality control was out, so the potassium needs to be redone. Or someone realizes the blood specimen was hemolyzed and the reported potassium value is too high. "That’s something fairly easy to track with most computer systems—and one example of how to check error rates," says Dr. Wilkinson. "There’s not an exact correlation with actual errors, but the rate of corrected reports is a reasonable indicator or surrogate for the error rate."

Histology labs can make sure their block counts are correct. "With most automated systems, the histology techs label the blocks, and at the end of the day they may have blocks for which they can’t account," says Dr. Wilkinson. "This may indicate an error in labeling." A lab could also track the number of times it fails proficiency testing.

The study didn’t look at laboratory turnaround times, which can also reveal a lot about customer satisfaction and the impact of lab testing on patient care. "If a lab only performs testing 8 to 5 Monday through Friday, anything that comes in after that will be run later," Dr. Wilkinson says. "The laboratory staff could handle a huge volume of specimens very efficiently during the weekdays, but the doctors may need some of those test results sooner."

"Labs that are going to give their emergency department, operating room, or ICU 15- to 30-minute turnaround times will have to be ’overstaffed’ relative to the volume of testing to avoid having people waiting around for test results," he adds. But that’s changing to some extent with the new high-volume chemistry and hematology instruments—random-access analyzers—that can handle stat specimens as soon as they come in the door.

The study did not delve into the management decisions and details that could account for the wide variation in productivity among laboratories, says Dr. Wilkinson. For example, the study didn’t survey whether the labs were batching specimens, which tends to boost labor productivity figures. Nor did the study look at whether participants had implemented Lean manufacturing principles or other productivity improvement methods.

"When you create a Q-Probes study, you’re stuck between a rock and a hard place," Dr. Wilkinson explains, "because you want to collect data with enough granularity to it to draw meaningful conclusions that will be useful to laboratories." But the data-collection instrument can’t be so daunting that it discourages labs from participating.

"So there’s both a science and an art to creating a Q-Probes," he says. "This probe doesn’t tell you why productive labs are more productive—it tells you simply that some labs are more productive and that there is wide variation in productivity that goes beyond the relationship between volume and productivity."

Laboratories can make deliberate decisions about where they want to fall on the productivity continuum. Dr. Wilkinson’s approach would be to focus on achieving productivity near the median. "That way, the lab knows it’s not a slug on the one hand, and that it isn’t pushing people beyond their limits on the other hand."

Yet he points out that his lab is in an academic institution with teaching and research missions. "And generally, laboratories in academic settings aren’t as productive as community hospital laboratories or a reference laboratory where a goal of achieving near the most favorable quartile might be more appropriate," he says. Dr. Valenstein’s laboratory, for example, strives to be just under the 75th percentile in productivity, while keeping a close eye on error rates, turnaround times, staff morale, and customer satisfaction.

"There is no absolutely right answer for benchmarking. It’s a tool," Dr. Wilkinson says. Labs have to set their own targets in terms of where they want to be on that continuum for productivity, cost per test, management span of control, or whatever they are measuring.

"Every situation is unique—and unique even to sections within a laboratory," he says.

Karen Lusky is a writer in Brentwood, Tenn. For more information about CAP Q-PROBES and how to enroll, call the CAP at 800-323-4040 or 847-832-7000 option 1#.