Choosing wisely from a plentiful market.

Make way for progress, say respondents to MLO's national survey, but don't just let it happen. A successful transition for each new piece of equipment often requires months of savvy planning. This article concludes a two-part report that began in last month's issue.

Automated analyzers aren't toasters; you can't just plug them in and get the crunch you want. Not even/instrument will fit each situation, not to mention whether it will fit physically in the available space. How do you know whether you really need it? Can you afford it in the first place?

There's only one way to avoid tripping over the machine itself or being tripped up by unforeseen problems: "Plan, plan, and plan some more," writes one respondent to MLO's exclusive national survey. "Have meetings to try to think of every contingency. Then dive in." (For an explanation of the survey methodology, see Part 1 of this article in the December 1996 issue.)

Think ahead

The time span from decision to delivery can take 3 to 6 months. Multiple or very large purchases may take up to several years. Furthermore, technology dates quickly. Those are among the reasons numerous respondents suggest planning far beyond current requirements to those of 5 years from now.

Check your prospective analyzer's ability "to provide an expanded menu based on your facility's future plans," says an administrative director. That's no easy task in these days of capitation and HMOs, respondents note. "You have to try to second-guess the future," adds a chief technologist.

Waiting until equipment degenerates is a mistake. "Make the decision before your current instrument is dead and too expensive to fix," advises Daniel M. Baer, MD, chief of pathology, Veterans Affairs Medical Center, and professor of pathology, Oregon Health Sciences University, Portland, Ore. (For other errors to avoid, see "What to look for in choosing automated instruments.")

If you have biomedical engineers on staff, Dr. Baer suggests, rely on their recommendations. They know best when repairs on a particular instrument are becoming too frequent or too complex to be cost effective.

Purchasing considerations

Asked about the relative importance of various factors in purchasing automated equipment, nearly all respondents pointed to accuracy (99%) and cost-effectiveness (98%) as crucial. (The questionnaire asked for a ranking from 1 ("not important") to 5 ("extremely important"); the percentages listed here reflect the totals of those who checked either 4 or 5.)

Ease of use was specified by 94%; turnaround time, 93%; low maintenance and customer service, 92% each; employee safety, 89%; and reduction of work hours, 79%.

If you have staff reductions in mind, writes a lab manager from Albuquerque, N.M., you may have to change more than equipment. "Automation by itself won't reduce staffing," the respondent notes. "It requires work redesign also."

Remain alert for the need to make changes, advises an administrative director in Tulsa, Okla., and "be willing to modify operations as automation affects them." The aim of purchasing decisions should not be merely to reduce staff, says an administrative director, but "to improve services - turnaround time, quality, flexibility, test menus - to reduce patients' length of stay and ultimately costs."

Low maintenance and the degree and speed of response by manufacturers should be priorities in all purchases of automated equipment, say many respondents. "Instruments are very similar in what they do, how fast they do it, and how much they cost," Dr. Baer says. "Customer service is a means of differentiating among manufacturers after you have eliminated others." Seek a company that can provide both hardware and software specialists who are close by, suggests a lab manager in Leeds, Me.

Another important criterion becomes obvious as soon as the instrument arrives, yet easily is overlooked in planning stages: space. "Provide bulk space for big pieces of equipment and an adequate work area for technologists working on them," proposes a section supervisor in Norwalk, Conn. "The physical layout of the lab must be factored in when decisions are made," an administrative director in Pompton Plains, N.J., concurs.

Centralize automated instruments rather than departmentalizing them, several respondents advise, since analyzers may be most functional if their use crosses traditional department lines. Also consider that benchtop analyzers may be smaller than those that stand on the floor, but they may take up valuable counter space.

Think, too, about how well the equipment will fit into its new culture. To make life simpler for the staff, says an assistant chief technologist in Falmouth, Mass., "Try to keep procedures as uniform and standardized throughout the lab as possible."

Look for instruments that use bar coding wherever feasible, several respondents recommend. "Pay close attention to calibration and maintenance requirements," adds another assistant chief technologist from Nashville, Tenn. Check the CPT coding and reimbursement levels for analyzers you are considering, says a lab manager from Mt. Pleasant, Mich.

Assess the intangible benefits of new instruments, such as enabling the lab to perform new tests or provide a new service, suggests an administrative director from Boston, Mass., "Don't inflate efficiencies gained with automation," however, says a lab manager in Baltimore, Md. who learned the hard way. "Service issues now occupy a great deal of the time we relinquished to automation. Paperwork also has increased greatly."

Make sure your lab's output is high enough for any instrument you consider, writes another lab manager in High Point, N.C. whose lab "wasn't big enough. Our equipment causes us more work every day at closing because of printing work sheets, doing sample tests, and checking sign-in logs." Small hospital labs, writes an administrative director in Bolivar, Mo., "must weigh the advantages of safety, faster turnaround time, and greater productivity against the increased expense of operation and service."

Price

Study startup costs and be aware of consumables, which can cost dearly, advises a lab manager from Haxtun, Colo., "Make sure you have enough test volume to justify added costs." An administrative director in Jackson, Ky., advises, "Do your homework on prices before vendors come to make their presentations."

In dealing with the vendor and speaking with other users, be sure to understand costs, "such as service contracts, upgrades, and the use of dedicated reagents," advises a lab manager in Butte, Mont. "Make sure you're given costs for billable tests, not tests in a vial or bottle," says a consultant in Lexington, Ky. "The vial may expire before you can use the full supply of reagent inside it."

Costs add up for shipping and handling of reagents, several respondents note; in your negotiations, see if the vendor will absorb that expense. In addition, says a chemistry section supervisor in Schenectady, N.Y., avoid purchase agreements that prevent you from buying reagents from cost-competitive vendors.

Don't be blindsided by price alone, though, advises another section supervisor in Leavenworth, Kan. Its significance may be overshadowed by that of your performance and maintenance requirements. "The cost of purchase is outweighed by savings in tech hours, satisfaction of MDs in turnaround time, and accuracy," says a section supervisor in Eau Claire, Wis. Attributes such as bar code readers (rated very important by many respondents) and clot detectors may offset small cost savings, says another section supervisor in Knoxville, Tenn. Another respondent points out, "Look at total cost of operation, not just cost per test."

To save money, "Keep equipment as simple as possible," writes a lab manager from Mt. Pleasant, Mich., "Look for low-maintenance items," service contracts will add to the expense.

Consider durability as well. "It does no good to skimp on price if the equipment will have to be upgraded or replaced in a short period of time," says a section supervisor in Santa Rosa, Calif. "On the other hand, don't get more machine than you need; it may not be the time saver you desire and could turn into a money eater."

"Get the best, not necessarily the cheapest," says an administrative director in Wichita, Kan. Methods must be sound, with good CAP or other proficiency testing scores, insists a lab manager in Great Falls, Mont.

Sources of information

Asked what sources they found most and least useful when seeking information on laboratory automation, respondents made it clear word of mouth was paramount.

Telephone research. For each instrument you want to buy, call at least a dozen people you respect and spend a total of at least 4 hours on the phone, Dr. Statland suggests. Approach people in various parts of the country, not just nearby, respondents recommend.

The experience of laboratories you approach should closely resemble that of your own lab, Dr. Statland says. To provide applicable information, they should be somewhat larger than yours - not much larger and not smaller. "If I do 300 of a certain type of tests per day, I want to talk to a lab that does 320 to 380 of them," Dr. Stat-land says. To be sure you're hearing unbiased opinions, ask whether the person is a consultant for the company in any way.

"Don't talk to just the customers the company recommends," says a section supervisor from Cincinnati, Ohio. Find a variety of customers." If you can discover a lab that once used the equipment you have in mind but discontinued it, a chief technologist in Temple, Texas, says, try to find out why.

Ask your contact pointed questions, Dr. Statland suggests, such as which other instruments that laboratory considered and what were the deciding factors for the one chosen. "Say, 'Tell me three things you like about it and three things you wish could be improved,' not just 'Do you like it?'," Dr. Statland says. Ask whether the reagents tend to expire before they're used up and how often results are aberrantly high or low.

Inquire how the company has been to work with, Dr. Baer suggests. Run through a checklist: How much downtime have you had? How much time elapsed from installation to routine use? Does the staff feel comfortable using the instrument? How does it compare with other pieces of the same complexity?

When talking to users of the equipment you're considering, writes a lab director in Watonga, Okla., "ask how the instrument performs on proficiency testing." Sign onto user's groups on the Internet, suggests a department chairperson in Miami, and ask away. In addition, trade journals can be a useful source for system information.

A good question to ask toward the end of your talk: "Have you had any experience that I haven't mentioned?" Unexpected factors inevitably pop up. One respondent, for example, points out that "some equipment is very temperature sensitive." A certain instrument in that section supervisor's lab "has been very temperamental" on that account.

Site visits. Talking on the phone is a good way to start, but the closer you can get to an instrument you're considering, the better informed you'll be. Once you have narrowed down your choices, pay site visits to see the instrument in action.

If possible, arrange for hands-on experience with the equipment during the site visit. For a frank appraisal, talk privately and confidentially with a technologist who uses the equipment but is not involved directly in purchasing decisions.

Exposure to too much information is like seeing too many houses or apartments in one day: It becomes hard to keep them apart in your mind. Putting your thoughts on paper will help, advises a laboratory supervisor in Harrisburg, Pa.: "Keep a running journal on advantages and disadvantages. By the time you see the last machine, they'll all run together."

Even better, have the instrument brought to you, says Marti K. Bailey, MT(ASCP), chief technologist in hematology and microscopy at University Hospital, Pennsylvania State University, Hershey, Pa. Vendors are making more such concessions than they used to, she has observed. "We have an in-house demonstration for a week or two before we decide if we want to issue a purchase order." One company refused; "They wouldn't bring it in, we wouldn't buy it," she says.

Vendors will do this even with relatively large pieces of equipment, Bailey has found, setting up "everything except interfacing. In chemistry, we have a lot of big equipment sitting here that we're trying."

Bailey's lab is considering bringing in a hematology instrument that the company has offered to customize in exchange for a 5-year contract. "It's very important to a company like that to build a client base and keep it," Bailey says. Once the instrument is in place, Bailey says, companies are very willing to send technical reps to do precision studies to make sure ali's well in the CLIA department.

Can every lab count on such wooing? Not necessarily. Bailey admits that Hershey Medical Center is "a showcase account. Maybe it's easier for us to get so much attention."

At least one respondent's lab, however, seems to gamer such treatment. "Make sure the tech is doing the work in the department," writes the MT/MLT from Beaver, Pa. "Pull the instrument into the laboratory and run it for about a month before making a decision." As a section supervisor from Corsicana, Texas, points out, "Working with equipment from day to day gives a much truer picture than that painted by anyone else."

"If you can't do anything else," according to Dr. Baer, examining the instrument at a trade show is "better than nothing." Trade shows do provide a good look at the actual size of equipment, a chief technologist in Temple, Texas, points out. But don't expect a lengthy discussion of your needs. At professional meetings, sales reps usually are too busy to explore any situation in depth. You may have to wait in line. "If the instrument is any good," Dr. Baer observes, "a lot of people want to see it."

Acquisition without purchase

Methods of acquiring automated instruments other than buying them are becoming increasingly popular. Some respondents, in fact, recommended that in future surveys MLO should include questions about leasing and reagent rental rather than just purchasing.

Many laboratories acquire instruments, for example, by entering into cost-per-test contracts. Based on the volume of use of relevant tests, the manufacturer develops a sliding scale pricing system, and the lab is billed according to how many tests are run.

Advantages of this approach are threefold, according to Dr. Baer. First, since the lab doesn't own the equipment, it doesn't have to come up with capital funds. Second, the arrangement may include servicing, so that the lab doesn't have to pay for a service contract. Third, it's a lot easier to trade up to a higher or newer model because the lab isn't locked into a particular piece of equipment.

"We've done this for 3 or 4 years and have not had a problem," Dr. Baer says of his institution. In addition, he says, "we write contracts that pool work from different hospitals," all of which use their own instruments. The resulting volume discounts are highly advantageous, Dr. Baer notes.

"One of the forgotten costs in purchasing is depreciation," which adds up to thousands of dollars, points out Bailey, who does cost accounting for her lab. Hershey, which previously purchased all its equipment, has increased reagent rentals in exchange for a free annual equipment update. "That's a bonus for us," she says. "I can't see buying equipment any more. I lease my car, too."

Reagent rentals or short-term leases are one good way to go, agrees an administrative director in Bend, Ore., because "the technology is changing too fast."

Facing interfacing

"The value of a sophisticated auto, mated instrument plummets if it doesn't interface with the computer," says Rodney Markin, MD, PhD, vice chair and professor of pathology at the University of Nebraska Medical Center, Omaha. "Without two-way integration with the lab computer," asserts a section supervisor in Long Beach, Calif., "any gains will be slight."

Connecting with the laboratory information system (US) is an essential part of installation, says Bailey, yet "it seems to be one of the hardest things the manufacturer has to do." That's not always the manufacturer's fault. At Bailey's laboratory, for example, it wasn't originally possible to effect an interface for the coagulation analyzer because the LIS "had no ports left." Recent purchases of a new LIS and new coagulation equipment permitted the connection.

"Obtain in your contract a guarantee that the instrument will interface with your LIS within a specified time or compensation will apply," says an administrative director in Bolivar, Mo., who reports that establishing an interface between the LIS and a specific instrument took the lab 6 months.

Determine whether the instrument you want has been interfaced with your type of LIS before, advises a laboratory business manager in Fort Wayne, Ind. "Contact users of the instrument, and confirm what LIS they have and any problems noted."

You may have to work from the other direction, says a lab director from Marietta, Ga., who advises, "Get a centralized computer that is interfaced easily first."

Horse before the cart

Think very hard about your needs before you leap, Dr. Markin urges. "The first stumbling block," he says, "is that people seem to think automation means hardware." In his view, by contrast, automation is "80% to 90% information system and 10% to 20% hardware." The ideal, he believes, is to "start with a software solution and work up to a hardware solution.

'The majority of people think that if they buy up-front specimen processing, they have solved the problem," Dr. Markin continues. "But there is a cutoff to make the cost justifiable. People just assume it will pay back. They need data to prove that." The answer, he says, is to "take it sequentially" and not try to do everything at once.

Furthermore, Dr. Markin says, many up-front specimen processing systems are too limited. If they accept test tubes of only one size, for example, they won't serve the lab's purpose later if industry standards are altered or the lab itself changes its preference. On the other hand, he says, "If the system is architecturally sound, it will be able to use all of the most common sizes." Only the volume of specimen drawn would change.

Many survey respondents similarly emphasize the need for accessibility and versatility. Look for user friendliness and the ability to change, delete, and especially add programs.

Another reason laboratorians buy automated instruments impetuously, Dr. Markin says, is that they haven't taken the time to understand their own clients and customer service requirements. The demands on a lab in a hospital that does transplantation, for example, are different from those where a specialty is cancer chemotherapy. "The 'One-size-fits-all' theory doesn't work," Dr. Markin says.

After seeing how much help labs needed in instituting automated systems, Dr. Markin and his colleagues developed an automation consulting service. They collect information, build a simulation model, and then overlay automation technology on it.

A collaborative effort

Many individuals at all levels of the laboratory are involved in the purchase of automated equipment, respondents say, and that's just how it should be. Solid thought and input from a multitude of participants should prevent most unpleasant surprises.

"Get the staff involved in the process from the beginning," writes an administrative director from Palm Springs, Calif. "Have bench-level technologists participate in the selection of the instruments," suggests an MT/MLT in South Fork, Pa., "because they will have the most connection with them."

Who suggests. The first step is recognizing the lab's need for an automated instrument. Asked who suggests or recommends such purchases in their own labs, respondents most often (68%) mentioned MTs, MLTs, and CLTs, the hands-on employees who may be best equipped to judge where equipment is needed. Others deeply involved in that process are department heads and supervisors (59%), chief or assistant chief technologists (also 59%), lab supervisors (54%), and pathologists (43%). Laboratory managers suggest equipment in about one-third (32%) of respondents' labs; lab directors, 29%; and lab administrators, 16%.

Marti Bailey has grown accustomed to the extent of higher-ups' participation in this early stage, but it was not ever thus. She was surprised 5 or 10 years ago, she says, when the lab manager recommended looking at a certain piece of equipment. Now "the lab manager, assistant lab manager, and even physicians are calling us." Such advice from physicians is particularly welcome, she says, because if the lab does acquire the instrument, "they've already bought in."

Information about new instruments comes from vendors as well. "Vendors that service us keep us much better informed than they did in the past," Bailey says. But one can't rely wholly on others, she says: "You have to read the lab journals religiously and look at all of the mail that comes across your desk."

Table 1

Purchasing budgets and contract costs for automated equipment

                                                 Service contracts
                         Annual budget,           and warranties
Number of beds           1996 (median)            1996 (median)

Fewer than 200              $52,000                  $20,000
200 to 399                  117,400                   50,000
400 or more                 225,000                   60,000

IT'S ALL RELATIVE: The larger the bed size and the greater the
number of tests produced at respondents' laboratories, the larger
were 1996 purchase budgets and amounts spent on contracts and
warranties. Nearly one-third (31%) of the laboratories had no
budget
at all - zero dollars - allocated for purchasing automated
equipment
in 1996, while the budgets of four out of 10 (40%) exceeded
$100,000. The rest (29%) had budgets of less than $100,000.

Who selects. The next step in the process, making the final selection, seems to lie with the laboratory manager, lab supervisor, and/or section head or supervisor. These are the individuals who, at MLO survey respondents' labs, typically decide which equipment to purchase (mentioned by 54%, 51%, and 47% of respondents, respectively). Other decision makers include lab directors (33%) and chief or assistant chief technologists (32%). The laboratory administrator decides in 26% of labs. Much less likely to make final choices are pathologists (16%) and MT/MLT/CLTs (13%).

Consultation with hands-on workers, however, is crucial. "Point number one, with an asterisk," Dr. Baer says, "is to involve the people who are going to use it" in deciding what you'll buy. His own laboratory has made efforts in this regard, he says, but "we didn't do it as well as we should have. We didn't involve the evening and graveyard shift people who have to use instruments in a different way than the day people do."

Communicating with evening and night shifts, he suggests, might involve memos, brochures, and manufacturers' videotapes showing the equipment. "Let people know what you're considering," Dr. Baer says, and invite their input.

Who authorizes. Someone has to give final approval. That task logically falls to those with budgeting power: the lab director (52%) and lab administrator (49%). The lab manager authorizes equipment purchases in 40% of labs and the pathologist in 26%.

Before presenting a purchasing plan to the administration for approval, counsels a section supervisor, prepare a detailed report focusing on the financial benefits of adopting the new equipment.

Purchasing budgets and plans

Budgets for equipment varies so much among labs that it would be hard to tell from those figures how prevalent automation has become. About one-third of the laboratories surveyed had no budget for automated equipment in 1996, yet four in 10 had budgets in excess of $100,000. The median was $60,000. Service contracts and warranties had a similarly wide range (see Table 1).

Three out of four laboratorians (74%) say they expected to remain within their allotted budgets by the end of 1996. Only 5% anticipated exceeding their budgets. The rest were unsure.

The typical laboratory (median) planned to purchase two pieces of automated equipment in 1996. One would cost less than $50,000 and the other $50,000 to $100,000.

Not only the equipment itself must be paid for. "Make sure a budget is set up for maintenance agreements," says an MT in Lawrenceville, Ga.

The instrument arrives

Survey respondents were asked how long it takes to receive delivery of an [TABULAR DATA FOR TABLE 2 OMITTED] automated instrument after its purchase has been approved. The time lapse from decision to delivery ranged from a few months to more than 2 years. While budget size had little effect on this, price did. As price increases, so does the time lag (see Table 2).

Learning how to operate new equipment requires inservices, many respondents say. Often the vendor will provide them free of charge.

"Cross-train employees for both the lab's and the employees' benefit," advises a chief technologist from Grosse Pointe Woods, Mich. Training more than one technologist to handle troubleshooting is also recommended. "Make sure your staff is thoroughly trained on any new piece of equipment," warns an administrative director in Jackson, Wyo., "because it will break down only when everyone is out of town."

The need for backup

Once the new instrument has been installed, your laboratory may become highly dependent on it. Therefore, having adequate backup is crucial.

Dr. Statland lists several ways to approach this problem. Some people follow the "gemini approach," he says, obtaining two of everything. That's expensive, of course. "It may break the bank," he says, "but if it works, it's ideal."

A more common route is to know where else your tests can be done, such as at a sister hospital or commercial laboratory, in cases of prolonged downtime.

Third, consider inserting an expectations' standard in your service contract. Similar to an insurance policy, the standard calls for compensation if the manufacturer fails to produce.

A fourth solution, appropriate for relatively simple instruments only, is to keep spare parts on site and do diagnostic workups on the phone.

"Have battery backup for power fluctuations and outages," advises an MT in Dickiham, N.D. "Planned and unplanned power outages can ruin sophisticated electronic equipment," alerts a hematology section supervisor. "You must heed manufacturers' recommendations concerning dedicated circuits and the like to protect instrumentation properly."

Optimizing usage

While instruments have to fit in to your lab, you may benefit most if you'll also consider fitting the lab to the instrument, to some extent. "Reengineering departmental configurations is a natural outcome of automation," writes a lab manager in Lawrence, Mass.

To make the best use of automation, Dr. Markin suggests, allow it to chug along during nonstandard working hours. Reallocating volume to a different time of day, he says, can smooth the flow of specimens.

"Select equipment that's easily utilized by all three shifts," advises an administrative director. "How many Stats are performed? Can profiles be batched at night?" asks a lab manager in Philadelphia, Pa., who also recommends talking with the physicians who use the lab about their preferences in that regard.

Widespread acceptance

Overall, respondents to the MLO survey are enthusiastic about automation. "Don't hesitate," they write. "Just do it." "Go for it." "We should have done it sooner." "Hurry up." "It is essential to compete in today's market." "Without automation, you will not have a lab." (See "The lab manager's role in assessing new technology" on p. 64 of this issue.)

Keeping expectations moderate also is advised; automation, respondents say, is no panacea. "Automation will not fully resolve human error issues," warns a QI coordinator in Chicago. "Automation invites errors of a different type and magnifies the errors through electronic documentation." Similarly, automation will not solve problems related to personnel, quality, and procedures, says a blood bank section head in Battle Creek, Mich. An immunology section supervisor from Long Beach, Calif. summarizes, "Don't expect miracles."

With adequate thought and preparation, however, laboratories of virtually any size can benefit greatly from automation. As in all things, time and experience numb the pain of change. "We were anxious at first," admits a urinalysis section supervisor in Monroe, La., "but we're confident and comfortable now."

What to look for in choosing automated instruments

While everyone has to determine their own level of automation, respondents to the MLO survey issue a caveat emptor - don't necessarily be dazzle by the newest rather than the tried and true. For instance, instead of buying the latest and greatest piece of equipment, you may want to opt for an instrument that has more of a track record, they urge. "I never like to be the first kid on the block to have my secret decoder ring," agrees Daniel M. Baer, MD, Chief of Pathology, Veterans Affairs Medical Center, Portland, Ore., echoing comments from many MLO survey respondents. Instead, consider instruments with a history of accuracy and reliability, says a lab manager from Sacramento, Calif.

Installing a brand-new automated system may be exciting, says Bernard E. Statland, MD, PhD, of Statland Laboratory Consulting, Nashville, but it's challenging, too. A manufacturer may "learn on your back," he warns, by solving problems in the machinery as you encounter them. A cautious coagulation section head in Baltimore, Md. suggests, "Don't buy equipment until it has been in the field for at least 3 years." While that may be a bit long, considering how quickly technology has been changing, perhaps a year or more isn't too much to ask.

However, Dr. Statland says, "Don't buy something that's ready to be replaced." How can you know an upgrade is in the offing? You may hear about the baby-to-be through the rumor mill or deduce the situation because the instrument is several years old and "mill have a sense that the company is ready to flex its muscles somewhere else," Dr. Statland says. If so, see if you can negotiate for your contract to state that if a new model comes out, the manufacturer will upgrade your gratis.

Finally, Dr. Statland says, one mistake to avoid is failing to make sure the contract provides sufficient guarantees for service and support. Survey respondents agree, urging alertness to a vast array of details in this area. Even further, says another lab manager from Newton, N.J., "Be sure to have an 'out' clause in your contract if the equipment doesn't perform as expected or exceeds anticipated cost."

Marcia Ringel is a freelance health writer and editor in Ridgewood, N.J.