Putting the magic to work.
The microcomputer is most helpful for quality control, according to nearly half the panelists in computerized laboratories, and word processing is the second biggest asset. But these chores only begin to define the personal computer's potential. This part of our special report will examine some of the most promising applications, and some of the disappointments as well.
Figure I shows how the supervisory laboratorians in our survey rank the value of microcomputer functions. (The preceding article listed the frequency of microcomputer uses.) Widespread applications like quality control and word processing top both lists. But test reporting engenders the highest rate of satisfaction among users--only the fifth most widely used microcomputer function, it comes in third as most helpful.
At 120-bed City of Faith Medical and Research Center in Tulsa, Okla., the microbiology section's microcomputer interfaces with the hospital's mainframe information system, moving reports quickly out of the lab and onto patient charts.
"When we enter culture results, organism descriptions, or interpretive comments on the microcomputer, they cross the interface to the hospital's information system in a minute," according to City of Faith's microbiology supervisor, Edward Fisher. "Hard copies of our data are available on the floors almost immediately. We don't have to collate reports here in the lab, run them up to the floors, and distribute them." Fisher says that this has speeded up turnaround time considerably.
Using a commercial program originally designed for hematology and chemistry work, Fisher and his staff assign accession numbers to all specimens, then call up menus that offer a range of organism descriptions, names, and comments. These can be entered day by day as results become more definitive, by entering the accession number. In doing so, the staff can review work done previously on the same specimen. Fisher says the success of the system has encouraged the microbiology section to consider a microcomputer system that will manage hospital epidemiology reporting--a major data base.
Quality control--particularly in chemistry--claims by far the most computer time in clinical laboratories. Much of this usage results from the recent trend among manufacturers of including a microcomputer as part of their quality control packages. In fact, many laboratorians report that these units are used only to run the manufacturer's QC program.
Panelists by and large state that microcomputers make QC more manageable. Alice Redman, laboratory manager at the Sansum Medical Clinical, Santa Barbara, Calif., says that the lab has four such systems, in chemistry, hematology, coagulation, and RIA, all furnished "free." Each was from a different diagnostics vendor for optimum use in a particular section.
The hematology department head of a New England hospital lab comments: "Our QC program computes the mean, standard deviation, and variance for tests that are not part of a nationwide quality control program, including thrombin time, fibrinogen, and quality control for our backup system, a Coulter S. In addition, we use a program incorporating a T-test function when calibrating the Coulter to manual methods."
In a Midwestern hospital lab, the manager reports, quality control data are entered into an Apple and transmitted by telephone to the company providing the program. In return, the lab receives mean, SD, and peer comparison information.
At a large California medical center, microcomputers store data for test controls, print out graphs, log comments, and determine means, SDs, and coefficients of variation. Units are located right at the bench in the microbiology, chemistry, and quality control sections.
Workload recording is another task many laboratories assign to a microcomputer. David Van Vleck, workload coordinator at 475-bed St. Margaret Hospital in Hammond, Ind., used a spreadsheet program to create a model that lists all the laboratory's tests and profiles and then calculates test costs, based on workload units from another model, for both in-house and referred specimens.
Van Vleck notes that nearly all the laboratory's computer projects for the past year relate directly to the expected impact of DRGs. "We're using the computer extensively to calculate productivity and WLUs, and to run a hospitalwide productivity report," he says, nothing that other hospital departments have taken notice and may purchase their own microcomputers.
The lab at 420-bed Theda Clark Regional Medical Center in Neenah, Wis. conducted a cumulative comparison of expenses versus cAP workload using a personal computer, lab manager James Bracker reports. The program highlighted strong points and a few problem areas. In the highly automated chemistry section, for example, it revealed that technologists generated $90 of income for each hour worked, whereas toxicology managed considerably less and failed to meet its expenses. With new, more efficient instrumentation and two reassigned technologists, the toxicology section's cost ratio improved significantly.
When it comes to the capabilities labs with microcomputers want to have, inventory control heads the list, cited by 34 per cent. Another 31 per cent would like to expand their systems into workload recording. Budgeting, staffing, and scheduling are also noiminated for future handling by microcomputer, as shown in Figure II, especially in view of the new prospective payment system. Let's take a closer look at how some labs use microcomputers in these areas to help them cope.
As DRGs come into place, laboratorians are increasingly called upon to justify even the most minor expenditures. The bottom line for inventory management is to keep just enough supplies on hand for a specified period. If the lab purchases too much, it runs out of storage space and risks outdating; if it underbuys, it can't provide necessary services. Either mistake means budget variances and unhappy administrators.
The survey results underline this concern. While nearly a third of the panelists would like to add inventory control to their computer capabilities, 21 per cent already have, and 6 per cent name this function as the microcomputer's greatest asset.
Roman Golash, microbiology supervisor at Evanston (Ill.) Hospital, is one of them. The microbiology department at this 750-bed hospital had the usual problems juggling its $250,000 annual expendables budget--before an Apple IIe tackled the chore. Since last summer, Golash reports, supply virtually always meets demand.
To keep microbiology inventory under control, the department is developing three programs based on the weekly or monthly ordering frequency for each product. The first program will monitor average product usage per week or month. The second will record quantities ordered for each period, and the third will detail actual usage.
"When ordering, we want our actual usage to meet our guesstimate based on previous usage," Golash explains. If actual consumption comes in too high or too low, he adjusts the next order accordingly--an ongoing challenge when the monthly consumption of Bactec bottles ranges from 800 to 1,300, and the weekly need for blood agar plates fluctuates between 800 and 1,400. "We have a standing order for many supplies," he notes. "When a trend develops, we just call the supplier and adjust the next shipment."
The department also runs a daily inventory monitoring system. It's working well, but Golash looks forward to making it work even better with programs for trend analysis and for inputting supplies as they're withdrawn from storage. Although the system doesn't save that much time compared with the old manual method. Golash says it does produce much better data. "Potentially, it's a big time-saver--once we become more sophisticated programmers."
Budgeting is a primary concern for supervisory laboratorians, so it's no surprise that some 25 per cent of the panelists would like microcomputer assistance with this fiscal headache. Nearly 30 per cent of the panelists rely on their computers at budget time, and 9 per cent list budgeting as the single most helpful application.
In this area alone, the microcomputer can simplify a wide array of complicated tasks, and some panelists are using it to the hilt. Earlier, this year, the laboratory at Valley Children's Hospital in Fresno, Calif., directed by Stephen H. Kassel, M.D., submitted its first computerized budget for approval. "We prepared a four-inch thick book, complete with line-by-line justifications," Stan Schofield, clinical laboratories manager, comments. The administration was impressed with the ease of review and approved the entire $7 million lab budget without a major revision, Schofield reports happily.
The laboratory achieved this victory by putting its microcomputer through a vigorous workout. The details are worth examining. "Our hospital setup is rather unusual and makes budgeting somewhat complicated," Schofield explains. The 138-bed tertiary-care pediatric facility provides comprehensive care, including open heart surgery, and testing that extends to toxicology and virology. The laboratory's staff of 140 FTEs expend more than 2.8 million workload units annually to support in-house, outpatient, and reference testing for these high-demand specialties.
The budget program, developed last year with the aid of Chris Downey, the lab's programmer-analyst, offers several capabilities specifically tailored to the lab's needs. The computer performs spreadsheet and other analyses. It handles all cost accounting, provides test count projections, and tracks supply ordering patterns, indicating given quantities on hand. If the number of patient days goes up, the microcomputer points out which supplies must be increased.
"We can now project our costs so that we don't exceed our operating budget," Schofield says. The microcomputer has a record of the lab's spending patterns for the past three years and can accurately estimate any of its costs.
The laboratory developed a 10-factor program to handle cost analysis of structured charges. "We simply plug in the numbers and update the figures whenever necessary," Schofield notes. This gives the lab strong advantages when renewing vendor contracts. "By knowing the exact cost per test, we can quickly see how any price change--in reagents, say--will affect expenditures."
The microcomputer also helps in deciding on a major instrument purchase. It evaluates an instrument's cost with special emphasis on cost-effectiveness and rate of payback. These data greatly simplify selection and justification.
When fiscal questions do arise, Schofield says, "We can have the figures we need to answer them." Comprehensive monthly reports also keep top management up to date on laboratory activities. Best of all, the microcomputer has given the lab at Valley Children's Hospital more control of its financial destiny.
In short, the lab's investment is paying off. It took several months to develop the program and input that data, but since then the lab has saved months overall, cut budget time in half, and produced more valid data. None of that could have been accomplished manually in any length of time. With the computer, next year's budget "is going to be a piece of cake," Schofield says.
As Figure II has borne out, staffing and scheduling are a major worry, especially now that each work hour counts more than ever. Some 17 per cent of the panelists would like to turn staffing over to the microcomputer, and 24 per cent would delegate scheduling.
William Closson, Ph.D., chief of clinical services at 625-bed Brunswick Hospital, Amityville, N.Y., has computerized both functions, with uniformly successful results. Using modified Apple software with a TRS-80, the lab has created a fair schedule for weekend rotation, put an end to staff squabbles, and virtually eliminated excessive overtime, all in a short six months.
The system is basically simple, Closson says. The microcomputer maintains a record of previous schedules and approved requests for vacation time; the rest is left to mathematical probability. When the computer drafts the weekly schedule, it automatically assigns four technologists for weekend duty. Technologists work only their assigned shift--day, evening, or night--with no trading allowed. When the computer prints the schedule, it automatically assigns Thursdays before and after the weekend tour as compensatory time.
"If the TRS-80 says it's your turn, you pull the duty," Closson notes, adding that the computer is eminently fair and not swayed by excuses. Chemistry technologists generally work one weekend a month; those in other departments have a slightly longer rotation. Schedules, posted two weeks in advance, take only 10 minutes to prepare.
The microcomputer has also slashed unnecessary overtime. "Unless you want technologists to punch out in the middle of an assay, it's hard to eliminate all overtime," the director says. "However, you can keep it under control--for use, that's less than 50 hours per week for the entire lab, or an hour a week for each of our 50 technologists."
By inputting time card data, the laboratory can track weekly overtime patterns by day and shift. Six weeks into the program, Closson noticed a high rate of overtime on Sundays. At that time, only three technologists served on the weekend roster--not enough to meet the heavy demands of Sunday afternoon presurgical testing. The laboratory solved the problem by adding a fourth technologist and no further scheduling adjustments have been needed.
Staffing has been similarly stable. "We reassigned a few people in the beginning," Closson says, "and now the computer simply confirms that we're on the right track."
For all their versatility, microcomputers are only as useful as their software. At present, 65 per cent of the laboratories with microcomputers purchase ready-to-use software, 38 per cent design their own, 13 per cent have it custom-made by an outside supplier (Figure III). Eleven per cent also swap programs with other labs.
While many panelists are satisfied with commercial software, a number of others say it's too general for their laboratories' needs. "There aren't too many programs for the blood bank," comments Susan Stewart, laboratory supervisor at Detroit's 991-bed Harper-Grace Hospital. "We simply don't have the capability or time to develop our own, and even if we had several hundred dollars to spend on custom software, it's hard to explain exactly what we need to a computer expert who knows nothing about labs."
This software problem makes it hard, in turn, to justify purchasing a microcomputer. "Without the programs, we're having trouble convincing administrators that our bright ideas will work," Stewart says.
The major drawback to custom software is the price tag, panelists say. With prices ranging from $250 to $2,000, the laboratory at 384-bed Northwest Texas Hospital in Amarillo is doing some heavy comparison shopping, reports technologist Bruce Jones. In the meantime, Jones has developed his own quality control program fro evaluating lot numbers, now doing well in a trial run in hematology.
The Veterans Administration is in the process of developing a comprehensive software system to insure standardization among its 172 laboratory facilities, states Linda Akens, a cytotechnologist at the 1,088-bed Bay Pines (Fla.) VA Medical Center.
The laboratory staff at a 200-bed Philadelphia hospital developed a memory-tickler file for preventive maintenance, according to the laboratory director. "The daily QC person simply hits a function key, and the lab's PM schedule comes up automatically--and keeps coming up until we tell the computer that the work is completed," he explains. The lab saves $2,500 annually by printing its own test tube labels using a program developed by a staff technologist.
To help teach medical students, residents, and technologists all about anemia, the laboratory director at 606-bed North Carolina Memorial Hospital in Chapel Hill developed his own version of a video game. "He calls it 'Bloodhound,' and it includes 10 case studies," says Judith Watson, chief supervisor of the hematology lab. "The goal is to diagnose patients while ordering the fewest possible tests."
Some panelists exchange software on an informal basis, while others belong to specialized users' groups. Such groups may be affiliated with professional organizations; more often, they are operated by computer hackers. One, the Micribiology Computer Users Group, had its first meeting at the 1983 American Society for Microbiology convention and now has 452 members and a newsletter, reports Robert Manasse, microbiology/serology department head at 274-bed Memorial Hospital in Cumberland, Md. The College of American Pathologists has also established a software exchange.
At the University of North Carolina, according to Judith Watson, computer enthusiasts from all university departments meet to share information, trade programs, and help each other work the bugs out of their individual systems. Those with special areas of expertise also participate in a university-sponsored tutorial program.
Gary Jones, assistant chief technologist at 140-bed Queen of the Valley Hospital in Napa, Calif., points out that several information services and users' groups have sprung up in response to the ever-increasing popularity of personal computers. Those interested in obtaining a particular program or other data simply use a phone modem to contact the host computer, enter a prearranged password, and request the desired information. "Some of the commercial services cost as much as $100 per hour," Jones says. "The hackers generally provide the data for free."
Mostly, panelists pick up new software by trading with colleagues, both internally and outside their institutions. It is also increasingly common to borrow a program for a laboratory tryout prior to implementation. One blood bank supervisor says her lab does this routinely to avoid purchasing imcompatible or duplicate software. A chief technologist reports that her lab recently evaluated various data base and spreadsheet programs, found the best seller outdated and inflexible, and ultimately selected "Brand B."
A few nontraders question the legality of using "free" software, and with good reason. As with videocassettes, there's a growing market for pirated commercial software. Somewhere between patenting and pirating are the microcomputer sales representatives who "share" computer software, not necessarily developed by their firm, as a sales incentive or bonus.
Panelists, while generally enthusiastic about their microcomputers, are quick to point out other drawbacks besides the lack of decent software. Too many technologists and too little computer time is one common complaint. Respondents in some labs say that every technologist wants his or her own terminal. In others, bench personnel have absolutely no interest in learning, using, or even thinking about the computer. Some panelists cite a lack of administrative support as a major stumbling block, and several complain that they simply don't have time to master microcomputers properly.
Computer problems run a wide and frustrating gamut. "We're being renovated, and the power goes out frequently," says a laboratory supervisor at a large southeastern medical center. On top of that, she adds, the old and soon to be replaced laboratory information system crashes daily. "This totally disrupts our printing and storing functions, but we seem to manage."
A Midwest laboratory manager reports that theft is a major concern. "It hasn't occurred yet, but the portability and value of the small computers make it a definite consideration." To make sure the system doesn't wander, her lab bolts down microcomputers and locks away data disks.
Despite these large and small disadvantages, microcomputer use in laboratories is bound to grow, especially now that prospective payment has created a massive need for cost containment and sophisticated data management. Of course, as we pointed out in Part I of this special report, prospective payment also appears to limit the funds for microcomputer acquisitions. How big an obstacle is that?
An independent laboratory director didn't know, but he did have one suggestion--seemingly sound in light of our findings. "If you don't buy a microcomputer, at least buy stock in a company that makes one."
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||microcomputers in the lab, part 2|
|Publication:||Medical Laboratory Observer|
|Date:||May 1, 1984|
|Previous Article:||Microcomputers in the lab: the sudden boom.|
|Next Article:||Microcomputers in microbiology: a matter of special needs.|