Printer Friendly

Heighten efficiency with an integrated bar code system.

Heighten efficiency with an integrated bar code system

Implementing this worthwhile mode of identification requires an organized plan and plenty of interdepartmental cooperation.

Every day, laboratories match specimens to names, names to tests, and tests to results. Human error can creep in at every step. One way to improve the situation is to adopt bar coding, a system that is widely used to record prices in grocery stores and other shops, check out library books, track warehouse inventories, and identify airline luggage.

Even though manufacturers have been adding bar code readers to analyzers for about four years, the bar-coding capabilities of most laboratories remain too limited to improve efficiency, accuracy, or productivity. In blood banks, among the first hospital departments to use bar codes, applications have tended to be rather primitive. For the most part, blood banks have used this potentially far-reaching mode of identification solely to identify blood units. That is going to change, and for good reason: Bar coding is a versatile system that is widely applicable in the laboratory. * Early efforts. It was my observations at the supermarket, in fact, that led me to institute a bar coding system where I formerly worked, at the laboratory in the Department of Veterans Affairs Medical Center at San Diego. Our lab was among the first to exploit the full potential of bar coding.

In 1984, I constructed a system that integrated bar-coded specimen identification with our large automated chemistry and hematology instruments and the laboratory computer.[1-3] The result is that bar coding now saves the lab nearly $260,000 a year. Specimen identification is considerably more accurate, and efficiency and productivity have increased.

Other labs have since applied some of the same principles to blood collection and chemistry testing, with comparable improvements in efficiency.[4] Last year I joined the Department of Veterans Affairs Medical Center in West Haven, Conn., as chief of laboratory service. Among my first projects was to get the bar code system up and running. Early results indicate that in many ways it will be as successful as the program in San Diego.

Bar coding, while a powerful tool, cannot produce substantial cost savings or improvements by itself. I think of bar coding as an essential building block in the construction of a modern laboratory. The system must be carefully integrated with laboratory instrumentation and the hospital or lab computer system to be effective. Figure I shows a stylized diagram of our overall system and illustrates the interrelationships of its essential components. * Distributed test ordering. An essential step in achieving high efficiency is to implement distributed test ordering--that is, to have all test orders processed on the wards or in the clinics. From a management standpoint, it makes very little sense to squander the limited time of our laboratorians on the tedious process of entering patient information and test requests early each morning, just when a massive number of blood specimens arrive.

Most of these test orders are routine and have been placed by ward clerks, nurses, or secretaries on multiple wards or clinics during the previous 24 hours. It is far more logical to share test-ordering chores by having the ward personnel enter test orders immediately. This procedure spreads the work among many hospital staff members and over a much longer period of time. Test ordering via CRTs is easier and faster and eliminates the paper hassle and other problems associated with processing handwritten lab slips. * Phlebotomy. Another key to improving the lab's efficiency and service is to provide blood drawing on the floors at least three times a day. Many readers will say there is no way they can oblige without increasing the staff. Not true. The lab's newly improved efficiency--and a minor redefinition of jobs--makes it possible. Placing bar-coded labels on blood tubes at the bedside throughout the day, for example, more than offsets the extra phlebotomy rounds.

We print a blood drawing list along with the bar-coded labels three times a day. The list, which provides such basic information as the patient's name and location, specifies the tests ordered and the number of tubes needed. Full information labels show both text that can be read by people and the bar code. The code contains a unique accession or specimen identification number assigned by the hospital or lab computer.

Each morning, the phlebotomists take their lists and labels, draw the blood from the patients, and attach the bar-coded labels. Other rounds take place at 11 a.m. and 3 p.m.

For outpatients, the phlebotomist prints the bar-coded labels just before the draw, which takes place in the lab's phlebotomy area. As soon as the lab receives specimens drawn by ward personnel, technologists print labels and place them on the tubes.

More than 90 per cent of our tests are now performed on specimens labeled with bar codes. For miscellaneous low-volume tests, such as CEAs, AFPs, and ferritins, we use 10-part perforated labels to accommodate the aliquots of serum. The main label goes on the specimen tube; the aliquot labels show the patient's name and Social Security number and test done. Although special chemistry analyzers do not incorporate bar code readers yet, I believe they eventually will. When they do, all our specimens will be bar coded. * Chemistry. Once bar-coded tubes arrive at the lab, the only processing needed is centrifugation. Our chemistry analyzer is fully interfaced with the LIS in a true bidirectional mode. All patient information and test requests are automatically transferred to the analyzer without any intervention from a technologist.

After centrifugation, the technologist removes the tube's stopper and places the original clot tube in the instrument. The instrument then reads the bar code and performs the test. Results are sent straight to the computer. As soon as the results have been reviewed and verified, the data are available on CRTs throughout the hospital and clinics. Results are printed automatically in selected areas, such as intensive care units and emergency rooms. * How automatic? In researching bar code systems, don't be fooled by snake oil sales reps who claim their computer systems are bidirectional. Closer examination may show this isn't so. In some systems, patient information and test orders aren't transferred to an automated chemistry analyzer until a technologist logs on and requests the transmission. From an operational standpoint, the transfer should be invisible and independent of human involvement.

Both of our labs at Yale use the Hitachi 717 chemistry analyzer. This instrument is ideal for our purposes because it meets several essential criteria. It has a bar code reader, a serum level sensor that allows direct sampling from the primary clot tube, and random access selection of up to 35 different tests. Other features include refrigerated reagent storage, a high throughput of 750 tests per hour, and bidirectional communication. Reagents may be purchased from many commercial sources or prepared in house. * No more Stat lab. Integrating bar coding with our chemistry analyzer and LIS made it possible for us to eliminate our Stat lab altogether. An analyzer that's fast enough permits all tests to be done in the same location.

When a designated Stat specimen comes in, it is centrifuged immediately and placed in the instrument's specimen tray so that it is first in line for the bar code reader. If a routine specimen occupies that position, we simply move it to the next available slot.

The analyzer's high throughput and large test menu are essential factors in the system's success. As a bonus, erroneous IDs are a thing of the past.

The automated chemistry analyzer handles most of our therapeutic drug levels, including theophylline, phenobarbital, phenytoin, procainamide, and NAPA. We use EMIT technology for most of these tests.[5] We have developed a method for diluting the EMIT reagents with buffer, restoring NAD and G6P levels, elevating the temperature to 37 C, and extending the reading period over several minutes.

In our experience over the past year, we have found that the diluted reagents remain stable for at least four weeks. Equally important, we have to calibrate the procedures only once every three to four weeks. Controls are run just once on each eight-hour shift.

These therapeutic drug monitoring tests are no longer performed on a separate instrument at an extremely high cost. Running them along with the electrolytes and enzymes on the general chemistry analyzer saves a tremendous amount of money and personnel time, reduces specimen handling, and yields much faster results. This modified protocol cuts the cost of TDM to 10 cents per test and makes Stat testing available around the clock.

With regard to the debate over distributed versus batch testing, we've found that the traditional approach is no longer efficient for our laboratory. The lab offers much better service when technologists run as many tests as possible as soon as the specimens arrive rather than splitting the batch and storing some to test later. * Advantages. The many benefits of bar coding in the laboratory are summarized in Figure II. The testing at the San Diego facility became far less labor intensive; we were even able to trim several FTEs by staff attrition. I anticipate more modest gains at the West Haven site, which had already tightened the staff considerably before instituting bar coding.

At the bench, technologists report a dramatic improvement in working conditions. Testing used to entail a great deal of processing and specimen splitting to prepare aliquots for the Beckman Astra, Technicon SMAC, Du Pont aca, or Abbott TDx. Processing a 3 a.m. specimen requiring a large panel of chemistry tests plus theophylline and quinidine was extremely labor intensive. Stat tests were performed on the Astra, drug levels on a TDx. The specimen was then refrigerated for a few hours, with the remaining tests run on the SMAC during the day shift.

Such a specimen was handled many times by many technologists. It was poured into multiple specimen cups--all labeled correctly, we hoped--and analyzed on several instruments. Results were distributed irregularly over several hours, depending on when testing was completed.

Today, most tests are run on a single analyzer as specimens arrive. This approach eliminates the bulk of specimen splitting and the need to handle the same specimen several times or to move the serum from one instrument to another. Physicians are satisfied because they obtain results in a few minutes, even at 3 a.m.

Our substantial staff reduction was the major source of cost saving. With our former system, there had always been a great deal of physical movement around the lab. Things are much quieter now. It takes very little activity to generate the same amount of work.

The four major problems related to implementing bar coding are to find good printers, obtain the needed software, encourage interdepartmental cooperation, and combat that old laboratory standby, resistance to change. * Printers. The first major problem, poor visual quality of the labels, is the easiest to rectify: Choose appropriate hardware.

Necessary equipment includes a desktop computer for the instrument interface and two bar code printers. Many laboratorians looking to conserve their budgets buy the cheapest printers they can find. This is an unfortunate and shortsighted mistake.

Professional printers use thermal paper to produce clear, distinct bar codes. The ideal bar code fits conveniently on the label with plenty of room for the text that has to be read by human beings. In contrast, the large codes produced with dot matrix printers have irregular edges and ink irregularly distributed across the label. Ribbons must be changed every few days, but nobody ever seems to remember to do this.

The cost of professional-quality printers has plummeted over the past several years, making it easier to buy equipment of high quality. One interface computer and two professional bar code printers cost a total of around $5,000--far less than the $15,000 I paid just four years ago. Don't try to save money by forgoing a backup printer; the security is well worth the extra $2,000.

We keep our main printer in the blood drawing area where outpatient draws are processed. The backup unit stands ready in the lab. It's not idle, since we use it to print labels for specimens dropped off in the lab. * Software. It is impossible to design and implement a modern and efficient bar coding system without automated instrument interface software and software to drive the printers. Depending on circumstances, developing this software can be relatively easy or extremely difficult.

For example, only the vendor can make software additions to a commercial HIS or LIS. Purchasing customized software from most vendors costs time and money. Since customizing software is not a big priority for vendors, obtaining such a program can take from three months to a year. * Lack of cooperation. The greatest obstacle to attaining a highly efficient laboratory is inadequate cooperation between hospital departments. Full cooperation of the ward administration is essential in establishing a workable test ordering system.

In San Diego, we had trouble implementing ward order entry because of the middle management bureaucracy in the medical administration. The chief of the ward clerks, for example, had vetoed the plan, claiming the clerks were "too busy" to use bar codes. The clerks, on the other hand, thought bar coding was a great idea. The lab sidestepped such hospital politics by assuming responsibility for teaching ward clerks how to order lab tests on their CRTs and by providing day-to-day troubleshooting and support on the wards.

Equally important is to obtain full cooperation from the hospital or laboratory computer section staffs, whom you'll need to write the required software or to help the lab write its own. Since most computer sections are overworked and understaffed, getting customized software in a reasonable amount of time is not very likely. I was very fortunate at the VA. With access to the LIS and the full cooperation of the computer staff, I was able to write all the necessary programs myself. * Resistance to change. The final and particularly troublesome major problem in implementing a bar code system occurs in every lab: Laboratorians often pose tremendous resistance to anything new. Many find it more comfortable to cling to old procedures and habits just because "that's the way it has always been done." Clinical laboratories traditionally exhibit an incredible amount of internal inertia. To achieve a highly efficient and modern laboratory, managers and staff alike must be farsighted, willing to shake things up, and--yes--brave. * Gold Rush. To my mind, there is no question that labs will universally use bar coding in the 1990s. I think of it as the Laboratory Gold Rush of the future. Once labs discover just how much time and money can be saved, they're all going to want a bar coding system.

It isn't always easy to implement bar coding. Actually, the equipment is simple enough; it's people who pose the challenge. The key is to garner support within the lab and outside of it--the cooperation of the computer staff in particular is essential--and then mesh all these diverse components so that they work together smoothly. If you can accomplish this, the rewards may exceed your highest expectations. [Figure 1 and 2 Omitted]

[1]Neeley, W.E., and Griffin, S. Bar codes and the clinical laboratory: I. Introduction to bar code symbology. Informatics Pathol. 1: 128-132, 1986. [2]Neeley, W.E., and Griffin, S. Design and implementation of an inexpensive bar code identification system for Coulter counters. Informatics Pathol. 2: 118-130, 1987. [3]Neeley, W.E. Design of a bar code identification system for the clinical laboratory. Informatics Pathol. 2: 159-162, 1987. [4]Tilzer, L.L., and Jones, R.W. Use of bar code labels on collection tubes for specimen management in the clinical laboratory. Arch. Pathol. Lab. Med. 112: 1200-1202, 1988. [5]Sung, E., and Neeley, W.E. A cost-effective system for performing therapeutic drug assays: I. Optimization of the theophylline assay. Clin. Chem. 31: 1210-1215, 1985.

William E. Neeley, M.D. is associate professor of laboratory medicine at the Yale University School of Medicine in New Haven, Conn., and chief of laboratory service at the Department of Veterans Affairs Medical Center in West Haven, Conn.
COPYRIGHT 1990 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1990 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Neeley, William E.
Publication:Medical Laboratory Observer
Date:Mar 1, 1990
Previous Article:How to support your supervisor.
Next Article:We were guest stars at the science center for Lab Week.

Related Articles
Strategic planning for an integrated bar code system.
Joining the bar code revolution.
Creating a bar code chemistry system.
Wristbanding for positive patient ID.
Bar coding: the ideal system.
Efficiency gains with bar coding and interfaced instruments.
ID Detective can help owners monitor visitors to buildings.
Surgical Instrument Management.
Sessions improve production output for major soft drinks manufacturer.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters