Printer Friendly

An LIS is not all pluses.

An Lis is not all pluses

Laboratories install computer systems for a number of reasons and usually after long and careful study. Our lab was no exception, but the outcome was disappointing.

Time and motion studies show we need about 5 more FTEs to handle extra work caused by use of the computer. In addition, we've had to restructure the work done by both our clerical and technical staffs, and there's now a full 24-hour delay in reporting results back to the floors--two days if charting of results is included.

Added costs attributable to the laboratory information system totaled $292,000 in the first year alone. If we can't make some significant changes soon, the costs will mount to approximately $1.2 million over five years.

This all makes it sound as if we didn't pick the right system, but having talked to others in laboratory management, I know our experience is by no means unique. We are seeing the hidden flaws in laboratory information systems, problems you can expect to encounter no matter which system you purchase.

Bridgeport Hospital acquired and installed its LIS last February after a lengthy search. The system seemed to be a natural solution to a brewing problem: We expected sustained growth in the volume and complexity of laboratory test requirements, and were appalled at the prospect of an even denser paperwork jungle than the one already surrounding us.

We expected the computer to help make laboratory reporting more timely and medical records more orderly while also improving the quality of management information. We carefully studied systems that were within our budget and finally chose an LIS that seemed to meet our needs and allow for growth.

After a 2$p1/2$p-month training and installation period (including parallel operation of the computer and manual systems for three weeks), we went live with the computer alone. In another month, we were about 80 per cent efficient in the use of the system and becoming increasingly aware of its disadvantages and limitations.

The first conclusion we reached was that installation of an LIS has an immediate and incalculable impact on the organization of the laboratory. A previously well-organized laboratory will become disorganized and probably have to be restructured to make work flow compatible with the computer. Ideally, it whould be the other way around, but it never is.

We found that the LIS increased rather than decreased our staffing requirements, particularly in the clerical area, thereby reducing the lab's productivity. The first signs of the problem were fatigue, confusion, and lower morale among the clerical staff, all due to the increased workload. The LIS is not hooked up either to the hospital information system or to terminals at nursing stations, so the clerical staff has to enter demographics on approximately 450 patients and 2,000 specimens every day, phone State to floors, and field calls from the floors for results.

The computer system has not eliminated manual logging of specimens. Even though an analyzer with a bar code reader can pick up a specimen ID, only a flow of information from computer to analyzer would automatically tie that ID to patient demographics. Right now, this must be done manually. The interruption of work flow by Stat tests makes it even more imperative to verify patient IDs. With these manual procedures retained in the presence of the LIS, there is at least a 50 per cent increase in the clerical component of the test workload.

Such considerations quickly made it clear that we had to reorganize the clerical staff based on revised concepts of workstation definition. We identified the tasks the clerks had undertaken to support the system, then developed a workstation model to distribute specific assignments and improve the work flow. For example, one clerk now enters all new patients, another handles outpatient entries into the system, and a third phones the floors with results taken from a Stat printer.

We also had to put on an additional clerical person at night to take care of printing reports. Results used to be written on "shingle' slips that appeared on the wards within about three hours. Now, with overnight printing, results don't get back to the floors until the next day.

The printing can't be done during the day because it would compete with all the other work going on and cause a slowdown. This is true for most computer systems: Because of the way data are collected, report printing must be batched. In our case, the mainframe database is tied up for about five hours of printing.

Granted, we added to our LIS headaches with problems of our own-but what hospital doesn't? Complications were caused by a same-day surgery program that blossomed just after we installed the computer system. Demand increased for presurgical testing on the day of surgery. With surgical patients starting to arrive at 7 a.m., there was a direct conflict between the hospital's and the same-day surgery program's needs for laboratory services first thing in the morning.

This led to surgery delays of 30 minutes and more, at a cost of about $300 per half-hour beyond the scheduled time. The only way to solve the problem was to move a chemistry analyzer and a technologist into the one-stop surgery department.

Preadmission testing might have eased the strain, but at the time we installed the LIS, collection sites for such testing had not been established around the city. (Many patients found it inconvenient to come to the hospital for such testing.)

Phlebotomy was even harder hit than the clerical area by the advent of the LIS. We were already short two phlebotomists out of an allocated staff of eight when the computer came on the scene. This shortage had caused long delays in getting specimens to the laboratory, even though medical technologists were lending a hand on the floors. The delays lengthened when phlebotomists and others had to start logging on to the LIS as they brought specimens into the laboratory.

We were able to patch together a solution by hiring some temporary phlebotomy help and using student runners to get specimens to the laboratory.

One of the system's major flaws is in the way it interfaces with the various instruments. We can't make up a workload list for a specific analyzer because the computer doesn't distinguish between, say, glucoses from two different instruments. This also affects quality control since there's no way to review results by instrument.

As a result, we had to reorganize the work in chemistry. For example, we will run a big batch of tests on a high-throughput instrument, then have a technologist verify each determination and manually, with one stroke per result, pass the data on to the computer. We are reaching the limits of what we can do to compensate for the problem, however. This is one of the areas we re working on with the manufacturer to resolve to our satisfaction.

What should be clear by now is that one can't even consider the information management value of an LIS until the work flow problems are solved. It's also clear that such a system creates significant, essential clerical costs. Even more important, with the addition of extra staff and reorganization, most laboratory information systems still offer a lower standard of performance than good manual systems. So we must ask ourselves whether we can find a justification for acquiring an LIS. At this point, I think not.

I have made a thorough analysis of our LIS, including a study of the costs it adds to our operation (Table I). As I indicated earlier, these costs are dramatic-- $292,000 the first year, dropping to $224,500 in subsequent years for a five-year total of nearly $1.2 million, at very conservative estimates.

In the DRG era, the laboratory must provide timely results without wasting resources. A laboratory information system can be justified if it has a positive impact on work flow, reduces the fixed overhead costs of operation, and perhaps even helps the laboratory create new services effectively and efficiently. In addition, the lab must be able use the computer to examine 1) the efficiency, quality, and value of the services it provides and 2) the relationship of lab services to the medical activities it supports.

This is not possible with traditional mainframe computer systems. The only solution is the introduction of relational database technology and its real-time application to lab operations. Some of these features are available right now, but not all of them in a single system. If they were all available, I estimate they would reduce our projected added costs to under $150,000 over five years.

Unfortunately, laboratory information systems as they now exist cannot be justified either on a cost or a medical basis. Indeed, the LIS conceptual framework may put the laboratory at a disadvantage in the DRG era.

Table: I Computer-added costs in chemistry
COPYRIGHT 1986 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1986 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:installing a medical laboratory information system
Author:Bernstein, Larry H.
Publication:Medical Laboratory Observer
Date:Nov 1, 1986
Previous Article:Debugging a laboratory outreach program.
Next Article:Quicker QC on a small microcomputer.

Related Articles
Strategic planning in selection of a lab information system.
Laying a solid foundation for LIS selection.
Saving time with combined microcomputer applications.
Find the right LIS ... with EASE.
Available in most labs, fully utilized by few.
Creating a bar code chemistry system.
Laboratory information systems: expensive, complicated and cranky.
How well does LIS reach all 4 corners of the lab?
Laboratory information systems: continue to add features that contribute to maximizing personnel and cost containment. (Product Focus).
Traditional lines blur between clinical and AP lab systems.

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