Printer Friendly

How we automated our hematology lab; careful preparation enabled this laboratory section to computerize CBCs, differentials, and RBC morphology without a hitch.

Our laboratory was guilty of discrimination for years. Chemistry enjoyed the benefits of a minicomputer for result reporting, work listing, statistics, and billing while other sections operated manually. Today, we're correcting that inequity with a computer system of our own design, now at work in hematology.

I described how we developed our chemistry computer system in the May 1978 issue of MLO ("Our Lab's Minicomputer Does More and Costs Less"). We had time-shared for six years before that and had come to relyon the computer to increase the speed and accuracy of our work. We planned to expand use of our leased NCR 9040 minicomputer with 128K memory to other sections, especially hematology, which together with chemistry constitutes 80 per cent of our workload.

But in mid-1983 there were still three obstacles to expanded computer operation. First, we had long-standing plans to replace our 10-year-old Coulter S. It didn't make sense to buy an instrument/minicomputer interface until we chose a new automated hematology system and had administration approval to buy it.

Second, the alternative to an interface--entering all the parameters of the CBC, differential count, and RBC morphology manually through a video display terminal (VDT)--was ruled out as too burdensome. And third, it would have been difficult to justify the cost of hiring outside software experts to design and write the programs.

Nevertheless, the prospects of higher productivity, a reduced error rate, and other benefits beckoned. Stringent cost control legislation in our state hastened the decision to computerize hematology. (For an explanation of Massachusett's third-party payment changes, see "Is This Test Really Necessary?" by Dr. Bernard Statland, MLO, May 1983.)

The first hurdle was passed when we received authorization to purchase the Coulter S-Plus IV analyzer. That solved the problem of the interface, too.

To mee the computer software challenge, we got administration approval for assistance from our 294-bed hospital's data processing staff. Two programs had to be developed--one to process the hematology instrument's results and the other to record differential counts.

The plan required an initial outlay of less than $2,000. All we needed were two video display terminals at $600 each and a microprocessor costing $500 for the instrument/minicomputer interface. It wasn't difficult to get approval for those expenditures.

While waiting for delivery of the VDTs and the interface hardware, we constructed the hematology data base. It contains the names of our hematology tests, their reference values, charge codes, and delta checks. The computer already held such information as hospital room and physician codes in the chemistry data base.

Each test was given a mnemonic code name. Keeping the names as succinct as possible reduces keyboard time for each requisition. For example, we labeled the complete blood count C; the factor VIII assay, F8; and prothrombin time, PT.

We designed worksheets and patient comulative reports (Figure I). The reports also accommodate manual entry of results fromless frequently ordered tests, such as coagulation assays. In addition, abnormal results are flagged with asterisks. The formats essentially followed the model of our tried and tested chemistry system, which lessened the odds that the program would be plagued by bugs.

We chose pink paper with a white pin stripe for the hematology reports. Chemistry printouts have the same design on blue paper. The two colors quickly distinguish the source of a report.

We had a month before the instrument/minicomputer interface software would be ready. That gave us time to train technologists in use of the system. They worked with a dummy disk pack during slack work periods, manually entering results for fictitious patients. This mock exercise helped allay the anxiety felt by many staff members who were unfamiliar with computers. Actual patient files would have raised fears of making mistakes.

After individual training sessions, technologists were left to work their way through the system step by step, discovering and correcting their own errors. I remained in the background to answer questions. This training took longer than other methods might have, but I believe it made the learning stick.

Enthusiasm snowballed. Technologists made small wagers on the best way to solve procedural problems. They induced us to put up a bulletin board on which they could post common questions and answers about the system.

Meanwhile, the programmers were writing the software. They knew that our principal concern was to operate the section in our customary manner, without sacrificing speed or accuracy. To help them understand how we operate, we invited them into the lab for hands-on experience running specimens through the Coulter and simulating a differential count. (We also hired two consulting engineers, who modified the instrument interface to meet the mini-computer's specifications.)

The programmers delivered the software on schedule, and we were ready to go live by the beginning of October 1983. When we did, there wasn't a single bug in the system--our experience in computerized chemistry had paid off! Today, we routinely process all CBCs and differentials using the computerized system. Figure II shows the system's configuration. I'll describe how it works.

Processing begins with logging of the specimen. Every specimen gets an accession number, beginning each day with 1. Hematology averages 125 specimens a day, so to simplify this procedure and insure that no specimens share a number we have a hand-held, consecutive numbering stamp machine.

In typical operation of the hematology analyzer, the operator introduces the specimen and puts the requisition in the instrument's printer. Thirty second later, CBC results are printed on the requisition, along with a Coulter test number.

That second identification number is the key to our system. It tags the results while they're held in the interface microprocessor, or buffer, until they're sent to the minicomputer for storage or reporting. The buffer permits the operator to run a batch of specimens without having to pause and perform keyboard work after each CBC.

With the instrument interface, the operator can also perform a Stat CBC and rush the result to the floor before logging in the specimen, because the patient is already keyed to a unique Coulter number. As planned, this is a system that we control--not one that runs us!

If a technologist decides that a CBC needs to be rerun, the patient's accession number is written on a blank requisition and inserted in the printer. The specimen is reentered and run. Then the operator decides which set of results to record.

When we are ready to transmit CBC data to the minicomputer, using either of the video display terminals, we call up the results entry program. The accession number is entered, and the VDT displays the patient's demographic information.

Next, the technologist enters the instrument's test identification number taken from the printed requisition. The screen displays the results, held in the microprocessor buffer, along with the patient's previous CBC result retrieved from the minicomputer (Figure III).

The technologist confirms the correctness of patient and test data, accepts the results, sends them to the minicomputer, and moves to the next requisition slip--all in a few seconds. Although it isn't necessary to assign an accession number before running a specimen through the hematology analyzer, the operator can't store a patient's test data until that has been done. Another program feature displays the day's Coulter numbers and each patient's results.

The program for performing the differential count was more difficult to design. We conceived of the VDT and keyboard as a tallying instrument.

To start with, the technologist calls up the differential count program and enters the accession number. The computer displays the patient's demographic information--name, age, room number, physician, etc. The technologist enters his or her initials and the VDT displays the results of the patient's Coulter analysis and previous differential counts (Figure IV). Then it's ready to accept the latest differential count.

We assigned one letter key to each type of cell--N for segmented neutrophil, for example. A small adhesive tab with each cell's name is attached to the corresponding key. With the microscope at one hand and the keyboard and VDT at the other, the technologist counts 100 cells, continually pressing the different cell keys to keep a running tally. The final count can be revised before it is recorded if the technologist makes an input error.

the platelet estimate line appears next. The operator strikes the return key for normal, I for increased, or D for decrease. This entry will be recorded alongside the Coulter platelet count result on the CBC, with enough space for a comment.

The last entry is for RBC morphology. Normal morphology is noted by pressing the return key. If morphology is abnormal, the technologist hits N, which prints "See note," and enters the morphology description by depressing the assigned key. The numeral 1 key displays the word "anisocytosis," for example. The degree of abnormal morphology is then entered by striking a numeral key--pressing 2, for example, will print "2+."

The program continually guides the operator by flashing messages at the bottom of the screen and displaying a master list of cell types and corresponding keys.

Once all results are entered, the technologist can enter interpretive comments under each test result and correct the differential count. Finally, the computer operator presses P to send the results to the patient's file in the minicomputer.

Today, hematology enjoys the benefits of computerization long available to our chemistry section. The departments share the minicomputer, and we combine their results on a number of documents, such as outpatient and interim ward reports, although not on patient charts.

Besides yielding fast, standardized, and accurate results, the system is simple to operate. It makes it possible to run the hematology section with only two technologists when vacation or sick time cuts into staffing. We have also eliminated most transcription errors.

Paper costs approximately $30 a month--the system's only ongoing expense. On the other hand, we have accomplished a major cost saving. A one-day audit showed that we were losing as much as $20 a day in unbilled charges through clerical errors or when additional tests were ordered by phone without any billing notation. Now charges are automatically added in for all tests recorded on the computer, and a complete daily report for the billing department is printed out during the night shift.

What will the future bring? Plans call for a hospitalwide computer system that will place VDTs on every floor. The nursing staff will then be able to send test requests or call up results by VDT. This should further streamline our work and eliminate even more errors.

Our homemade hematology system has turned out to be more than palatable. In fact, we think it's a blue-ribbon recipe for meeting today's demands for accuracy and productivity in the lab. The key ingredient was a pinch of resourcefulness.
COPYRIGHT 1985 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1985 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Zwirko, Michael A.
Publication:Medical Laboratory Observer
Date:Sep 1, 1985
Previous Article:How to handle a product problem; when a lab product doesn't perform as it should, alert the manufacturer.
Next Article:A competency-based orientation program for new procedures.

Related Articles
Streamlining blood counts with a microcomputer.
kAn in-house proficiency survey for WBC differentials.
Internal proficiency testing for hematology.
A redesign of RBC morphology reporting.
National survey results: automation is everywhere.
Streamline your automated hematology laboratory.
Streamline your automated hematology laboratory.
Streamline your automated hematology laboratory.
Implementing preanalytical automation.
Size up this critical medical laboratory automation component.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters