REX: Extralaboratory cycle of information processing from laboratories to care units

SJ. DARMONI, CRIH Haute Normandie, Bld Gambetta, F76031 Rouen Cedex, France; P. MASSARI Département de l'Information Médicale, CHRU de Rouen, 1 Rue de Germont, F76031 Rouen Cedex, France; P. ALLAIRE, CRIH Haute Normandie; JL. CAFFAREL, CRIH Haute Normandie; M. MONCONDUIT, Service d'Hématologie, CHRU de Rouen; M. BALDENWECK, CRIH Haute Normandie; P. HECKETSWEILER, Policlinique, CHRU de Rouen.

Abstract: REX, developed in the Rouen University hospital, allows the complete computerization of extralaboratory cycle of lab test numerical results from laboratories to care units. This cycle includes a real time transmission on terminals as soon as the test is realized and a batch print every evening. In laboratories, most of the lab tests results are directly acquired from automated devices. This application is integrated into our hospital information system. It is operational since the second semester of 1989. Every day, 18 000 lab tests results are transmitted by REX and 10 000 transactions by health professionals are done. Beyond amelioration of speed and reliability of the extralaboratory cycle of lab test results, REX decreases work time linked to phone transmission in laboratories and care units.

I. INTRODUCTION

The transmission of lab tests results, from laboratories to care units, still depends most often on phone calls for emergency tests, and on paper for other tests, which means errors and lost of time.

Data processing with its communication capabilities can improve the extralaboratory cycle of lab tests results. A software (REX) integrated into the Hospital Information System (HIS) has been developed and has been used since the end of 1989 for the transmission of numerical lab tests to care units of the Rouen teaching hospital.

II. DESCRIPTION OF THE REX SOFTWARE

A. Conception

REX software has been developed in cooperation with physicians, biologists and computer scientists. The first step was the modification of the laboratory information system (LIS). For this purpose, since 1988, we have been using a French software (LAM6 of the SARIC company) which is implemented on a Bull minicomputer (DPS6). This computer is connected to 50 terminals and 15 automated devices in the five main laboratories.

The lab tests results produced by automated devices are directly integrated into the software. For all lab tests, LAM6 makes the first step of validation, taking into account (a) the normal range of values, defined by biologists according to age and sex; (b) prior results of the patient and (c) clinical findings. For each result out of range, biologists must validate them.

REX has been developed, using Cobol in the CICS environment. REX manages information between the laboratory computer and the main computer of our hospital (IBM 3090 200J in the MVS/ESA environment). This computer manages a network of 250 terminals and 150 microcomputers located in care units and offices.

B. Functionalities

REX is interfaced with our administrative patient management software (GEMME). This functionality is specially useful when the patient is transferred to another care unit during the laboratory cycle of information.

REX has two mains functions: a real time display on screens located in care units and a batch print of lab tests results.

1) display of results: Every REX end-user owns an identification which allows, after input of a password, access to authorized HIS applications. After choosing REX, the end-user has to specify which authorized care unit he wants to work with. Then, he will see on the screen the list of patients present in the care unit, ordered by room number. He can obtain with a specific PF key the list of patients present the previous weeks. REX displays the results of the last known sample. So, results can be seen in real time. An on-line help is available about normal range of values and units of numerical results.

Two display modes are possible: a chronologic mode and an historic mode where the last eight samples are displayed in columns. The lab tests are separated in three categories A, B and C. Each category is displayed in a different screen: category A, for "routine" numerical lab tests such as blood cell count or creatininemia; category B, for other numerical lab tests such as protein electrophoresis and category C, for literal lab tests and comments. The status of results, normal, out of range, not validated is displayed in different colors. The concept of validation is more important than the concept of normal range. A non-validated result will be sent a second time to REX after validation by a biologist. As soon as a request is registered in a laboratory, this information appears in REX and every requested and registered lab test will be marked.

For the two modes, three types of navigation are possible: display of patient's previous lab tests, of patient's other categories lab tests, and of next patient's results. Chronologic mode displays lab tests in function of the sample time. The first screen of the chronologic mode corresponds to the latest known sample. Historic mode displays the last eight known samples of numerical lab tests.

2) print of results: The print mode is very similar to the historic mode. This print is done on a batch mode around 4 pm on seven laser printers located in each building of the v hospital. Only validated lab tests are printed.

III. RESULTS: IMPLEMENTATION AND EFFICIENCY

The operational implementation of REX has been realized in the second semester of 1989 in six pilot care units and extended to all care units of the Rouen University hospital at the beginning of 1990. This hospital comprises three centers separated by 5 kilometers, with 39 care units and 2500 beds. Each day, an average of 2200 patients files are processed with an average of 8 lab tests by patient.

In October 1991, an average of 10 000 daily transactions by health professionals were done. Real time access to lab tests of a patient is available until 90 days after discharge from the hospital.

REX allowed a significant decrease of work by laboratory secretaries; before REX, one person was affected to phoning emergency results; today, these phone calls nearly disappeared. REX improved reliability by minimizing the rate of transmission errors, as phone calls were replaced by computerized transmission and handwritten copy of results by hardcopy.

IV. DISCUSSION

There are two cycles of information processing in LIS [ 1]: intralaboratory cycle, and extralaboratory cycle from laboratories to care units.

Intralaboratory softwares, most times, manage independently the informations of each laboratory and more rarely take into account informations from several laboratories [ 1]. LAM6 manages this intralaboratory cycle. A more or less intelligent validation of results could be integrated using expert systems interfaced with the laboratory management software [ 2, 3].

Extralaboratory cycle of information processings are rarely reported in the literature [ 1, 4, 5]. The computerized transmission of lab tests results increases the quality of the extralaboratory cycle and decreases the time lost during this cycle [ 6]. It can rely technically on hospital network, phones lines and videotext. The goal of computerization could be display or print of results, or both as with REX.

The ergonomics, specially the absence of input on the keyboard and the presentation of the results on the screen using different colors, and the quality of the terminal network in care units are necessary to implement with success a software to manage the extralaboratory cycle of information. In our experience, owing to the 400 implemented terminals, and although the number of beds in the Rouen University hospital (N=2500) exceeds its American equivalent, and are dispersed in three locations, the operational implementation of REX was easy, imposing a very light training for a limited number of end-users.

In our experience, the integration of extralaboratory cycle of information processing into the HIS allows a simple connexion, common with the others applications of the HIS, it respects the security rules defined in our hospital: each end-user has an unique identification on the mainframe, with a specific password. . This integration allows access to lab tests results from the DIAMANT application which is in charge of the common medical record.

Benefits of this software are difficult to estimate objectively and in the aggregate. Care quality has been improved by the decrease of extralaboratory cycle time, of lab tests to care units, and by the minimization of transmission errors. Time linked to phone calls are drastically decreased in care units and principally in laboratories: the redeployment of one secretary by laboratory and the suppression of one Sunday resident duty prove it.

Therefore in our experience REX is a performing system. It has been set up with success in 1992 in three other French hospitals. It is limited to the transmission of numerical results to care units but it should take into account literal results. In the future, computerization of lab tests prescription which needs, according to us, the computerization of all orders, will permit the entire extralaboratory cycle of information processing.

REFERENCES

[1]. JW. SMITH, JR. SVIRBELY, "Laboratory Information Systems," In: Medical Informatics, Computer applications in health care, Shortliffe EH & Perreault LE Eds, Addison-Wasley, 1990, pp 273-297.
[2]. DP. CONNELLY, ST. BENNETT, "Expert systems and the clinical laboratory information system," Clin. Lab. Med. 1991; 11: 135-151.
[3]. T. GROTH, H. MODEN "A knowledge-based system for real-time quality control and fault diagnosis of multitest analyzers,' Comput. Methods Programs Biomed. 1991; 34: 175-190.
[4]. HL. BLEICH , C. SAFRAN, WV. SLACK, "Departmental and laboratory computing in two hospitals," MD Comput. 1989; 6: 149-155.
[5]. KE. TATE , RM. GARDNER , LK. WEAVER, "A computerized laboratory alerting system," MD Comput. 1990; 7: 296-301.
[6 ]. K. MORITA, K. IKEDA, "Evaluation of a microcomputer-based clinical laboratory data acquisition system linked with a minicomputer-based patient data management system," J. Clin. Monit. 1988; 4: 48-58.