Printer Friendly

A database design for a laboratory handbook.

Last August, when Iraq invaded Kuwait, I was teaching in the clinical laboratory science program at King Abdulaziz University in Jeddah, Saudi Arabia. I was just completing a two-year stint as assistant professor on assignment from the School on Health-Related Professions of the University of Alabama at Birmingham, under a contract between the university and the Joint Commission for Economic Cooperation (JECOR). JECOR signs contracts with the Saudi government to provide U.S. expertise and training to Saudi nationals under programs that are fully funded by the Saudis. (JECOR is still going strong.)

I was allowed in the lab but not in most other buildings on that side of the university, which is for men only. Besides teaching, my responsibilities at the university included acting as a QA consultant to the chief technologist at King Abdulaziz University Hospital laboratories. Among our projects was to update the handbook used by lab staff and by physicians and nurses who ordered lab tests and received results.

The information in the handbook, a sparse collection of unnumbered typewritten pages shoved into the backs of drawers at nursing stations, was three years old. Each lab section had written its own little chapter, listing the tests offered and the type of specimen preferred for each.

Unfortunately, the handbook had not done its intended job. The lab was receiving an average of three or four unnecessary phone calls a day and the wrong specimen type about once a day. A comprehensive handbook that was easy to use, we believed, would improve the situation.

The chief technologist and I held discussions with a committee of the eight physician consultants who managed the various lab sections. (Other than consultants from other countries, there are currently no Saudi Arabian clinical pathologists.) After two or three months, we had outlined our goals. These were to improve the handbook, to create a system for yearly updates, and to reduce the inappropriate specimens and phone calls received by the lab.

* Database. In considering ways to achieve these goals, I began to realize how useful it would be to create on my PC a database of the tests and specimen requirements for the handbook. Once the information was entered into the computer, it could be updated easily each year or as needed.

For this task I used dBase IV (Ashton-Tate Corp., Torrance, Calif.), simply because it was available at the university. As it turned out, arbitrary key combinations and inadequate on-screen help made dBase difficult to learn and use. The discussion that follows will illustrate features of my database design in a way that would allow any relational database software to be used in setting up a similar handbook.

* Structure. The database structure that I created for the project can be seen in Figure 1. I used the same structure for all lab sections, which varied in size and type from histology to parasitology.

Two elements of the design are functionally related: the use of multiple database files and the use of code names. Both elements were created to speed data entry and editing.

Using a separate database for specimen codes and names, for example, facilitated data entry. To save data entry time, I assigned each specimen type a one-letter code: "B" for whole blood, "S" for serum, and so on. Urine was a special case. I assigned "U" to random urine tests and gave two-character codes to timed ones, such as U2 for a two-hour urine test. When the timing itself required two digits, such as 12- and 24-hour tests, we coded them without the "U" as 12 and 24.

When completed, the database contained 198 kinds of tests, of which about 120 (roughly 60%) were to be performed on serum samples. Typing "S" instead of "serum" therefore saved 480 keystrokes.

Another example of this keystroke saver was the test tube database. Each tube type was entered once and assigned an abbreviation. Example: A mere four keystrokes, 5GRN, will order a test that requires a 5-ml green-topped tube containing lithium heparin. Using a separate database for tube types streamlined data entry while paving the way for changes in the future.

* Linkage. Another reason for using multiple databases and code names was to incorporate flexibility into the report layout. Relational database software can link field names in two or more databases. The field name TUBE-CODE would identify data in the tube and test databases. Consequently, the printout of a test result report can include the type of preservative as well as the name of the tube used. The code name itself is printed only if desired.

Including both the code and name for specimens, tubes, lab sections, and tests in the relational database saves keystrokes while providing flexibility for the report layout. For example, the chief technologist decided to print the specimen name but not the specimen code for each test. To avoid confusion, however, since many doctors and nurses commonly refer to tests by test code (e.g., CK) rather than by name (creatine kinase) or vise versa (glucose, not "glu"), both the test code and the test name would be printed in the handbook.

* Field names. The fields in the test database used to identify panel or timed tests (PANEL_TIME and CONTENTS, listed fourth and fifth under "Test database," the third subsection of Figure I) were created to inform (or remind) physicians what tests would be performed when they checked a given box on the requisition slip. One panel test, the cardiac profile, consisted of CK, AST, and LDH. Another, the CBC, included a hemogram, platelet count, and differential. Examples of timed tests were glucose and lactose tolerance tests.

The comments field in the test database was created for miscellaneous remarks about a test. These might include clinical indications or laboratory scheduling requirements. The memo field type--a special kind of character field in which a large number of characters may be stored for extensive notes--permits reference ranges to be stored for each test. It was decided early in the process, however, that ranges would not be included in the handbook at that time because studies were lacking on reference values for the Arabs who lived in the area. For the time being, the lab would use the levels given on package inserts. Yet everyone could see the value of adding that information in future editions, so we allowed for this.

* Field types and widths. The choice of field type--character, logical, float, or memo--is based on the kind of information to be stored. Character fields can store letters and numerals. Logical fields store yes and no or true and false responses.

The float (short for "floating point") field type stores only numerals, to a predefined number of significant decimal places. The only category for which we used a float was for the cost of the test (third from the bottom under "Test database" in Figure I), allowing for a total of nine numerals, including two digits after the decimal point ("9 Dec 2").

Field widths were based on trial and error. We wanted to keep the columns narrow, yet minimize confusing abbreviations and include all important data.

The department database was expanded beyond what is shown in Figure 1. Additional fields were used for such data as hours of operation, phone numbers, and names of the medical consultant(s) and technical supervisor attached to that section.

* Gathering data. Once the database was designed, each section of the laboratory compiled the necessary information. We provided a simple form to facilitate the gathering of data consistently throughout the lab. These forms were returned to me. As time permitted, I entered the information into the database.

First, I looked through all the forms for specimen and tube types and entered them in the specimen and tube databases. Then I entered the test information using the codes that had been defined earlier in the specimen and tube databases.

After completing data entry for each section, I printed a copy of those records for that section's review. Because the laboratory sections returned their original forms and approved the pages I had completed at various intervals, when it was convenient for them, I was never inundated with stacks of entries and edits at any one time. I was able to enter and revise the 198 tests in the final database within a few months.

* Layout. An example of the proposed layout of the laboratory handbook is shown in Figure II. The field descriptions appear in the first column. The sequencing of the fields is defined in the report layout function of dBase IV.

If desired, individual fields can be marked not to print. Although we used tube codes for ease of data entry, for example, we commanded only the tube names and preservatives to print out. Similarly, department and specimen names but not codes were printed. Since the report layout allows sorting on a field, it would be possible to print an alphabetical listing by test code or name.

* Format. For the format, we chose a compact size and shape that would slide easily into the pocket of a standard-issue white hospital coat. The pages would be printed in 14-point type on a laser printer and taken to the medical illustration department at the university for 72% reduction, duplication, and binding into spiral-bound booklets. They would then be distributed around the lab and to the medical and nursing staffs at meetings we had already set up.

Because we planned yearly updates, we decided to print only a few more copies than we actually needed--about 75. My husband, a graphic designer working in the university's medical illustration department, designed an attractive cover on desktop publishing graphics software.

Everything was going nicely when the Persian Gulf upheaval intervened. Although my husband and I had thought of spending another year there, we changed our minds and returned home in September 1990. To my amazement, I learned a few months later that a U.S. laboratorian had been found to replace me in January 1991, as international politics heated up further.

When I left Saudi Arabia, the record layout still wasn't finished. My next step would have been to send the report to a file on disk and then to print the pages via WordPerfect (WordPerfect Corp., Orem, Utah). We planned to use the search-and-replace function to change Y to Yes and N to No, then run a final spelling check before printing out the final pages.

The university was shut down during the war and reopened in April. Classes will continue through August to make up for two lost semesters. Packing all that material into a few months hasn't left much time to get back to the handbook. Meanwhile, the final pages are under review.

Like other projects in the country, work on the handbook will resume as soon as possible. I hope the lab sends me a copy. Jean Houger Hengesbaugh, M.S., CLS(NCA) The author is a Salt Lake City--based systems analyst with Superior Consultant Company, Inc., Farmington Hills, Mich.
COPYRIGHT 1991 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Hengesbaugh, Jean Houger
Publication:Medical Laboratory Observer
Date:Jul 1, 1991
Previous Article:Promotional health screening as a marketing tool.
Next Article:Cross-training: boon or bane?

Related Articles
Take in 5600 foods in one sitting without gaining a pound.
Quality improvement in the information age.
FBI Laboratory Publications. (Focus on Technology).
Our machinery database is live online! (Editorial).
NEHA Publications Clearance Sale.
Nist publishes online e-handbook of statistical methods. (General Developments).
World's largest MSDS database--fast internet access.
Genetically Engineered Mice Handbook.
EDA for IC implementation, circuit design, and process technology.
CRC handbook of enthalpy data of polymer-solvent systems.

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