We created our own data base management program for microbiology,using Applesoft Basic language.
Although it is slower than commercialsoftware, it is an inexpensive way to maintain current statistics ondepartmental activity, including organisms cultured and susceptibility patterns. Our microbiology department serves a 500-bed medical center.Statistics, compiled manually, had been dropping lower and lower on thelist of prioritiers after prospective payment led to layoffs andattrition. When up-to-date, monthly susceptibility summaries could nolonger be provided to clinicians and the infection control committee, weturned to our Apple II Plus. Till then, The microcomputer had been usedexclusively for automatic recording of data from our blood culturinginstrument. The data base management program was actually constructed by authorKay McConnell with input on microbiology’s needs from herco-authors. McConnell had learned Applesoft Basic at home on her ownApple II Plus. Data entered on each specimen include patient name, lab or specimenaccession number, room, age, billing number, admission date, requisition date, specimen type, organisms cultured, and susceptibility patterns.
From all the individual test records, we generate complete monthlylist of microbiology testing, sorted by patient name, room number (forinfection control purposes), specimen type, and type of organism. Inaddition, we generate overall statistics on identifications andsusceptibilities by specimen type and by organism. The information is stored on a floppy disk, with printouts forbackup.
The Apple II Plus has dual disk drives and 48K RAM, and we usean Epson FX-100 printer. Here are different aspects of the program: * Code files. These three files are directories covering thenumerical codes for organisms (101, E. Coli) and the alphabetical codesfor specimens (BRON, bronchial) and antibiotics (GM, gentamicin).
Whileany number of specimen and organism codes can be specified, theantibiotic file is limited to a maximum of 25 for space reasons, sinceall the drug codes have to be shown in each report on susceptibilities. If we want to modify the antibiotic file, we must wait until thebeginning of a new year. Computer generation of monthly and yearlysusceptibility statistics depends on the codes and their order.
* Data entry. The data entry format on the computer screen (Figure1) corresponds to the sequence in which informatin appears on ourrequisition form. The specimen code is entered alongside SPEC.Antibiotic codes are read vertically.
The first code is just the letterP for penicillin; the second is AK for amikacin; the third, AM forampicillin; and so on. Beneath each of these, on the organism codeline, is an S for susceptible, R for resistant, or a period if the drughas not been tested. When several consecutive drugs are not tested,periods are easier to count than blank spaces.
Up to four organisms per specimen can be entered at one time. Ifany more organisms are cultured from the same specimen, a second recordmust be created. This was one of the points we paused over whendeveloping the program.
How many organisms would each culture yield?Rarely more than four, but frequently more than three, a review of pastreports revealed. At the top left of the data entry screen is an exit option,allowing a return to the main menu. At the bottom of the screen, theuser has a chance to enter the displayed data into the main patientfile, revise the data, or delete it all. Also at the bottom is thelatest total of patient records for the month. Data entry is performed whenever any employee has a spare 10 or 15minutes.
It is the only employee time devoted to organism andsusceptibility statistics now. The Apple and the Epson do the rest. * Search and edit. The main file can be searched by either patientname or specimen accession number.
Data for an individual test recordare displayed on the screen, and the following options are given:continue search, edit data, delete data, or return (to the main menue).Continued search by name calls up all the specimens cultured that monthon the same patient. Continued search by specimen number allows us tosee any added organisms cultured on a specimen, beyond the maximum offour in th first record displayed on the screen. Searches are usually conducted with some editing purpose in mind.For example, when we find out the identity of a patient originallyadmitted as John Doe, we go into the files and change the name.Occasionally, a physician will ask for all the culture data on apatient. Retrieval takes less than a minute.
We also have an Apple program called FID that enables us to copycomputer files and programs for backup. * Calculations and printouts. Among the availabe screen displaysand hard copies are: a CRT listing and printout of the names andcomputer-record numbers of all patients in a monthly file, for use as anindex; a CRT listing of statistics on a single organism, needed when aphysician asks for drug susceptibilities; and printouts of all data inthe main file, sorted by patient name, by floor, by organism, or byspecimen. These last printouts serve as hardcopy backup to the disk. In additin, we have monthly and annual printouts of statistics byorganism and by specimen. The former shows the number of times eachorganism was cultured and the percentage of organisms tested that weresusceptible to each drug (Figure II).
The data are used bymicrobiologists, clinicians, and the infection control committee tocompare susceptibilities among new and old drugs and to monitorincreasing antibiotic resistance by organisms. The printout of specimen statistics lists each specimen type, howmany were cultured during the month or year, and the percentage of thosecultured that grew any of the coded organisms. Infection control isparticularly interested in this information.
There are drawbacks to our data base management software. For onething, a wide carriage printer is needed for some of the printouts. Foranother, the program is slow compared with commercial software, as wehave mentioned. With 338 entries in the main patient file, 119 entriesin the organism file, 22 entries in the antibiotic file, and 21 entriesin the specimen file, it would take: 29 seconds to search by namethrough the entire patient file; 2 minutes, 40 seconds to load from thedisk to the computer and sort by patient, and 14 minutes to print out; 2minutes, 17 seconds to load and sort by floor, and 13 minutes to printout; 60 minutes to sort and print out by organism; 10 minutes to sortand print out by specimen; 66 minutes to calculate and print out monthlystatistics by organism; and 38 minutes to calculate and print outmonthly statistics by specimen. Now for the advantages.
Since we already had the Apple II Plus andsince the program was written by a laboratory employee on her own time,the only supply costs we have run up are for 14 floppy disks per yearand printer paper. Even though the software is relatively slow, it is still muchfaster than our previous system of manual sorting and calculations. Wehave cut 50 per cent off the time technologists and volunteer hospitalworkers used to spend in compiling our statistics. Since all data arenow kept on floppy disks and a small collection of printouts, we havealso reduced the space needed to store our data. We can continually monitor microbiology data and getup-to-the-minute interim reports on our overall activity.
With programming knowledge and ability to handle random accessfiles in Applesoft Basic, the program can constantly be updated andimproved. Modifications can be accomplished within days–sometimeswithin minutes–therefore providing us with a program tailored to ourneeds.