key: cord-0823761-83f0ruku authors: Wagner, Joseph E.; Warriner, William J.; Bradfield, Sylvia A.; Farrar, Patricia L.; Morehouse, Lawrence G. title: A computer based system for collection, storage, retrieval and reporting accession information in a veterinary medical diagnostic laboratory date: 1984-12-31 journal: Computers in Biology and Medicine DOI: 10.1016/0010-4825(84)90033-7 sha: 3b019c97245182a2f2a638b91e7c66c269096a3a doc_id: 823761 cord_uid: 83f0ruku Abstract Substantial data collected from large numbers of accessions, the need for comprehensive reporting of negative as well as positive laboratory findings, and the necessity for obtaining rapid diagnostic correlations prompted the development of a computer based system of accession data management for collection, storage, rapid retrieval, reporting, concording, and administrative compiling in a state-university Veterinary Medical Diagnostic Laboratory. Increasing numbers of accessions and a desire to store and rapidly retrieve information on each accession have prompted veterinary hospitals, clinics and diagnostic laboratories to develop computer-based data management systems [ 11. One such system has been developed at the University of Missouri-Columbia Research Animal Diagnostic and Investigative Laboratory (RADIL) section of the Veterinary Medical Diagnostic Laboratory at the University of Missouri College of Veterinary Medicine. Annually, a large amount of accession data is generated at the RADIL. In 1982, for example, there were 1293 accessions with a total of 14,803 animals examined at the RADIL. The development of a computer based system of data management has made the storage, compilation, and rapid retrieval of large amounts of diagnostic information possible. This system is capable of handling large volumes of diagnostic data such as results of histopathology, serology, toxicology, virology, parasitology, necropsy, and microbiology examinations as well as demographic-zoographic/patient data. This system uses the full screen capabilities of IBM 3278 series computer terminals to display a blank panel (essentially a blank form for recording laboratory results) which is filled in by the laboratory technicians at the major data generating stations throughout the laboratory. Preliminary, final, and supplemental reports or select diagnostic data for each accession are printed by the computer directly from this information, and copies of these reports are sent to the person(s) responsible for submitting the accession. Individual accession records are kept in a VSAM database (IBM program product) and archived to magnetic tape every quarter. Information from these records is abstracted by the computer for an annual concordance catalogue index and other administrative reports. During the development of our data storage and retrieval system, several design objectives were conceived and followed. These included: 1. The system should be easy to use, and require little training; 2. The system should allow for easy production of summaries and year-end reports, preferably using an existing file processing package such as Mark IV (Informatics, Inc.); 3. The system should allow for easy update of accession records; 4. The system should interface with the existing concordance index program [2]; 5. There should be little or no keypunching or other data entry required to produce an 325 annual or other compiled concordance indexes. Data for compiled reports should be acquired passively from existing files originally created to store laboratory data for reporting purposes ; 6. Preliminary, supplemental and final reports should be produced in English (as opposed to numerical codes), and in letter format suitable for mailing to commercial research animal producers, investigators and/or owners, and referring veterinarians. DATA ENTRY The data entry system described herein requires a large capacity computer. This system runs under the time sharing option of the University of Missouri's Amdahl470 V/8 running IBM MVS/SP release 3 and JES2/NJE release 3.1 operating systems. Data entry and editing is by means of seven IBM 3278 model 2 full screen computer terminals and one full screen 3276 controller-terminal. These are located in major data-generating areas of the RADIL, i.e. the accessioning area, the necropsy, microbiology, and serology laboratories. The full screen capabilities of these terminals are used for accessioning and data entry. The computer displays a blank form (panel) on which the CRT terminal operator enters appropriate data. To avoid confusion, the panel is displayed in low intensity characters while operator entries are displayed in high intensity characters. During subsequent screen display of data and for updating accessions, the CRT displays the panel with such information filled in as is currently present in the accession record. The operator may then change or delete this information by typing "over" it, or may add more information to it, or both. Each panel has its own designated "free text" area for continuation of selected fields as well as general comments. A set of 13 panels are in use, with provisions for adding more panels when needed. Typically, each panel, except the demographic-zoographic and summary-concordance panels, comprises a report of data from one section of the laboratory. Each panel type is shown in this paper with data entered in italics from a fictional accession (No. 12345). The actual report generated and printed from entered data is also presented for each panel. Because all laboratory findings are treated as confidential information, data entries on these panels do not represent actual accession material, but are merely for demonstration purposes. Demographic-zoographic panel ( Fig. 1) When an accession is presented to the RADIL section of the Veterinary Medical 12 13 14 15 16 17 18 19 20 21 22 23 24 ,ll,l111lll~llLllll1111111111111111~1111~~~~~~~~~~~~'~~~~~~"'~~"'~'~'~~~~~~~~~ Diagnostic Laboratory, demographic and zoographic information is immediately entered by a data controller or data entry operator from information on a form submitted with the accession. This panel ( Fig. 1 ) includes such information as the investigator's and/or owner's name and address, the referring veterinarian's name and address, type of specimen submitted (whole animal, slides, fixed tissue for histopathology, swabs for culture, etc.), number of specimens submitted, species, strain, age, sex and name of animal (if applicable), and accession history. Owner or investigator, and referrer names may be entered as a 3-digit code number which the computer "looks up" in a directory, replacing the number with the appropriate name and address on all subsequent CRT displays or printed reports. This insures consistency and accuracy in names and addresses. The demographic-zoographic panel ( Fig. 1 ) and summary-concordance panel ( Fig. 13 ) always occur exactly once per accession, and together with certain control information, form the "base segment" of the accession. In addition to the base segment, for each accession, there are several types of "subordinate segments" which may occur independently of each other any number of times, or not at all. Each subordinate segment is represented by a panel type, as described herein. Necropsy panel (Fig. 2) The necropsy results panel (Fig. 2) includes space for recording results ofprenecropsy and necropsy examinations for one or more animals. Figure 2 includes data entries for 20 animals ; 10 adult females and 10 juvenile males (arrow, line 5). Reports of negative findings and normal necropsy observations, as well as reports of the kinds of techniques used (such as the kind of blood collection method used, arrow, Fig. 2 , line 8) can be entered by a code number, thus reducing data entry time. Through use of a directory, the code number appears as a statement in English when printed on the report. If the data entry operator needs more space than is available on the panel to enter or report a finding, a free text reference can be entered in the appropriate field, eg. the "I A 1" at the arrow on line 9 in Fig. 2 . This allows the operator to continue the report of a finding in a free text area (eg. "I A 1 EYES."in Fig. 2 (a) on line 3). The computer will, on the final printed report, replace the free text reference with the text from the free text area (arrow on Fig. 2(b) ). Organs and tissues can be designated as normal (N) and/or "flagged' as having been sent to either the histopathology (H) or microbiology (M) laboratory subsections, or both (X) (arrows in Fig. 2 on lines 14-19) . In Fig. 2 , line 8, a code number was entered to indicate the method of blood collection used ; "1" (arrow) when taken from the directory and printed on the report reads, "JUGULAR INCISION". An entry of "2" would read, "ORBITAL BLEEDING", and "3" would read, "AXILLARY INCISION". Code numbers are also used for several other entries, such as general appearance (line lo), skeletal palpation (line lo), andexternal lesions (line 11). Through use of a directory these code numbers are replaced by "canned" statements on the final report ( Fig. 2(b) ). If these animals were without significant gross lesions, a code number would have been entered in the field "NGL IN ANY SYSTEM" (arrow in Fig. 2 on line 13 ). Depending on which code number was used, the final report would contain a statement telling which organs and tissues had been examined and found without lesions. There are five different panel formats for entering microbiology results. The first is designed to show detailed microbiological culture results from individual organs or sites (Fig. 3) . Shown in Fig. 3 and 3 (a) are typical culture results for a group of animals in which nasopharynx and middle ear were cultured on mycoplasma agar and on blood agar and incubated in a 10% carbon dioxide environment. The second format for entering microbiology results shows the results of antimicrobial sensitivity testing (Fig. 4) . In this accession, an antibiotic sensitivity test was performed on a Pseudomonas aeruginosa bacterial isolant from the middle ear of animal A, and its relative sensitivity to 11 antibiotics was determined (Figs 4 and 4(a) ). The third format reports microbiology results in a tabular form (Fig. 5) . The bacterial isolants commonly cultured from laboratory rodents are listed on the panel and there is room for adding two additional isolants on the same table. In the event that culture methods employed would not detect certain microorganisms, an "x" placed in the space immediately preceding the genus and species of the bacteria causes elimination of that line (organism) from the final report (Fig. 5, lines 7 and 13 ). Thus Citrobacter freundii and Pasteurella pneumotropica do not appear on the final report ( Fig. 5(a) ). Negative results are tilled in automatically by the computer, and are overtyped if positive results are found. 11111,111,,,,11111,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,(,,,,,,,,,,,,,,,) PNEUNOTROPI CA The fourth microbiology panel type (Fig. 6) reports group results in a tabular format, but the genus and species of the bacteria must be filled in by the data entry operator (Figs 6,6(a) and 6(b)). A specialized form of microbiological examination panel is provided for reporting positive or negative results of tests on large numbers of fecal specimens cultured for Salmonella spp. and Pseudomonas spp. (Fig. 7) . This panel is designed to facilitate tracking of culture results by supplier and colony (building and room). The results shown in Figs 7, 7(a) and 7(b) indicate that the first five samples were negative for both Pseudomonas and Salmonella, the sixth was positive for Pseudomonas and negative for Salmonella, samples 7-15 were ,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,',,,,,,,,,,,,,,,,,, Fig. 7(b) . Microbiology panel type 5, final report. negative for both, sample 16 was positive for both, and samples Pseudomonas but negative for Salmonella. Parasitology panel (Fig. 8) 17-20 were positive for The parasitology panel (Fig. 8) is for recording results of parasitological examinations performed. Both "EXAM METHOD" and "SPECIMEN EXAMINED" may be entered as rt al. I,,,,,,,,,,,,,,,,I,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, ,,,II111111~1111I(IIII.IIIIIII:I~IIIIII111(1"""""'1"'1"""""'1""~~~' '11,,,,,,111,,,,11,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,, either a code number or in English (Fig. 8) . On the printed report the code numbers are translated into "canned" responses ( Fig. S(a) ). For example, a "SPECIMEN EXAMINED" entry of "1" (arrow in Fig. 8 on line 6) refers to the perianal area ( Fig. 8(a) ), and an "EXAM METHOD" entry of "6" (arrow in Fig. 8 on line 7) indicates microscopic examination by cellophane tape impressions ( Fig. 8(a) ). ;-8. WAGNER lllll, , , lllllll, llll, , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , llll~l~ Fig. 11 . Histopathology panel, type 1, completed 1,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,,_,,,,,, lllllll,l,l,,,l~ Serology and ELZSA panels (Figs 9 and 10) The serology panel (Fig. 9 ) includes complete serology results only for those tests commonly performed, i.e. hemagglutination inhibition (HI) and complement fixation (CF) tests. In this case, a serological examination of aminals A-J was performed, using complement fixation tests and hemagglutination inhibition tests (Figs 9 and 9(a) ). The ELISA (enzyme linked immunosorbent assay) panel is used for reporting results of mouse hepatitis virus (MHV), rat coronavirus (RCV), Sendai virus, or Mycoplasma testing for serum antibodies (Fig. 10 ). An ELISA to test for Mycoplasma pulmonis antibody was performed on the sera of animals A-J (Figs 10 and 10(a) ). (Figs 11 and 12) There are two formats for entry of histopathology results. The first of these results in a tabular report (Fig. 11) is entered in complete sentences or narrative form on this panel, and printed out exactly as entered. This format is useful for lengthy descriptions of lesions. Summary-concordance panel (Fig. 13) The summary of findings panel (Figs 13, 13 (a) and 13(b)) includes a brief summary of laboratory results from other panels, along with a summary diagnosis and diagnostic information for concordance indexing. In the upper right hand corner (Fig. 13) is the "ARCHIVE?" field. As long as this field is left blank all panels with this accession number remain on line in the master file. This field is left blank until the accession is completely processed and all reports have been mailed. When the accession has been fully reported and all reports are mailed, this field is marked with an "X". Quarterly, accessions marked with an "X" in the "ARCHIVE?" field are removed from the master file and placed in a magnetic tape file. About 25 technical and professional personnel enter data into the RADIL system described herein. After data has been captured through entry on the various panels, individual accession reports are generated interactively using a LA1 20 Decwriter II computer terminal located in the laboratory but connected to the host computer via low-noise phone lines and 1200 baud modems. One to three or four part carbonless paper is used in this terminal to provide printed copies of reports ; one for the referring veterinarian, one for the owner or investigator, and one for the RADIL files. Reports can be designated either preliminary, final, or supplemental, and may optionally include only designated panel types. Alternately, accessions can be printed centrally on campus on an IBM 3800 printing subsystem for twice daily delivery to our laboratory and subsequent mailing or distribution. Regardless of where accessions are printed, all mailing or distribution of reports is done by a data controller who controls the flow of accession material and subsequently generated data. At any time this key individual knows of the status of any accession that has entered the laboratory. Many telephone inquiries can be answered by the on-line Data Controller. Accessions are easily "tracked" through the laboratory. A special "Culprits" program flags any accession that has been in the laboratory over two weeks. The advantages of the system described herein are many. Clean typed reports are issued. It is convenient and easy to provide complete reporting of diagnostic tests performed, i.e. positive as well as negative findings are reported. Multiple copies are generated without the need for preprinted forms. Errors are corrected electronically by data entry operators. Computer stored and accessed directories of complete mailing addresses of referring veterinarians and selected clients are entered by three digit code, thus speeding up data entry and improving accuracy and completeness of addresses. Use of window envelopes eliminates the need for addressing envelopes. There is no typing of reports per se, rather, data is entered from the laboratory by laboratory technicians or a data entry operator. Accession history and demographic-zoographic data is acquired for concordance indexing without the need for reentry. All data is held in a form accessible through Mark IV and specialized programs can be prepared for summarized reporting using an optional "display" mode. Additional panels can be created if needed, thus, allowing for future expansion. Availability of comprehensive user's manuals and interdigitated Standard Operating Procedures of laboratory procedures greatly facilitate training of new employees. There are also several disadvantages to this system. It requires a high capacity computer running under IBM's MVS operating system, and IBM's 3278 series terminals. These are expensive, necessitating sharing with other users. This can produce prolonged response times during periods of maximal usage. Development of such a system as this requires the services of a computer programmer as well as considerable time. Data processing and storage costs are quite high. Nearly all laboratory technicians must be trained to enter data into the system. Consistency in terminology is necessary. Users' manuals must be prepared and updated periodically. The programs are extensively interdigitated, making certain types of program changes difficult. High case loads and the necessity for obtaining rapid diagnostic correlations prompted the development of an electronic computer based system of accession data management, storage, and retrieval in a large state-university Veterinary Medical Diagnostic Laboratory. This system is capable of handling large volumes of diagnostic data such as results of histopathology, parasitology, necropsy, and microbiology as well as demographic data. This system uses the full screen capabilities of IBM 3278 model 2 computer terminals to display a blank panel (essentially a blank laboratory results form) which is filled in by laboratory technicians or data entry operators at the major data generating stations throughout the laboratory. Final reports are printed directly from this computer stored information. Individual accession records are kept in a VSAM data-base and archived to magnetic tape every quarter. Information from these records is abstracted as needed by the computer for an annual concordance index and other administrative reports. The management ofclinical and laboratory data in veterinary medicine A computerized system for retrieval of case information in a veterinary diagnostic laboratory organ or tissue examined microscopically in routine disease surveillance accessions (Fig. 11, line 4) . Each cell in the table may be left blank (indicating that an organ or tissue was not examined for that animal), filled with a dash or minus sign (indicating no significant microscopic lesions were found in that organ or tissue for that animal), or a description of the lesions found can be entered using a two digit number from a directory (Fig. 11) . To speed data entry and insure consistency, a directory of descriptions of common microscopic lesions was programmed which the computer interprets to narrative statements in English on the final report. The operator thus need only enter the code for a given lesion, for example, when one enters "54", the computer will translate it and " MULTIPLE SYNCYTIA IN ABSORBING EPITHELIUM COVERING VILLI PATHOGNOMONIC OF MOUSE HEPATITIS VIRUS INFECTION." will appear on the printed report (arrow on Fig. 1 l(b) ). Descriptions of lesions not on the list can be entered by means of a free text reference (arrow in Fig. 11 on line 14) in the appropriate cell and by entering the appropriate text in the free text area for that panel, labeled with the same free text reference ( Fig. 11 (a) I""""""" in the College of Veterinary Medicine. His research activities have centered around infectious diseases of animals and poultry (where he has senior or co-authored more than 70 papers in refereed journals), and in toxic effects of fungal poisons. where he has authored several papers and is co-editor ofa threevolume series on mycotoxins and mycotoxicoses. He has had interest throughout his career in computerization of data from animal disease diagnostic laboratories.