A Computerized Poultry Serology System

The routine administration of a serology laboratory serving the poultry industry is a highly complex activity. Records of sample receipt must be kept as well as the required assays scheduled in an efficient and cost-effective manner. Results must be recorded and accurate reports despatched to the correct person or persons in a timely manner. This must all continue against the background of peaks and troughs of sample arrival and changing patterns of assays required. The availability and affordability of powerful micro-computers led us to investigate their use in facilitating the administration of the serology work.

The first option considered was to use a general-purpose data-base programme. Some of the types of calculation were found to be too complex for some data-base programmes, other programmes were felt to be rather complex for the laboratory staff who were, at the beginning of the project, inexperienced in the use of computers. A purpose-designed avian serology system was then considered. However this system only included Elisa assays from one supplier, and did not allow the incorporation of agglutination and HI assays. It also relied on a flock-coding system which could be quite powerful but which relied on the lab staff knowing and correctly entering the codes (with no built-in checks on the accuracy of this coding). The number of farms serviced by the laboratory (c. 300) made this difficult. Having discarded these "easy" options it was decided to embark on the design and programming of a purpose-built system to satisfy our objectives. Because of the author's previous programming experience it was decided to write the software using the QuickBASIC system (Microsoft).

Initially it was intended to simply automate the existing procedures for accessions (groups of sera received) and recording of results. In time the system grew to include:

1. Recording of each accession as it comes in, with the basic information required to monitor the serological work (identity and number of samples, tests,storage requirement).

2. Ability to generate lists by farm, by test required, or of all sera.

3. Recording results at the bench in the most efficient and convenient manner appropriate to the usual tests. For rapid serum agglutination tests and haemagglutination-inhibition tests this involves manual data entry, while for Elisa tests the results can be automatically read by direct cable connection to Dynatech plate readers (MR4000, MR5000, or MR7000). Results from non-standard tests can be recorded manually directly in the report using a standard text processor (MS-DOS Editor). Results for the main tests are, in addition, recorded in a data base file (all results recorded during one month in one file).

4. Preparation, editing and printing of interim and final reports without further data entry but with the facility of adding and editing comments.

All assay results are reported in a standard format. Sera within each group (pen, house, source flock etc.) are summarized as shown in Figure 2.8. Individual sera are placed in up to 12 groupings of increasingly intense reaction. For HI tests these will be Log2 of titres, for some Elisa tests they will be titre groups derived from a formula applied to the S/P ratio, for others it will simply be bands of S/P ratio. In all Elisa tests in which there is a recognized "suspect" range any such sera are placed in group 1. For all quantitative assays a mean titre group and standard deviation is presented for each group.

Figure 2.8. A sample serology report format

Presenting the data in this standard fashion facilitates comparisons between different assays and groups. The result is almost as graphical as histograms but allows much more data to be presented on a single sheet report. A bonus resulting from the implementation of this system is that all data generated become readily available for analysis with little extra effort. It is ,for instance, relatively easy to periodically produce summary reports by company, type of bird and assay. To extract and the summarize the serological data presented in the latter part of this dissertation would have been next to impossible without this facility. The ability to summarize the data allows both the veterinarian and poultry company management to have a much better overall picture of serological response.\par \par The user documentation or manual for the system software described here is included as an appendix to this dissertation (Appendix C.)