AMS for quality control – a user's story
Xchange newsletter – February 2009
In addition to improving efficiency, the right middleware can offer the laboratory extra functionality and also help maximise performance and quality. Here Susan Shanley, technical manager in Biochemistry at Glasgow Royal Infirmary shares her experience of using the Analyser Management System (AMS)* for quality control.
All change
As part of the contract signed between NHS Greater Glasgow and Clyde (as it’s now known) and Abbott Diagnostics in 2005, AMS was installed in four laboratories within the North Glasgow University Hospitals NHS Trust. Our previous biochemistry middleware had been in use for more than 10 years and there was some concern that the simultaneous change in both automation and software would be extremely disruptive. However, we were pleased to find that the transition was relatively painless as excellent analyser training was given by Abbott, and the AMS software proved to be extremely user friendly and flexible.
Establishing a QC program
It was particularly important for us to have an integrated QC program, as our hospital LIMS does not support raw QC data. At the time, our only method of collating data was copying the raw QC data from the analyser and manipulating it manually to get the figures we needed. This was a very onerous task taking up to three days per month.
AMS core functionality includes a QC program that can be tailored to meet the lab’s own needs. Abbott and Omnilab were both extremely responsive to our requests. Through effective communication and on-site presence, a suitable program was developed. It is extremely flexible.
With the program appropriately configured, QC results from all analysers connected to the AMS are fed into a database. From here, we can extract a variety of data, both statistical and graphical, quickly and easily.
Daily QC
Current QC data (for any connected analyser) is instantly accessible from the main AMS menu. I can easily view the QC results for a whole worklist (whether it’s chemistry, immunoassay or haematology) or I can filter the results down to individual analysers. QC’s with failed Westgard rules are highlighted, together with details of the analyte, for easy identification (figure 1).
Figure 1
I can select a specific analyte and view its Levy Jennings graph including the assigned Means, 1, 2 and 3 SDs (figure 2). Derived data (such as the calculated mean, SD and CV) are also displayed. Consequently, actual data is shown alongside expected data.
Figure 2
Also from this screen (figure 2), the ‘compare levels’ option allows me to choose a QC level and compare the results from different analysers or, alternatively, I can view different levels of QC from the same analyser. This helps me establish the source of any problems that may arise.
The ‘list of measurements’ function is another particularly useful feature. Simply double-clicking on a graph replaces it with the actual data. This too helps identify the source of any errors. If a mistake has been made (switching two QC levels for example), these results can be deleted. For accreditation purposes, this data is stored in a ‘deleted’ file and so complies with the audit trail required for CPA.
Generating QC reports
The ‘report manager’ function enables various reports to be produced. The ‘six month report’ is a particularly important one for me as I use it to report back to the North Glasgow quality control sub-committee. It can be used to show everything from broad overviews to very detailed examples.
Multi-site networks
In North Glasgow, the AMS system is on a single server so all the processes described can be performed for more than one site. This is particularly helpful when changes in lot numbers or QC material are made.
Conclusion
Overall, the AMS and QC package makes my life much easier. Not only does it save me time in collating data and generating reports, but it also helps to quickly identify the source of any problems for rapid resolution.
* provided by Abbott Diagnostics in partnership with Omnilab srl