Empower IVD Globe Meeting Report
On December 10th 2014 the first Empower IVD ● Globe user meeting was held in Ghent. The aim of this meeting was to report on the status of the Empower project, which promotes the quality, stability, and comparability of global In Vitro Diagnostic (IVD) testing. More than a year ago, we were excited to help spread the word about this project. It's even more gratifying to see the results of the pilot stage.
Empower IVD ● Globe – a status update (meeting report)
Kenneth Goossens on behalf of the Empower team.
On December 10th 2014 the first Empower IVD ● Globe user meeting was held in Ghent. The aim of this meeting was to report on the status of the Empower project, which promotes the quality, stability, and comparability of global In Vitro Diagnostic (IVD) testing. As mostly users of the project had registered to attend the meeting, it was the intention to have a lively discussion: to learn what they think of the project, and how they use it; to know what they appreciate in the project, what is missing or what they would like to see changed. Communication with the involved parties is namely a key aim of this project.
In general the attendees confirmed the benefits of the Empower project for the laboratory, the manufacturer, and finally, the patient. If the project will be successful, it will provide global evidence about field performance of IVD tests. This will allow laboratories to compare with their peers and set benchmarks for their own performance. Manufacturers can address issues of test performance and aid laboratories in creating optimal protocols for best test stability and peer group comparability. Finally, the project can aid in global quality advancement and harmonization of IVD-testing. Additional information on the topic can be found online (www.stt-consulting.com) and in a previous invitation for participation on the guest assay section of the Westgard site (http://www.westgard.com/patient-percentile-project.htm).
The meeting was attended by approximately 80 participants. Most of them were clinical chemists (and coworkers), but also representatives of IVD manufacturers and Laboratory Information System (LIS) providers were present. Most participants were from Belgium, but there were also some attendees from France, Germany, Ireland, Turkey, and the UK. A discussion was held on four different topics: i) the Master Comparison results of the 2014 survey and the requisites to organize a new survey; ii) the relevance of Patient Percentile Monitoring/Percentiler for medium sized labs; iii) the relevance of mild transaminase elevations and the corresponding need for correct monitoring; and iv) Patient Percentile Monitoring: present and future. For presentation see: http://www.stt-consulting.com/news.php?rubriek=8.
Results of the 2014 Master Comparison (MC) survey – an EQA study that makes use of single-donation native sera – were summarized. In this study 8 different analytes were studied (alkaline phosphatase, ALP; alanine aminotransferase, ALT; aspartate aminotransferase, AST; lactate dehydrogenase, LDH; gamma-glutamyl transferase, GGT; potassium; sodium; and chloride) in 126 labs representing 8 different peer groups of equal size, and 5 manufactures. The results indicate the need for i) phasing out the non-IFCC (International Federation for Clinical Chemistry) assay variants; ii) improvement of the quality, comparability, and standardization of the ALT assays; iii) better comparability of the GGT & LDH assays; iv) improvement of peer comparability at low concentration levels; v) single-manufacturer efforts for improvement of comparability for nearly all enzymes examined. Peer performance and assay comparability were generally good for all electrolytes. An exception was the negative bias of the Roche assays for chloride.
Although results of this survey prove to be highly valuable, the costs for conducting the MC surveys are quite high for a merely descriptive exercise. Therefore, the next survey might be postponed in order to give the respective manufacturers sufficient time to address issues that were observed in the past 2012 and 2014 surveys. However, the attendees deemed the MCs in general of great value and clearly expressed the wish to continue, even though the financial burden for its organization is high. They considered their own measurement costs reasonable, even if it would be the intention to tackle immunochemistry assays in future surveys. Therefore, the organizers will look for funding, possibly by industry or EQA partners.
Then the Percentiler was presented, which is an on-line tool for patient percentile monitoring. It consists of daily monitoring of medians of measurement results for twenty commonly measured analytes in outpatients. A demo version can be found on: www.thepercentiler.be (login: DEMOLAB; password: demo1234). First, a guest presenter, clinical chemist Timothy Ghys from the Hospital AZ St Lucas Gent talked about the relevance of patient percentile monitoring for medium sized labs. His hospital was shortly presented in terms of bed-size, laboratory services, and quality management efforts to situate the use of the Percentiler in a mid-sized hospital laboratory. He confirmed that the Percentiler was valuable to monitor assay variability. It is an added value in the overall quality control strategy, in particular, for mid- to long-term quality management. Nevertheless, data analysis needs to consider population effects and test volumes ("size matters"). In addition, there is a need for solid peer groups.
Next, the value of the Empower project was considered in an ALT case study. Literature indicates the relevance of mild elevations of transaminases (ALT and AST) in nonalcoholic fatty liver disease (metabolic syndrome), which becomes more and more prevalent as obesity rates increase. A case study (female developing metabolic syndrome) was presented with the following ALT values (U/L): 17, at the start of yearly check-up; 44, diagnosed as development of metabolic syndrome; 28, after six months of treatment; 18, after long-term treatment. This case study showed that, for correct monitoring of disease development and treatment, stable and comparable assays are needed. However, in the Percentiler application, significant lot-to-lot changes are sometimes observed (for example: ALT 35 to 22 U/L, in several steps), which can jeopardize patient management. The significant between-manufacturer differences observed in the 2014 MC survey can similarly be confounding if a patient moves.
Finally, the current status of the Percentiler (i.e., participants, peer groups, IT-connectivity, and user interface) and the theoretical concept of the project were presented. Currently 125 laboratories with approximately 250 instruments take part in the survey. Results were shown indicating that the quality specifications selected by the project organizers can be achieved. These specifications are based on biological variation, but take into account what is currently achievable (state-of-the-art). In addition, the existence of the Flagger was emphasized. This online application has the potential to translate analytical quality into quality of medical decision making using locally important cut-offs, but it requires that the IT-connectivity enables participants to send their daily hypo- and hyper-flagging rate automatically. The presentation was concluded with a summary of the project benefits, such as real-time monitoring of (in)stability, strengthening the communication between the laboratory and manufacturer or clinician, establishing realistic quality specification goals, and providing a basis for cross-manufacturer comparison.
Attendees expressed the wish for additional information on peer group sizes and potential benefits of making the distinction between instrument types and method principles. Although the 1st request (providing updates on peer group size) can be met, the 2nd one proves more difficult as the survey is too small to split up the analytes according to measurement principle or the peer groups according to instrument types. Attendees wondered about the possibility to expand with other analytes (e.g., from hematology). Although this is not excluded in the future, current time management and available work forces are prohibitive. However, in the context of the IFCC project for standardization of thyroid function tests chaired by Prof. Thienpont, FT4 and TSH will be added in the near future, but only for medium to large sized laboratories on invitation. Although some peer groups are not yet available, it was stressed that some analytes don't require a peer target, as manufacturers use the same reference intervals. This is, for example, the case for sodium, calcium, potassium, magnesium, and inorganic phosphate.
In conclusion, participants deemed both the Percentiler and MC surveys useful tools for their quality management and incentives for manufacturers to improve assay comparability and stability. The current intention is to keep the Percentiler free of costs as long as possible. The MC surveys will continue if sufficient funding can be found.