PAMM is small but mighty. Short for the Laboratory for Pathology and Medical Microbiology in Dutch and located in Veldhoven in the Netherlands, the institution performs diagnostics on cassette samples for five hospitals and 500 general practitioners in the country. "In the world, that isn't so significant," says PAMM's head of bacteriology, Theo Liebregts. "In Holland, though, it's quite big."
With so many hospitals and doctor's surgeries across the southern part of the Netherlands relying on PAMM to process samples, the laboratory has made a conscious effort over the past decade to invest in lab automation to boost the accuracy of its reports and increase overall sample throughput. In this regard, its partnership with Copan Italia and MLS, the distributors in Benelux, has been crucial, with the company supplying PAMM with its WASPLab automated sample-processing solution, supported by the PhenoMATRIX system module. The reason why Liebregts chose the company over its rivals in the marketplace is clear.
"At the time, its competitors could all deliver very big systems that were very, very expensive," he explains. "Meanwhile Copan could deliver, as it still does, a solution tailor-made for your lab size and procedures. It understood what our needs were a lot more than its competitors did." The positive fruitfull and continuing collaboration between Copan and PAMM has served to improve Copan's solutions, and has helped PAMM to remain efficient and competitive in an increasingly challenging market.
PAMM installed WASPLab in 2014, says Liebregts. "We needed a way of increasing the speed of diagnostics with the same number of people, but we didn't want to do so at the expense of accuracy," he says. "That made us decide on an automated solution."
WASPLab works by producing high definition pictures of the plates before and during the incubation period, delivering them to lab technicians without the need to physically handle them for examination. This serves to drastically reduce the capacity for human error during the interpretation phase, and increases sample throughput. When PAMM's technicians began working with the WASPLab, the lab was processing around 3,000 samples per week. Two years later, that throughput has increased to 5,000. Its staff numbers, meanwhile, have only grown by 10%. "We could only have achieved this gain in capacity by embracing automation," says Liebregts.
The addition of the PhenoMATRIX algorithm, meanwhile, allows for more sophisticated image analysis by evaluating the growth present on an individual sample plate. Rules on the sorting of negative and positive plates can be tweaked by the user. This reduces the time it takes for lab technicians to read individual samples, freeing them for other duties within the lab and enables them to apply a constant quality refinement with deep learning techniques.
The impact of these technologies on the patient is obvious to PAMM's head of bacteriology - as automation increases sample throughput and allows staffing costs to be redistributed, diagnostics will become even cheaper and faster. Liebregts believes that his lab can help realise this ambition, at least in the area it covers, very soon indeed.
"Six months ago, we were using the PhenoMATRIX module to automate image analysis of chromogenic agars," explains Liebregts. "I believe that within one year, the software should be able to handle the reading of more than 50% of our plates. That's a target that I really want to set, and I believe we will be able to reach that."