Abstract
How a Japanese pioneer of laboratory automation re-organized his workflow
This report is based on an interview at the Seirei Mikatabara General Hospital in Hamamatsu with Shokichi Furuhashi, director of the Medical Laboratories and Isao Yoshida, chief manager of clinical chemistry conducted by Prof. Georg Hoffmann from the University of Munich, Germany
The Seirei Mikatabara General Hospital in Hamamatsu was second after the University of Kochi, Japan, to install a self-made conveyor belt for workflow automation. 10 years later, in 1996, they moved into a new building, and quite surprisingly they decided to install a stand-alone “automation solution. In an interview, conducted in spring 1998, we conducted an interview to find out, why the conveyor belt pioneers changed their minds:
Furuhashi: When we moved to the new laboratory, we had to rethink the necessity of each and every piece of equipment primarily for two reasons, one being space constraints and the other being financial restrictions. In my eyes, a conveyor is not necessarily an essential part of an automated laboratory.
In addition, our old conveyor belt, which was built by our own engineers, was not fast enough to meet the goals of modern emergency testing.
As a last issue, I would like to mention, that given the configuration of our new laboratory, the conveyor would have interfered with the movement of the tecnicians, since the new laboratory is L-shaped. With its curves, the conveyor would have been in their way.
Yoshida: In those times, the blood was drawn on the first floor and the laboratory was on the second floor. To reduce labor and to speed up the the process, the conveyor transport system was connected with the large size multi-channel analyzers on the second floor. We installed two conveyor belt lines, one for routine analysis samples and another for STAT analysis samles.
Yoshida: The number of emergency samples has increased from about 60 twelve years ago to more than 400 as of today. This is due to an increase of the total volume of 300 to 700 per day as well as a tremendous shift towards emergency testing, which represents now 60% as compared to 20% in those days. It seemed to be faster and even more labor saving to carry these samples manually to the second floor. So the conveyor became less and less useful.
In addition, whenever the conveyor had troubles, we had to shut it down for about two hours to get the problem fixed, and this had a very negative impact on the laboratory's performance.
Furuhashi: Our mission is, to deliver all laboratory results within less than 30 minutes after the samples have arrived at the laboratory. In order to maintain such a short turnaround time with a conveyor belt solution, you need a low number of high throughput analyzers, otherwise, you will create traffic jams on the conveyor all the time. Such analyzers are very expensive. In addition, conveyors are less flexible in distributing specimens among various analyzers of the same type than humans.
Our new philosphy is, to have a higher number of smaller analytical instruments in the laboratory and to maintain flexibility by putting more emphasis on people rather than machines. We now use three Olympus AU 600 for STAT and routine analysis and an OLA 6000, which splits the primary samples into aliquots. The labelled secondary tubes are then distributed manually to as many of those analyzers as are actually needed. In general, we keep all three analyzers in operation, except in those instances, where maintenance is performed.
Hoffmann: How does your current workflow work practically?
Yoshida: We still use a conveyor belt, but only for transportation purposes from the blood drawing site to the laboratories. There, all specimens are centrifuged in the conventional way and loaded onto the OLA 6000. We use Olympus 10-whole racks for this purpose. The machine reads the barcodes, so that it knows which tests need to be done and how many secondary tubes are needed. It prepares the appropriate number of secondary tubes, labels them and dispenses the volume, which is needed for analysis.
Yoshida: In general between 200 and 600 μl. The volume of the send-out samples, which go to SRL (special reference laboratory) is 500 to 1000 μl.
Furuhashi: The secondary tubes get sorted into routine samples and STAT samples in 30 output channels. The operator takes the racks out and carries them to the appropriate analyzers. Besides the three AU 600 this includes also the Olympus ASE 630 electrophoresis machine, the immunochemistry analyzers Lumipulse F, Axsym and ACS 180 plus several off-line channels for manual testing and sendouts. There are also three error channels for those samples that clotted or contain insufficient volume.
Furuhashi: If the workload is high, all analyzers run simultaneously with full capacity. So, we do more analyses than before in a shorter time with less people. Despite the high number of samples, we process today, we need only five people in the chemical and immunochemical work area (two for front-end processing plus blood gas and glucose testing, two for immunoassays and electrophoreses and one for chemical analyses on the three AU 600 instruments).
The Olympus Automation System OLA 6000 can be used as a stand-alone aliquotter or as an integrated module. Our picture shows an installation at the University of Tokyo with a conveyor belt link.
