The automation of laboratory (lab) processes is a long-established trend that has helped clinicians to get the results of tests faster and, therefore, improve patient outcomes. It has also enhanced the efficiency of lab testing, which increases its capacity and can have a noticeable impact on costs by, for example, reducing the length of a patient’s stay in hospital or lowering the need for unnecessary tests.
Equipment manufacturers have been working on lab automation for decades, but their innovation has taken a significant step forward in recent years; however, with each advance comes a new set of questions about what lab automation can achieve and the factors that influence the decision to invest in expensive new systems.
"There are big advantages to lab automation," explains Robert C Jerris, director of microbiology at Children’s Healthcare of Atlanta. "It standardises functions and allows you to maintain consistency. New systems deliver better-quality results than older systems and they improve throughput, which is a major benefit. Turnaround times are better and efficiency increases as the hands-on time for the technologist falls.
"These systems come with a high cost, but the results are better for patient care. A big question concerns what better outcomes we see, but in microbiology, which is my field, I see in the molecular arena a new set of standards worldwide. Companies have strong databases of needs for new lab and patient outcomes, and they are enabling multiplexed tests for things like respiratory specimens and assays for each pathogen. We used to culture viruses, which could take 30 days or more, now it takes just hours."
Huge potential
Lab automation encompasses all of the R&D processes that optimise lab technologies to deliver new and improved systems. It is a field to which many disciplines contribute, whether it be academia, clinicians or commercial research companies and engineers. Their output can inform the design of many types of innovation in devices, instruments, software algorithms and methodologies. It is a diverse and innovative field with
huge potential.
Since the mid-1990s, when the world’s first automated lab-management systems emerged, lab automation came to include new systems to improve single processes, and those that could bring together different processes and manage them more efficiently while integrating them. It is this level that many of the most exciting developments are now happening, though many challenges remain.
"I am a very robust user of lab automation in diagnostics and microbiology, and I’ve been in the business for 30 years, so I’ve seen everything from test tubes to the sophisticated systems we are using now," says Jerris. "I have seen the slow progress in the early days of lab automation, but I have also seen it gather pace in the last five years.
"Things have come a long way, but there are still many challenges, like standardisation in processes for the identification of substrates. We are dealing with biological living agents that don’t read the rules. They modify, so the algorithms to identify them must evolve. Lab automation is slow to adapt to evolving, mutating biological agents; however, there have been advances in antibiotic susceptibility testing, which started out as a manual process."
Joined-up thinking
According to Jerris, there have been many advances in the three separate stages of diagnostics and microbiology on which lab automation has focused, but these stages have not yet been truly linked.
In the preanalytical phase, there has, over the last five years, been progress in automating streakers – devices that draw a bacteria-laden needle across the surface of a solid culture medium. In the analytical phase, the automation process has improved the processes of susceptibility testing and identification. In the post-analytical phase, big steps forward have been made in the processing of data and the ability to apply rules to that data to avoid errors such as the use of drug combinations that are incompatible.
"What is lacking in microbiology is the seamless integration of those three phases," says Jerris. "If that is achieved, then it will be similar to the stage that has been reached in automated immunology and chemistry.
"The challenge now is that we are not ready for complete automated identification of organisms from specimens; for example, they still require plating for epidemiological typing. So, for the next ten to 15 years, we will still need cultures. We need something that – even if it does not achieve complete automation – allows us to observe colonies on the plate and analyse them for susceptibility."
One area where Jerris sees important advances is in the application of video image processing systems. These have the capacity to observe organisms and recommend the next stage of the process to which they should be sent, such as spectroscopy or susceptibility testing; however, in terms of systems that integrate the three phases of the analytical process in the seamless way Jerris would like, some companies are coming close, but there are still gaps to close.
"For microbiology, we are close with two systems – Kiestra and Copan – but they are not seamless," he says. "Complete automation may not be for everyone and I have questions at all three stages, but gradually we are working through the concerns that people have."
One of the advantages of the Kiestra Fully Automated Microbiology Testing System that Jerris refers to is that it focuses heavily on traceability and visibility throughout the entire testing process. The system is powered by 23 individual PCs using customised software, which the company claims has an error rate of less than one in a billion for some devices. The system logs each action throughout the testing process to provide an audit trail for each sample dish.
The sample processing system takes orders from the sample inoculation benches, processes the request for different dish types and labels each dish appropriately. It has over 1,000 test-type variations in its database. After sample inoculation, the dishes are sent to an automatic media-pouring unit, and the molten media is then poured aseptically into each dish before being cooled and sorted according to incubation temperature. The dishes are then delivered to the workbench.
The Copan Walk-Away Specimen Processor for liquid sample processing in microbiology automates all aspects of processing, planting and streaking, slide preparation and enrichment broth inoculation. It works on a modular basis and the company intends to introduce new modules, capabilities and applications to meet the needs of clinicians.
Costs vs benefits
The capability of the latest systems is evidently increasing, but they come at a price. Cost is always a big hurdle to overcome with any new technology and advocates, like Jerris, of lab automation are keen that the cost of developing and implementing more advanced automated systems should be viewed in the right light.
"The kind of technology we have now, such as the systems in the analytical phase that speed up the identification of organisms, can affect the decision about whether to use antimicrobials and not use antibiotics injudiciously," he explains. "The overuse of antibiotics is a personal bugbear of mine, so I think that on its own is a very important issue.
"If you can identify a virus quicker, then you will use antibiotics less and the patient’s stay in hospital will be shorter, with less need for additional tests and imaging. All of these things can bring costs down. That will be the trend of the future, though there are still some limitations. The benefits outweigh the limitations, but whether to invest in lab automation always comes down to a business decision. The cost question depends on the individual system and not everyone looks at ‘avoided healthcare costs’ rather than just ROI."
Considering the costs that could be avoided by using automated systems in the lab would change the equation behind the business decision, but many other factors also come into play.
"Some systems, like Kiestra and Copan, suit a large lab, but others could find their way into most hospitals, such as video imaging systems," says Jerris. "Components like upfront processing might suit medium-sized labs. Basically, you have to consider the scale of work that is going to be done.
"Automated identification and susceptibility testing could deliver huge economies of scale, but if you are only doing 50 tests a day, then the costs will not be justified. If you are looking at automation in the pre-analytical phase in my profession, then it might be easy to forget that there is a lower specimen volume in paediatrics and that needs to be addressed. You need to look at the complete package in order to justify the costs of lab automation, and many of the limitations depend on your patient population."
Clear vision
There is a huge number of variables in play in any investment decision concerning lab automation, and those decisions must be made with great care and long-term vision. Looking at the long term, it is clear that new developments in lab automation will occur frequently, and the scope for those advances is vast as lab automation covers a wide variety of processes and applications. Jerris hopes that that process of innovation will be steered by a clear understanding of the needs of clinicians.
"Lab automation can do some things better, especially susceptibility testing, which many people still do by hand," he says. "The question now is whether the equipment companies know the needs of clinicians well enough. What I would like to see in the future is more point-of-care tests that can be done at the bedside in order to get the results quickly to the patient. Doing tests as close to the patient as possible makes therapeutic decisions quicker.
"Advances in microfluidics and electronics allow us to do things much quicker than before, so the move towards point-of-care testing will accelerate with the technology that is already available. However, it depends on how regulatory agencies view these technologies, but that is something we have to deal with every day in this job. We need to focus on the seamless interaction of technologies in different phases and we must approach that challenge positively to optimise patient care."
