Automating the clinical microbiology lab

In spite of some exceptions, the clinical microbiology lab has been a late starter as far as automation is concerned.  It has also traditionally been viewed as ‘low tech’, especially when compared to its cousins in clinical chemistry or pathology. A variety of factors, however, have been converging to reverse such a situation.

Automation hampered by process complexity
One of the most important barriers to the automation of a clinical microbiology lab is process complexity. Unlike hematology or chemistry labs, which have little diversity in specimens and generally use standard collection tubes, microbiology laboratories need to work with a vast range of specimen types in a multitude of transport containers. The complex nature of  specimen processing and culturing and the ensuing lack of standardization have been major deterrents to automation.
Nevertheless, growth in the presence of automated technologies in clinical microbiology labs is now expected to accelerate as a result of several factors, above all rising demand. This requires agility and high responsiveness, making automation indispensable.

Ageing populations drive demand
Ageing populations with more-complex diseases and conditions require a growing number of tests – for example, to monitor implants and prosthetic devices for infections. The elderly also need greater care in medicating, since they are more prone to adverse drug events. 
In the year 2000, an article by Dr. Thomas T. Yoshikawa of the King-Drew Medical Center in Los Angeles noted that though “the major focus in infectious diseases for the past decade has been on young adults”, in the future “the vast majority of serious infectious diseases will be seen in the elderly population.”

Infectious disease, resistant bacteria
The rise in infectious disease outbreaks in recent decades is another factor driving demand for early detection by clinical microbiology labs, to contain their spread.
On their part, multidrug-resistant pathogens pose their own specific challenges. Delays in obtaining lab results leads to over-treatment of many patients – and increased antibiotic resistance.  In 2008, a team from Erasmus University Medical Centre in Rotterdam found that quicker microbiological lab turnaround led to “a significant reduction in antibiotic use” in a study of almost 1,500 patients.  This finding assumes considerable significance when one takes account of the fact that 5 years later, another study found that just 3% of community-acquired respiratory infections in the UK were guided by laboratory results.

The role of budgets and cutbacks
In an era of budgetary cutbacks, financial considerations too have reinforced demands for the automation of clinical microbiology labs. There is some irony here. Given the nature of a hospital business, it has been easier for administrators to assess the productivity of their clinical laboratories, determine return on investment (RoI) and justify new outlays – via quantifying and benchmark tests and staff numbers. Such an exercise has, in general, already been conducted for other hospital labs. It is now the clinical microbiology laboratory’s turn.

The above considerations are summed up in an article in the December 2013 issue of the journal ‘Clinical Chemistry’ which quotes Gilbert Greub of the Institute of Microbiology at the University Hospital in Lausanne, Switzerland.  He says that the key reasons for the Hospital’s decision to move toward a fully automated laboratory consisted of a shortage of financial resources and the concomitant increase in activity of the Hospital’s clinical diagnostic microbiology laboratory “of about 4% to 12% per year.”

Workflow improvement, staff shortages
Indeed, improvements in workflow and quicker test results are also directly related to growing automation. One of the most important collateral effects of this is the freeing up of staff for other work. 
In May 2009, the ‘Wall Street Journal’ warned about “the shrinking ranks of skilled lab workers” in the US, which pose “a potential threat to the safety and quality of health care”.  Hospitals, it continued, said that “it can take as much as a year to fill some job openings,” while an American Society for Clinical Pathology (ASCP) survey found average job-vacancy rates topping 50% in some states.
The ASCP survey also illustrated another interesting fact. Laboratories which were affected by new technologies found a decreased need for as large a staff. However, 75% of respondents said they were not affected by new technologies.  In other words, not only does automation seem to be an answer to staff shortages. There is also a lot of untapped room for growth.

Europe faces staff shortages too. A report by Belgium’s University Hospital at Leuven highlights the challenges of an ageing workforce, alongside major waves of retirement which have started recently and are expected to continue for several years. The problem is exacerbated by a decline in interest in labs as a career and the presence in the workforce of fewer young recruits. As a result, the paper warns, there is a “trend towards employing less-trained technicians.”

Transferring skills to points of need
The benefit of automation in the face of labour shortages is to utilize the skills of medical laboratory professionals where they are most needed and to automate tasks that are repetitive and do not require the comprehensive skill set of a trained professional.
For example, a laboratory could use an automated system for mundane and repetitive tasks such as “planting and streaking of urine samples and other liquid specimens,” while assigning a lab technician “to perform Gram stain review and processing of more-complex specimens, such as tissue.” 
At the other end, boredom can also be a problem. In a non-automated environment, lab staff frequently complain of poor turnaround (TAT), referring to the duration or idling time between inoculation of media and microbial growth.  By shifting monotonous tasks to automation, while assigning higher-skill tasks to a technologist, the laboratory reduces boredom and increases productivity.
The May 2009 article by the ‘Wall Street Journal’ quotes Dr. Carol Wells, director of the clinical laboratory sciences programme at the University of Minnesota in Minneapolis: “Many tests are automated, but that doesn’t mean a lab monkey can do them.” The machines, she continues, need careful monitoring. Should they “spit out a result” which does not make sense, only a skilled lab technician can catch a possible discrepancy and determine what is wrong.

Liquid-based microbiology, MALDI-TOF mass spectrometry drive demand
Automation of the clinical microbiology lab is also being driven by supply-side factors.
Among the first is the advent of new technologies, such as liquid-based microbiology and mass spectrometry. Liquid-based microbiology allows specimens of varying viscosities (e.g. stool or sputum) to be homogenized into a liquid phase, in order to enable greater consistency in the inoculation of medium.  Specimen elution from recent flocked-style swabs into liquid phase has also resulted in a significant increase in the release of viable organisms from the swab, in other words resulting in greater sensitivity for detection of microorganisms. 
The second technology is matrix-assisted laser desorption ionization–time of flight (MALDI-TOF) mass spectrometry. This permits the accurate and rapid identification of microorganisms isolated from clinical specimens. MALDI-TOF procedures “are highly amenable to automation because they are relatively simple, do not change based on organism, and are reproducible.”  In addition, target plate spotting and extraction of proteins “can be standardized for most organisms, and when combined with automation, automated crude extraction using the on-plate formic acid extraction method can be performed with minimal staffing.”

Specimen processing as entry point

One of the first points of entry by new technology in a clinical microbiology lab consists of front-end instrumentation to automate and standardize initial specimen processing. Automation of the front-end makes it possible for tests to be conducted as soon as specimens arrive, obviating the need for a separate ‘stat’ lab. 
Nevertheless, the impact of automation in specimen processing is not necessarily uniform, and depends on the needs of a particular laboratory. One study by researchers at Penn State College of Medicine and Medical College of Wisconsin estimated break-even point of more than 4 years for a barcode-driven, conveyor-connected automated specimen processing system which included plating and streaking, incubation, image acquisition, and digital microbiology.

Full- versus partial automation

As with other types of clinical labs, automation of the clinical microbiology laboratory can be classified into total (or full) laboratory automation or modular automation. Most automation systems have traditionally been designed for the larger, high-volume laboratory with substantial specimen throughput requirements. More recently, automated processing units have been designed for the pre-analytical section of smaller/medium-sized laboratories.
In effect, smaller labs can choose to automate only some of the processing steps. The investment required for automation makes such a choice imperative. The report by the University Hospital at Leuven cited previously notes that its own implementation of ‘full lab automation’ cost €3.2 million with €1.7 million in capital investment and another €1.5 million for refurbishment. Time needs to be also factored in. The Leuven study notes that their automation entailed 1 year in preparatory work and 1 year for adaptation. Against this, the savings achieved were equivalent to 3.8 full time employees (FTE).
Some hospitals begin small and scale up. For example, the University Hospital in Lausanne, Switzerland, started with two stand-alone automated systems for microbial identification. It moved after a few years “to a fully automated laboratory, by adding the missing pieces to the puzzle, i.e., smart incubators, high-quality digital imaging, an automated colony-picking system, and all required transport belts in between.”

A glimpse of the future
Today, state-of-the-art clinical microbiology labs have the potential to automate nearly all areas of testing, including inoculation of primary culture plates, detection of growth on culture media, identification of microorganisms, susceptibility testing, and extraction and detection of nucleic acids in clinical samples.   In the future, process standardization is expected to be reinforced by high-resolution digital imaging and robotics, and take automation in the clinical microbiology lab to wholly new frontiers.