Automation in diagnostic microbiology

Keywords: automation, continuous, culture automation, diagnostic, on-demand, microbiology, remote plate reading, shared technology, virtual.

My thanks to the guest editors for the invitation to share my journey and insights regarding automation in diagnostic microbiology (ADM). ADM is not that novel. It is more ‘old’ news rather than breaking news. It is not my aim to compare and contrast. What I would like to do is to focus more on how automation in the diagnostic microbiology laboratory (DML) can be used, its impact on service delivery and subsequently on the profession. When you stand at the door of a DML (Fig. 1), what do you see? The past or the future? Do you see a lot of instruments or hardly any? Is the space recognisable as a DML or does it look more like a biochemistry laboratory? Just as importantly – what can’t you see? A virtual DML evolving out of the ADM.

Certainly, instrumentation is a feature of the modern DML, typically performing multiple microbiology diagnostics in the same space. Instrumentation has brought it together by automating testing, whether urine microscopy, blood cultures, identification and susceptibility (ID&S), molecular testing and now even culture-based technologies. These have become more common place tools for the ADM laboratory managed through integration with the Laboratory Information Management System (LIMS). The impacts are staff savings, faster standardised results, interfacing LIMS, automatic reporting and possibly interpretation. Integration of automation (e.g. blood culture with ID&S, molecular, and MALDI-TOF) opens further opportunities for the provision of more rapid, meaningful clinical information. For most ADMs, this has been an evolutionary process driven more by a desire for efficiency and to proactively manage increasing workloads and restrictive budgets. The efficiency gains however precipitate consolidation. The same driver creating local health districts with governance over several hospitals and organisations to achieve consolidation efficiencies.

The options available for consolidating testing to take advantage of automation efficiencies are limited – either centralise the test and it is not performed on site or provide the technology on site. Molecular methods were seen as needing specialised skills and culture technology was hamstrung by growth requirements making referral a viable option. The cost of the technology was a further limiting factor given the smaller volumes on-site and availability of suitable logistics.

A centralised approach to consolidating DMLs had occurred within our region around 2008 and a number of regional and metropolitan DMLs closed. Regional units performed very limited tests on site and the rest were centralised, leveraged by suitable logistics, the 24/7 operations in our laboratory and the high level of automation in place which combined to minimise the clinical impact of transferring microbiology specimens off site. There was a distinct feeling of loss for these sites that extended to an actual loss of expertise and professional development of staff as well as to the clinicians and community. The appeal to work in regional sites diminished further adding to the difficulty to attract staff.

With the development of culture based automation (CBA) using digital imaging in the DML we were presented with a realistic alternative to centralisation that was previously unavailable – remote access. In October 2015 we went live with the BD Kiestra^TM Total Laboratory Automation (BDKTLA) and journeyed into a different world of microbiology. The principles are the same, but they are achieved through automation of processes and digitisation of images using monitors connected to the system. A more inclusive approach is to share the technology off site. Ironically, forwarding specimens for processing on CBA has the same impact for all sites including the host. Everyone forwards their specimens to the CBA for processing. Unlike centralised models, the referring laboratory can read plates digitally as if they had the automation physically on site. It also has access to the other host technologies – instrument-based susceptibilities, MALDI-TOF and molecular methods. The plate reading and interpretation is decentralised, and they become inclusive with enhanced professional development. The technology can also separate the work into folders that are site specific. There is minimal difference to having the technology on site. Participating sites have internet access to the technology and associated efficiency benefits. The virtual laboratory. They are advantaged by freeing up staff to expand on-site service and utilise quiet periods to support the remote reading. Staff are rostered over the 24/7 period to read and report plates digitally rather than in hand. Instead of the laboratory being just 30 sq m it becomes 6000 sq km and is at the forefront of ADM and the evolving virtual DML. It is a work in progress, not without continuing challenges needing significant recalibration of mindsets, change management, workflow review, communication and adjustments. It does not really stop.

Today we have evolved into a continuous processing, on demand DML operating 24/7 – testing, plate reading, ID&S and reporting. Everything on our scope is available over the internet and remotely. With two participating regional sites and a team of staff to work from home providing 24/7 support remotely – virtually. They are mutually complementary with flexible working arrangements making it easier to accommodate changing circumstances and commitments. This was especially helpful during the COVID-19 pandemic.

Due to the historically specialised nature of molecular processing and equipment use, molecular biology was developing into a separate subdiscipline, centralising molecular testing. However, automation of molecular methods into self-contained instruments integrating extraction, amplification and detection without introducing contamination has simplified the technology. It became accessible and affordable for the DML. This provided the opportunity to place the technology at referring sites to not just improve turnaround times (TATs) but to better manage workload to the referral site. It challenges the traditional centralised mindset of work efficiency to a decentralised model with improvements in TAT and equity of service access for remote communities. The opportunity for mini hubs over shorter distances has emerged providing local referral of specimens. This was certainly evident during the COVID-19 pandemic with coronavirus testing performed at regional sites on automated molecular instrumentation providing rapid results for regional public health units and enabling workload to be better shared than gridlocked into a central distant location.

Consolidating our platforms to one simplified automated molecular platform also allowed molecular methods to be integrated into our routine 24/7 service workflow to provide these assays on demand, without batching and easily used by out of hours DML staff. The samples were prepared ready to load when the platform was ready for the next run. Molecular testing evolved into a continuous process with no restrictions on menu availability with tests available on demand 24/7 in stark contrast to batch driven centralised models.

The 24/7 laboratory moved us beyond just providing rapid urgent results on request. It synchronised the DML to the continuous nature of automated testing, expanding our function over three shifts and the internet. This has avoided double/triple handling of specimens that occurs with batching. The added benefit was faster more relevant results that initially exceeded clinical expectations but quickly became the new norm feeding customer addiction to the improvements.

The shift by NSW Health to decentralising medical services and establishing Rural Referral Hospitals impacted not just workload but expectations of on-site services for regional and rural laboratories. The ADM opened applications to assist with workload and improvements in TAT via molecular automation and access to culture automation. This in turn provided offsets to expand local services by shifting the specimen processing off site but still retaining the reading and reporting – with the added advantage of access to the latest technologies such as MALDI-TOF. The rural community was receiving equity of service comparable to metropolitan and city communities.

The benefits were more obvious during the COVID-19 pandemic with the addition of coronavirus testing to Regional and Rural NSW. Timely rapid COVID-19 results were achieved enabling rapid tracking to protect the community the equal of any large teaching hospital and better than most. Regional testing also took a lot of workload pressure off referral laboratories. The same assays were used. The community felt significant support and not disadvantaged by delays in referring specimens to central testing laboratories. In turn additional support was provided by the referral laboratory via its automation of culture processing using the BDKTLA. This freed staff to focus on COVID testing while still maintaining current microbiology services.

Consideration of how ADM is implemented needs to be applied to the future. How it is implemented and how the profession of diagnostic microbiology will evolve and develop? Traditional centralisation could potentially decimate the profession limiting positions and training. It is the easiest option but not necessarily the best option. It is exclusive and no doubt appeals to some. It does not mean that the technology and what it offers should be ignored. So much would be lost. It is a case of how it is managed and the alternative approaches. As a discipline there is a broader consideration, the freeing up of sites, broadening service and buy into the technology sharing through the Virtual DML. An inclusive approach. Like all things there is a balance.

Automation in diagnostic pathology has typically been defined by the laboratory physical dimensions and location. Laboratories refer specimens to the central laboratory, which processes and sends out results. No further input by the referring laboratory. However, ADM has changed this and offers real alternatives.

The virtual laboratory is an inclusive laboratory sharing the technology amongst those that have access. Enabling enhanced professional development and improvement in services to the community. This can be fine-tuned, enhancing access to technology that was otherwise unaffordable and unavailable. There are undeniable challenges in sharing technology but these are surmountable.

Data sharing is not applicable as no new data were generated or analysed during this study.

The author declares no conflicts of interest.

This research did not receive any specific funding.