Automation in clinical bacteriology: what system to choose?

ABSTRACT

With increased activity and reduced financial and human resources, there is a need for automation in clinical bacteriology. Initial processing of clinical samples includes repetitive and fastidious steps. These tasks are suitable for automation, and several instruments are now available on the market, including the WASP^® (Copan), Previ-Isola (BioMerieux), Innova (Becton-Dickinson) and Inoqula (KIESTRA) systems. These new instruments allow efficient and accurate inoculation of samples, including four main steps: (i) selecting the appropriate Petri dish; (ii) inoculating the sample; (iii) spreading the inoculum on agar plates to obtain, upon incubation, well-separated bacterial colonies; and (iv) accurate labelling and sorting of each inoculated media. The challenge for clinical bacteriologists is to determine what is the ideal automated system for their own laboratory. Indeed, different solutions will be preferred, according to the number and variety of samples, and to the types of sample that will be processed with the automated system. The final choice is troublesome, because audits proposed by industrials risk being biased towards the solution proposed by their company, and because these automated systems may not be easily tested on site prior to the final decision, owing to the complexity of computer connections between the laboratory information system and the instrument. This article thus summarizes the main parameters that need to be taken into account for choosing the optimal system, and provides some clues to help clinical bacteriologists to make their choice.

INTRODUCTION

Most clinical bacteriology laboratories are experiencing an increase in the number of samples to be processed on a daily basis. As an example, in our clinical bacteriology laboratory in Lausanne’s university hospital, the number of samples to be investigated has steadily increased by about 4% per year, and this does not take into account the increased need for culture for epidemiological investigations owing to methicillin- resistant Staphylococcus aureus or to outbreaks of vancomycin-resistant Enterococcus [1].

However, human resources are not following this trend of increased number of samples, mainly because of strong financial pressure and resource shortages. Laboratory automation thus represents an appealing solution, especially for sample inoculation, which is a fastidious, repetitive process.

Pre-analytical handling of samples has been greatly improved in recent years by improved laboratory information systems (LISs) and increased use of bar-coding to trace samples and downstream processes, such as subculture, identification steps, and aliquoting [2–4].

However, despite improved LISs, the time spent on pre-analytical handling of samples and inoculation remains important, and in our own laboratory, which processes a mean of 300 samples per day and employs ten full-time laboratory technicians, we observed that about 24% of all technician activities related to sample reception, inoculation and Gram staining are dedicated to the inoculation of agar plates and broth. Thus, 50–70% of the time spent by full-time technicians may be saved by automation of this task.

METHODS

A first generation of automated plate streakers was developed more than 20 years ago [5]. However, the level of automation was still limited, and the automated inoculation instruments initially available, such as the Inoculab (Dynacon), had only unidirectional informatic interfaces.

Although updated with bi-directional connections with the LIS, the second- generation Inoculab LQH system (Dynacon) is only able to plate specimens starting from a single type of container (i.e. sterile urine container), and has a limited capacity of 38 inoculated plates, all loaded in a single silo [6]. Thus, these first-generation and second-generation systems were not developed enough to allow efficient, high-throughput and accurate inoculation of samples, including the following four main steps:

• selecting the appropriate Petri dish
• inoculating the sample efficiently
• spreading the inoculum on agar plates to obtain, upon incubation, well-separated bacterial colonies
• accurate labeling and sorting of each inoculated medium.

Very recently, third-generation instruments have become available on the market; these fulfill all of these prerequisites for automated handling of specimens in bacteriology laboratories. These new instruments include the WASP (Copan), Previ-Isola (BioMerieux), Innova (Becton-Dickinson) and Inoqula (KIESTRA) systems.

CONCLUSION

The current challenge for each clinical bacteriologist is now to determine the ideal automated system for his or her own laboratory. Indeed, different solutions will be preferred, according to the types of sample, and to the variety and amount of samples that will be processed with a given automated instrument.

However, this choice is difficult, owing to the limited data available in the scientific literature, partially because of the very recent availability of these systems in clinical laboratories, and partially because of the excessive level of confidentiality surrounding results obtained with prototypes. This review thus provides clues to guide the choice of clinical bacteriologists and provides a list of instrument characteristics that need to be considered (Table 1). This list should: (i) help for comparison between systems; and (ii) pinpoint the parameters that are important to choose an ideal automated system.