Analysis: Clinical Laboratory Automation

Evolution Of Automation

Prior to the introduction of automation in the 1950s, laboratory testing was a completely manual process that required technologists to hand-pipette specimen and reagent into glass test tubes, incubate, read reaction changes in a spectrophotometer, and extrapolate values from a calibration curve that was plotted on paper. These tasks are still performed today—only on automated platforms that eliminate manual intervention, significantly improving the reproducibility, accuracy, flexibility, and turnaround time of testing while lowering costs, eliminating errors, and enhancing patient safety through robust sample identification and test reporting technologies (Battisto, 2004; ).

The first automated analyzers were based on continuous-flow technology (i.e., reagents would continuously flow through the channels); they sequentially analyzed each sample in a single or multiple parallel channel configuration. The same group of tests was performed on each sample irrespective of which tests were ordered; in other words, there was no test selectivity. These devices were followed by variable profile analyzers that could perform a subset of tests from the master profile or suppress those that were not ordered. The one characteristic shared by these analyzers was that their operating cycles were sample-based—that is, they processed a set number of samples per hour irrespective of the number of tests ordered per sample. Another approach to testing was the batch analyzer (e.g., centrifugal analyzer), which could perform a single test on a large number of samples. These devices were “run-based” systems; once the batch testing began, it could not be interrupted to add a new sample. By the early 1980s, the inflexibility of these designs started to be addressed with the introduction of a totally new concept: the random access analyzer.

Unlike its predecessors, the random access analyzer selects sample and reagent randomly without regard to placement on the analyzer. This type of operation implies several unique characteristics that differentiate it from other instruments:

• 1.

  It has a single analytic pathway for all testing instead of dedicated channels for each test.

• 2.

  Tests are initiated sequentially based on what is ordered—that is, tests are not batched.

• 3.

  It has a flexible configuration to process each test’s unique parameters (e.g., wavelength, number of readings, reagent additions).

• 4.

  It performs tests discretely, each in its own cuvette.

• 5.

  Throughput is based on the number of tests run per hour, not the number of samples per hour.

Virtually all of today’s automated chemistry systems are based on random access design. This flexible design has allowed many tests that previously were performed on a variety of platforms to be consolidated into one.

Although the random access analyzer was originally introduced for general chemistry and homogeneous immunoassay testing, many of its operating characteristics were adapted in the 1990s to heterogeneous immunoassay testing. These tests typically require more flexible assay parameters, such as variable incubation times, reagent additions and wash steps, and unique detection systems (e.g., luminometer) that require testing to be performed in a specially designed unit. Because of these testing requirements, immunoassay throughput is significantly lower than that of a general chemistry analyzer. Although immunoassay analyzers were originally introduced as standalone units, they are also typically available as an integrated subsystem within a single chemistry/immunoassay platform.

After great strides were made in automating sample analysis, more attention was focused on the preanalytic processes that increasingly accounted for errors and workflow delays. These included sample labeling (or relabeling with bar codes), loading and unloading centrifuges, manual aliquoting (e.g., splitting samples to be used at several workstations or transferring samples into another container to be sent out for testing or used by in-house instruments), and sorting samples for different workstations. Each step could potentially lead to an error, such as a misidentified sample resulting in an incorrect report. Finally, the uncapping and capping of samples posed biohazard risks due to aerosol exposure and contact with blood and body fluids. In response to these issues, preanalytic automation was introduced in the form of standalone units or preanalytic processing modules integrated into an automated laboratory configuration.

In addition to refining the capabilities of automated analyzers, in the past 10 to 15 years, laboratories have primarily focused on automating the entire testing process from collection to reporting, with a major emphasis on integrating these activities and optimizing preanalytic tasks and sample storage. Total laboratory automation (TLA) is the culmination of this design strategy ( ), one that uses intralaboratory sample transportation systems such as conveyors or tracks to move samples between analyzers and into and out of processing stations that centrifuge, decap/cap, and store sample containers. TLA is discussed in greater detail later.

As we enter a new era in health care, with enhanced emphasis on value-based management, the diversity of laboratory automation solutions is well positioned to meet the needs of small-, medium-, and high-volume clinical laboratories ( ). Some of the factors that will continue to drive automation are presented in Box 6.1 .

Preanalytic And Postanalytic Automation

The Automated Chemistry Analyzer: Core Components

The automated analyzer is central to the operation of the clinical laboratory. Chapter 4 discusses many of the testing principles used by these devices. This section discusses various physical operating characteristics and design features that enable these devices to function in an automated manner ( Box 6.3 ).