Automation in UK clinical biochemistry

It seems to me that during the past 50 years one thing has changed the way clinical biochemists work more than any other: automation. The 50th anniversary of the Annals is a suitable occasion on which to review, through the prism of its archive, how developments in automation have unfolded during its lifetime. Currently, much of the focus of automation is on what happens to specimens immediately before and after they have been analysed. Thus core laboratory facilities have ‘train-tracks’ projected onto which are centrifuges, storage devices and of course the analysers themselves. The automation of the analytical phase itself is taken as a ‘given’ in this scheme of things. Yet it was not always so, and it is worth reminding ourselves of this.

Professor Tom Whitehead's inaugural professorial lecture from 1969 is a useful starting point. (Its publication in the Annals – just over a year later – is in itself noteworthy.) ¹ I can scarcely do better than quote from this in order to convey the impact of the first major analytical automation:

The invention of the AutoAnalyzer by a clinical chemist, Dr Leonard Skeggs, working in the Veterans Hospital, Cleveland, Ohio, was probably the most significant event which ever occurred in clinical chemistry. From an era where the performance of about 20 blood urea determinations was regarded as a reasonable half-day's activity by a technician, and one which demanded reasonable manual skill, we moved into an era where the technician could load a machine with sera, the machine would perform the analyses, displaying the results as a peak on a chart. An afternoon's work with virtually no manual skill could include, amongst other things, the analysis of 100 blood ureas with greater precision.

In a similar vein, a 1973 article tracked the changes in throughput in the Department of Clinical Chemistry at the Royal Infirmary, Edinburgh between 1950 (22,000 ‘determinations,’ i.e. measurements, by 11 technicians – preautomation – corresponding to 2000 determinations per technician per annum) and 1972 (779,000 determinations by 35 technicians – using 16 AutoAnalyzer channels – corresponding to 22,300 determinations per technician per annum). ²

Throughout the 1970s, the profession continued to embrace automation, recognizing the benefits it brought in terms of precision and workflow. A steady stream of papers in the Annals described the extension of automation to the measurement of specific analytes, and it was the focus of several Vickers Award lectures.^3–6 Formalized schemes for evaluating automated instruments used in clinical chemistry were published, outlining procedures for the assessment of mechanical and electrical features, measurement of analytical precision, carryover, cross-contamination, accuracy and linearity – in effect a prototype standard operating procedure for automated method evaluation. ⁷ There were also more fundamental studies on, e.g. aspects of continuous flow. For example, in the same 1971 issue of the Annals as Prof Whitehead's lecture, an original article examined the rate-limiting aspects of continuous flow analytical systems, developed a mathematical model to describe fluid flow in these systems and tested the accuracy of the model experimentally. ⁸ However, in an article written for the 21st Anniversary meeting of the Association in 1974, Broughton ⁹ sounded a note of warning that resonates today:

The major developments [in instrumentation] now come from specialist instrument firms. Many clinical chemists regret this because it takes some of the fun out of life, but it also means that industry will only make what will sell in large numbers. It also means that clinical chemists must first decide what they require and then persuade industry to make it. If they fail to do this they can only take what they are given…the profession as a whole seems to have been dazzled by automation, and to have come to believe that there is little it can do to influence the type of equipment, what it should do, or how well it should perform.

The author was clear that it should be the profession and not the manufacturers who should decide on the direction of travel. Nearly 40 years later, we can ask ourselves if the profession did indeed continue to play the leading role envisaged? The Annals archive would suggest not. If the profile of automation had waxed in the 1960s and 1970s, it began to wane thereafter. To be sure, papers describing automated methods continued to appear, but in a trickle rather than a stream. Of course, automation continued, but the close involvement of the profession in developments was less evident than before. There are several reasons for this. First, the undoubtedly successful automation of the analytical phase achieved (largely) between 1960 and 1980 inevitably and appropriately shifted the focus to pre- and postanalytical processing, a focus that persists up to the present time. Second, the substantial resources required for meaningful research and development in automation became increasingly difficult to procure in the context of a state of more or less permanent financial embattlement; for some time now it has been the exclusive preserve of the diagnostics industry. Third, the generation of biochemists who were at the forefront of automation in the 1960s and 1970s, cut their professional teeth in an era when in-house improvisation was quite common in a way that it no longer is (Table 1 of reference 9 provides an impressive list of eponymous instruments invented by clinical biochemists: Astrup acid-base analyser; Conway diffusion cell; Cotlove chloride meter; Delves cup; Evelyn colorimeter; Hartridge reversion spectroscope; Natelson microgasometer; Seligson burette; Thorpe nephelometer; van Slyke volumetric apparatus; van Slyke manometric apparatus; Folin and Wu tube). Finally, as Northam observed in 1980, the ‘poaching’ of National Health Service (NHS) staff by industry has occurred throughout the entire period of automation (‘over 1000 of our most experienced technicians left the NHS in 1964 alone’ ⁶ ) and continues up to the present day.

So what does the future hold? Contemporary clinical biochemists work in a regulatory environment that their predecessors would find suffocating; in particular, they would find the strictures placed by the In Vitro Diagnostics Medical Devices Directive ¹⁰ deeply frustrating (one is reminded of the phrase used by Evelyn Waugh to describe PG Wodehouse: ‘He will continue to release future generations from captivity that may be more irksome than our own’). Realistically, however, we cannot turn the clock back. This begs the wider question of how or whether clinical biochemists can regain the position of influence they previously enjoyed as a result of the changes to clinical management that were facilitated by the widespread implementation of laboratory automation. Perhaps future developments in technology – specifically in communication – will provide the basis for more ‘granular’ (i.e. more tailored, personal, one-to-one, ‘real-time’) clinical liaison than existing systems permit – an opportunity fundamentally to reverse what Northam called the ‘retreat into the laboratory [to develop and extend automation and data-processing techniques]'. ⁶ To conclude on another literary note, we should aspire to look forward with the same unwavering self-confidence shown in 1902 by James Joyce (aged 20, unproven talent) when he first met WB Yeats (aged 39, established literary figure): ‘You are too old for me to help you’. ¹¹