Browsing the seventies

Having recently taken up a new post and office, I was lucky enough to inherit a series of Annals dating back to the 1970s. I’m sure there are some out there with an older and more extensive archive but I none the less felt compelled to explore and share some of my discoveries with you.

In my laboratory we are about to embark on the purchase and evaluation of new analysers. I was consequently tempted by an article looking at the Greiner Selective Analyser. ¹ It described the evaluation of six of the 18 available methods in terms of reliability, accuracy and precision. Economy of operation, temperature control and electrical/mechanical safety were also assessed! The analyser was capable of performing up to 30 different tests, selected on the basis of a pattern of holes punched in card, and worked at a rate of 300 tests per hour. Compare this to today’s analysers with over 1200 tests per hour and an extensive menu of over 170 plus tests.

Correlations were performed against an SMA 12/60 analyser – there was not a difference plot in sight. I think the best part for me was (and I quote) ‘inspection was made to see if injury could be caused to the operator by electrical shock or mechanical means during normal operation of the machine…’. I’m assuming operator consent and ethical approval for this assessment was obtained?

Demand management in the 1970s presented some interesting ideas and concepts. ² The article was introduced by the question ‘If a visitor from a distant world arrived in the department of chemical pathology, one of his first comments may be to ask….’; the same may apply today. The author questions the continued survival of the chloride test and the value of producing results which were not asked for or tests requested only for their known availability. Sound familiar? The article suggests the issue is caused by clinical research, medical consultations and the laboratory working in isolation and the increase in automation. Diabetes, and the over-use of glucose in the diagnosis and management of the disorder, was used as an example. Alberti states that this issue can be traced back to when the first cases of diabetes were recorded by a group of Indian physicians in 1400 BC as ‘honey urine disease’. He suggested that measuring metabolic profiles (growth hormone, lactate, ketones, glycerol and alanine) may be more useful than glucose. Trends change and today, with the possible exception of ketones, these analytes barely get discussed in clinical practice in this setting. The value of HbA_1c in diabetes was not recognized until 1976 by Cerami and colleagues. ³

A couple of articles^4,5 regarding paracetamol measurement caught my eye as they highlighted how the concept of method evaluation has changed over the years. The first article described a modification of the photometric method of Glynn and Kendal to measure paracetamol. In these days before STARD (Standards of Accuracy in the Reporting of Diagnostic Studies) was dreamt of, the method evaluation involved a correlation study of 25 samples with gas chromatography. On the basis of the results the authors were able to conclude that this rapid method ‘should be considered as the method of choice in an emergency’. The second article was a slightly more extensive evaluation of a paracetamol assay using the indophenol reaction. It advocated taking samples at admission and 8 to 12 h, using both the estimated paracetamol half life and concentration to determine the likelihood of liver damage.

A synopsis of The Vickers Award Lecture 1974 entitled ‘Application of Analytical Methods to Automated Analysers’ presented some interesting reading. ⁶ The article described the development (in a Brighton NHS laboratory by Dr C Riley) of the first ‘high speed’ multichannel analyser. The prototype could measure serum protein using the biuret method and used an enzymatic method for urea. The advent of a purpose built colorimeter (Vickers D300 analyser, Vickers Medical 1972) increased throughput from 60 to 300 samples per hour. The first two-point reaction rate determination for the estimation of creatinine with alkaline pictrate (Jaffé) reagent was developed by Dr J Cook in 1971. ⁷ The author described the advantages of the kinetic approach to remove interference from this method.

In the same edition, Dr Riley commented on the key role scientists had played in the development of laboratory equipment. He disagreed with the point made in a lecture marking the 21st Anniversary of the Association of Clinical Biochemists ⁸ in which it was suggested that in the future the initiative (to develop new instruments) would lie with instrument manufacturers. How different the situation is today in the extent to which we are reliant on automation and the input of the analyser manufacturers. Has there been a shift in emphasis in the laboratory environment towards us looking more into the clinical value of our tests and away from method development? This I think certainly applies to the high throughput methods where demand is such to necessitate automation.