what the hoo-ha's about24 Feb 2014 22:22
A friend who knows I own some of these shares kindly sent me a note from Lidco's broker, below, and following that, the Times article. I hope this helps.
You may have seen the article in The Times on Saturday relating to Evil Knievil shorting LiDCO shares on the basis that the Optimise trial will show that the technology doesn’t work.
Dr Rupert Pearse, the lead investigator, presented the headline data at a conference in Paris last year (i.e. raw results before detailed analysis). This was webcast and I watched it, and a subsequent interview with him. There are several absolutely key points, and to understand them, one needs to know that for a result to be ‘statistically significant’ in a clinical trial it needs to have a ‘p value’ equal to, or less, than 0.05. This means (given the inherent variability in biological systems) that there is a 95% probability that this was a real result, rather than pure chance. A p value of 0.1 means that there is still a 90% probability that the result was real.
Back to the results of Optimise. To summarise what Dr Pearse said:
• Complications were reduced in the intervention group from 44.4 to 36.6 but just missed significance (p= 0.07).
• Chance of death at 180 days post surgery was reduced from 11.6% to 7.7% - but again not quite significant (P=0.079).
• Hospital stay was 1 day less in the intervention group – just missing significance (p=0.054).
So the results were all trending towards significance for improvement in patient care. Unfortunately the raw results just fell just short of statistical significance. Dr Pearse in his presentation made the following observations:
• The lack of significance was most likely due to the study having insufficient power due to the number of patients enrolled.
• There were a number of emergency/non elective patients entered on to the trial that might have skewed the results. It is not clear why these patients were entered as this was supposed to be a study on elective i.e. non emergency surgery patients. These patients were therefore included in contravention to the protocol, and excluding them is perfectly valid. These re-worked data have not yet been presented.
• If the first few patients treated in each hospital are excluded, i.e. accept that there was a learning curve associated with the protocol – then the results ARE significant.
So I’m afraid Evil Knievil is wrong. He should listen to what the clinician said, not what people are writing on bulletin boards. The paper with full data analysis and peer review has yet to be published. No doubt when it is it will take account of these last two points.
Then there is the recent AAGBI Guideline for the perioperative care of the elderly published in Janaury 2014, three months AFTER Dr Pearse’s presentation (Source: Anaesthesia, 69 (Supplement 1), 81-98). I quote verbatim, [this was published in our Morning