If someone is to be put on trial, it's essential that the process is fair. And make no mistake about it, in a CQC inspection GPs are on trial - a trial of our practice, our staff, ourselves, and all we have worked for since leaving school.
Even in the most minor of criminal trials, steps are taken to ensure that those forming judgments do so without preconceptions, yet the CQC is sending inspectors into practices having done exactly the opposite, having primed them to expect a practice which might be putting its patients at risk.
Even if they try to look at the practice with fresh eyes, it is inconceivable that they will not be swayed by their expectations on a subconcious level.
GPs demand CQC compensation
Added to that, the CQC is now under intense pressure about the validity of its banding process, and the easiest way to defuse some of the anger is to 'prove' that the bandings predict performance. This need not be a conscious process for it to be important - there need be no conspiracy, nor any bad intent, only the natural workings of the human mind combined with an unwillingness to accept the possibility of error.
Such worries are hardly controversial, as the whole concept of the blinding of clinical trials has come about in recognition of these issues. Just as researchers are not offended by the necessity for such precautions neither should staff at the CQC be offended. Being human and so subject to bias is not shameful, but not recognising that fact or protecting others from the harmful effects of it is.
So what needs to be done? The banding process is clearly damaged beyond repair and the CQC ought to recognise the harm that will be done not only to practices but also to their own credibility by pursuing it. From now on, practices need to be inspected in random order to avoid inspectors bringing preconceptions to the process, and the CQC needs to learn that there is a difference between the release of data and the sharing of knowledge.
Dr Andrew Green is a Yorkshire GP, and clinical and prescribing lead for the GPC
