The BMJ is campaigning against the harms of too much medicine. Yes, too much medicine can do plenty of harm. In the ensuing discussion one (slow) rapid response draws parallels with the harm caused by accrediting human behaviour. We knew that. But if you’re feeling powerless you can click to like it and contribute to the discussion.
Pathology laboratories have been purchasing accreditation to ISO 15189 and ISO 17025 standards from CPA and UKAS for years. More recently, its growth has spread into radiology and physiological measurement. “Quality” has been redefined strangely as compliance with these standards. Yet, as a consumer, no-one assesses the quality of any other product or service as compliance with ISO standards.
Do obsessively precise measurements and the collection of exhaustive records to enable inspection really improve patient care or is accreditation just an employment scheme for inspectors whose role is outdated? The value to ISO standards for products has not been clearly shown to extend to their systems of management inspection. The Advertising Standards Authority found against the BSI for its wishful claim that BS 5750 paid for itself in cost savings. In the USA, largely without ISO accreditation, labs fail to detect 1 in 20 pathogens.
ISO accreditation ought to viewed with the same suspicion as other medical interventions. Like other management fads it is likely to be wasteful and counterproductive. Accreditation is expensive and distracts from job purpose. Its efficacy, effectiveness and value have not been proven according to Cochrane’s principles. Accreditation has been designed to make verification very difficult but whether it is ethical to use it without rigorous testing is questionable.
If accreditation has the power to benefit patients it may also have the power to harm – probably through waste of financial and human resources and delivering a confidence that is false. The cartel of ISO inspection organisations needs to be subjected to validation from outside its membership and equivalent to that for pharmaceutical interventions.
1 in 20 failures leaves room for improvement but sounds like it might largely be explained by the normal distribution curve. If only the UK would publish how much better providing inspection fodder for UKAS makes lab performance!