Quality control: keeping a tight grip on systems

There is a paradox about safety in hospitals. In more than 20 years of research I often found errors in 5-10% (and sometimes far higher) of actions, and yet there are certain areas of hospitals in which there might only be an error every 1 in 1,000 or 1 in 10,000 actions. How can there be such a dichotomy?

Insight came in June. One of the great pleasures in my job is helping run our twice-yearly Improvement Science Development Group (ISDG). These leading international figures in improvement help mentor our Improvement Science Fellows, and provide thought leadership in the art, science and education related to improvement.

The insight came when one working group at the ISDG reminded us that at the same time as Deming was promoting quality improvement, another American ‘guru’, the remarkable Joseph Juran, was promoting a wider concept.

Juran was an engineer who proposed Company Wide Quality Management, made up of three elements: the ‘Juran Trilogy’. These can be envisaged as overlapping circles on a Venn diagram: quality planning (to meet customer need), quality improvement and quality control. The Health Foundation has long championed quality improvement, and there has been an increasing recognition of meeting ‘customer need’ in health care, however quality control has been the Cinderella of the family. Perhaps now is the time for Cinders to go to the ball?

Observing areas in health care where there are very low error rates, you notice they are run by scientists – pharmacy manufacturing, biochemical tests, areas of medical physics, for example. Why is this?

I think there are several reasons. First, these disciplines come from the organisational culture of laboratories, which deal with a range of materials and require high degrees of accuracy, achieved through meeting standards. Second, they involve processing physical things that are widely recognised as hazardous. Third, there is a framework of laws, acts and external regulators governing these hazardous materials, which can be a protection from local scrimping. 

When I ran a hospital pharmacy our production unit made products for industry, as well as the hospital group. Quality was maintained by a host of tightly choreographed and controlled factors: professionalism of the staff, the standard operating procedures, worksheets, multiple checks before the accountable pharmacist signed off the product, manufactured batches being checked in the quality control lab, which had also approved the processes and paperwork…and so on. We were regularly inspected by a regional quality controller. Nowadays these premises and processes are inspected in accordance with the Medicines Act by the Medicines and Healthcare products Regulatory Agency, which ensures high standards are maintained, even if the trust is in deficit.

It might be thought from the above that quality control only applies to the world of physical compounds that have to be managed, however it applies to processes as well. The problems with hospital infections came about through the failure of quality control processes around human hygiene and cleaning – we have had to enforce quality control mechanisms in these areas, which are slowly improving the quality of the system. 

Sometimes we have quality control mechanisms in our systems but do not recognise them as such. The evidence suggests hospital doctors make an error in around 1 in 10 prescriptions. How do we know? Because that is the proportion detected by the ward pharmacist – the quality control mechanism designed into prescribing in the late 1960s. How many errors do the pharmacists miss? How many do they detect too late? We do not know – we are not measuring the system properly and hence it is prone to decay. 

All systems need quality control. A ‘Drift to Low Performance’ is identified as a dangerous system trap by Donella Meadows in Thinking in Systems, her pragmatic and accessible primer on the simplicity and complexity of systems. 

Systems have balancing feedback loops that pull the system back to its goal if it begins to drift. That is part of our job, as humans, in the system. However if we begin to believe that things don’t work, or that it doesn’t matter if we let standards slip a bit and prioritise other things, then the system will begin to degrade. The less the system is seen to work, the less people bother maintaining it, and a downward spiral results. The solution: keep standards absolute – just what quality control methods are designed to ensure.

At a time when cash is tight it is easy to cut back, to pile more on the same staff, to cut corners. And the system will appear as if it can take it…for a while, and then the drift begins to become apparent – sometimes gradually, sometimes as catastrophic failure.

What does all this mean for the NHS? I think it shows that we need quality control in our systems. We need standards, we need people accountable for them being met, and who feed back poor performance so others can learn – both that they lapsed and that standards must be maintained. We also need the involvement of powerful external others that can ignore the internal pressures in an organisation.

There is a lot of sense in the Juran trilogy – let’s keep our ‘customer’ focus, let’s keep improving quality, and let’s keep our health care systems performing properly by controlling their quality.

Nick is Director of Research at the Health Foundation.