Can we measure the quality of nursing?

Is the work of a good ward nurse quantifiable? We asked Peter Griffiths, Professor in the Faculty of Health Sciences at University of Southampton and editor of the International Journal of Nursing Studies.

It’s always been important to understand the quality of nursing, because the things that nurses do are important to patients, with real impact – especially in the areas of safety and patient experience. But measuring nursing brings with it some unique challenges.

Indicators of quality

If you wanted to measure, say, mortality rates after cardiac surgery, you could find a clear link between the surgery and the outcome, focusing on an individual surgeon and the team around them. The link between the activity and mortality is reasonably clear: the activity is life threatening in terms of the operation and life-saving in terms of outcomes.

But in the case of ward-based nursing, you don’t have a single focal outcome. For many patients, the risk of death is essentially zero, so that’s not a good global measure (although what nurses do could be indirectly linked to any death). Factors such as presence of pressure sores and healthcare-associated infections are more closely linked to the quality of nursing care, and apply to many patients. But it’s not always straightforward to measure them.

Take, for example, pressure ulcers. These are linked to inadequate pressure relief, immobility and the patient’s general condition, so they are potentially a good indicator of whether nurses have been identifying need and delivering preventative care. But, unlike death, it’s surprisingly difficult to measure ulcers. The severity of an ulcer is subjective – it depends on the opinion of the person doing the measuring – and researchers don’t have access to high-quality routine data on ulcers, as they are not always recorded. We can measure some processes, such as the extent to which nurses use risk assessment tools, but the link between these tools and outcomes is not evidenced.

Another challenge is that if the ulcer were caused by poor-quality nursing care, you couldn’t link it to one individual nurse. This raises challenges when measuring patient experience too. If all the nurses are good except one, the patient’s overall feedback may be positive and that one nurse will slip through the survey. If someone generally likes the nurses, it’s hard to get them to be critical, even if their experience was very poor. I once spent months doing patient satisfaction surveys as part of a research project, and I’ve carried with me the memory of a ‘highly satisfied’ patient telling me: ‘The nurses are absolutely lovely, but sometimes they’re too busy to take me to the toilet.’

The solutions

When it comes to measures such as pressure ulcers, we need more work to risk adjust the data to recognise that different groups of patients differ in their underlying risk. For example, the risk of pressure ulcers for a young adult having their appendix removed is very low, so if they get an ulcer it could indicate that something awful is going on. A frail older person having hip surgery is at much higher risk, irrespective of the quality of nursing care. If we can solve this problem, it will be much easier to use these rates as measures of quality.

When it comes to patient experience, we need to do more to unlock feedback about all the nurses. If communication from one nurse in ten was poor, that’s what we need to know – not that overall communication from the team scored 9/10.

So is the work of a good ward nurse actually quantifiable?

The answer to the question posed by this article is: yes, we can measure quality in nursing. But, as these examples show, we can’t do it in the way that we would measure surgical teams or the performance of a whole hospital.

Currently, there are many local and national initiatives to measure quality in nursing, almost competing with each other. But they are under-resourced and underdeveloped for what they aim to do. We can’t measure everything, and gathering the data is costly. The challenge is to be modest: recognising that there are some things we can feasibly measure and some we can’t, and working on making measuring less costly by setting up a single framework, underpinned by rigour, going forward.