In safe hands? The need to know more about safety in health care

A paradox runs through our approach to safety in health care. A huge amount has been written and spoken about the need for an open and transparent safety culture. Encouraging staff to report safety incidents should enable risks to be reduced, and support learning and improvement. But in reality health care leaders find it difficult to acknowledge the extent of safety issues and actual harms which occur, and translating policies of openness into practice remains a challenge.

The National Reporting and Learning System has been collecting data on patient safety incidents since 2003. This plays a vital role in identifying emerging safety risks and setting national priorities for action, such as acute deterioration. But attempts to use the data for monitoring are fraught with difficulty. There is extensive evidence to suggest that incidents are under-reported, yet a high and increasing level of incidents does not necessarily equate to less safe care: it might simply be the result of changing patterns of reporting and recording. For example, improved awareness of sepsis as a safety issue means more cases are identified, and trends in harm data from the Safety Thermometer tell us little about national safety trends.     

A better understanding of the current picture regarding safety will help to improve the quality of care.

Measure safety rather than harm

A shift is needed in order to recognise that health care is a high-risk activity: instead of seeing incidents or harms as one-off events, we need to anticipate that problems will occur and think about how we can organise care to make it safer. A framework for measuring and monitoring safety has already been developed that works on this basis. Using discussions with health care providers, learning from other sectors, and findings from research, it starts with the assumption that health care is high risk, and takes a rounded view of how services can keep patients safe, encouraging learning from any harm that has occurred.

What does the data we have tell us?

As part of our QualityWatch programme with the Health Foundation, we have reviewed the indicators we use to monitor safety.

Improvement has been achieved and maintained for safety issues where there has been dedicated action over a number of years – for example in health care-related infections and cleanliness. There have also been recent improvements in the management of deep vein thrombosis.

But the pressure under which the NHS is now operating on a daily basis is evident in other safety measures. Given the progressive increases in people waiting over four hours to be admitted in an emergency, and the concerns raised by CQC about urgent and emergency care in acute hospitals, it is not surprising that patients with a hip fracture are waiting longer for their operation. There has also been an increase in pressure ulcers for patients who have been readmitted in an emergency.

There have been small improvements in measures of safety culture from the NHS staff survey, although there is worrying variation between care settings and staff groups. In particular there has been a fall in the proportion of medical and dental trainees who feel secure raising safety concerns, with only 12% strongly agreeing that they felt secure in 2017.

And there is still not enough engagement of patients in safety: over two-fifths of inpatients report that they were not told about the possible side-effects of medication before they went home.

Safety gaps

So the picture of progress on safety is somewhat mixed. But there are also important gaps in the information we have available to monitor safety.

For example, we were unable to identify any measures for tracking how well NHS organisations monitor safety on a day to day basis.  With the growing use of digital systems to record day to day clinical care, this is an area where we can expect to see developments.  There is limited information about safety in primary care at a national level, and we also have fewer measures of safety in mental health and community settings than for acute care.   We also know very little about the safety of healthcare for patients in their own homes – the setting where most patients spend their time. 

The growing number of people living with long-term conditions also requires a radical rethink of how we measure safety.  Most current measures of effectiveness of care are based on treatment of individual diseases, whereas most patients with long-term conditions have more than one condition.  Measures of optimal care need to take account of this, and balance disease related measures with patients' overall wellbeing and the treatment burden on patients.   

And finally, there is a need to find out more from patients about the safety of their care – for example, how reliably they can monitor their own condition, and their experience of worrying symptoms or services which are not coordinated. 

This is likely to challenge our perceptions of the safety of the NHS even more, but ultimately will increase our understanding of how safety can be improved from a patient perspective, rather than from a disease or service viewpoint.