Three things the NHS must do to manage clinical risk

I recently presented a clinical scenario to some GP colleagues – they were asked to comment on the management of a complex frail patient living alone. They could choose one of two responses – either to continue their current management, or immediately admit the patient to hospital.

Not surprisingly around 90 per cent of the GPs chose the first option. Clinicians tend to see continuity of care as best unless the patient has needs beyond their expertise.  

I then met with a large group of hospital doctors and posed the same scenario to them. Unsurprisingly the response was similar: keep managing the patient, no need to transfer care. However, there was an important difference between the two groups.

The GPs were asked if they wanted to immediately admit the patient to hospital, and the hospital teams were asked if they wished to discharge the same patient. Essentially, the GPs wanted to manage the patient at home and the hospital doctors wanted to manage the same patient in hospital.

How can that be? Two groups of medical professionals assessed the same complex case – and they chose an entirely separate place of care.

• Special report: Integrating all partners works
• Burstow: How to give people a life, not just a service

Disintegrated care

That makes no sense from a health provision perspective. It is not a surprise though – the new models of care are here to address the issue of disintegrated care.

However, before leaping into new structures, it’s worth having a think about why care ended up disintegrated. If we don’t understand that, we won’t design the right solution.

It is easy to blame the transactional basis of NHS care – the GP referral system, and the contracting consequences of that. However, among the many barriers to integration, the unrecognised one is an almost complete failure of the health system to quantify, define and share levels of acceptable risk as the patient transfers across the primary/secondary care interface – and that matters if we are going to integrate.

‘No one was addressing the risks around the interface’

Think of the scenario posed at the beginning – why did the hospital teams feel the patient needed to remain in hospital, when their GP colleagues were quite happy managing the patient in the community? It is because no one was addressing the risks around the interface.

Clinicians implicitly agree risk, rather than explicitly addressing and sharing it. An admitting GP might say “they need admission because I am worried about them”, and the hospital teams will talk about a patient being “a risky discharge”.

These conversations are vocalising perceived unacceptable risk but we only talk about risk when the perception is that an unacceptable level of risk has been reached. The conversation around risk is almost always about managing risk down. 

The problem with that approach is the missed opportunity around appropriate risk taking, and the lack of discussions on sensible mitigation of clinical risk. This is the conversation we need to start if we are going to truly integrate care.

Three things are needed:

• to recognise and share appropriate risk;
• to quantity what is acceptable risk – the clinical risk appetite; and
• vitally, to build a system where responsible clinicians who manage risk correctly are applauded and supported, not criticised, when adverse events occur.

Share the risk

Sharing the risk is the starting point. Every hospital admission or discharge should have an explicit understanding and agreement of risk, which is shared between the GP and hospital clinician.

That should, as far as possible, quantify and mitigate the risk. The risk assessment should be shared with the patient, and routinely documented as part of the discharge papers and in the patient record.

Clinicians need to be guided in their conversations by an agreed and accepted clinical risk appetite. An aspiring integrated care organisation (ICO) needs to formally address and attempt to quantity its clinical risk appetite in the same way it would do its organisational risk.

‘The risk assessment should be shared with the patient’

The discussion to formalise it should be between patient representatives, clinicians and managers. The ICO should also understand how it will respond to any future deviation from the agreed risk appetite – and that response should support the change in behaviour it wishes to see.

It would be all too easy for the ICO to define a risk appetite, but in reality manage the system in a way that pushes clinicians to be entirely risk averse.  

Managing and mitigating the adverse events that will arise then becomes vital. Unless our risk appetite is zero, then by definition our clinical system will see adverse events.

There are no prizes for risk taking in healthcare – the reverse is true. Clinicians find being the recipient of complaint and criticism hugely stressful.

Unless we truly support and encourage clinicians to do the right thing for the patient, and to take measured, appropriate and managed risk – not only will we harm patients through an over-conservative approach, but we will not succeed in truly integrating care.

Jonathan Inglesfield is medical director of NHS Guildford and Waverley Clinical Commissioning Group, and lead for the CCG’s Integrated Care Programme