Blog

COVID-19 back-to-work staff risk management requires intelligent automation

Sabu Samarnath
Sabu Samarnath
3 min read

The Thursday-weekly nationwide applause from the windows and doorways of the public attests to the love and appreciation for a National Health Service under severe stress. But for all the joyous cheering and clanging of pots and pans, the NHS needs more than moral support in the throes of this current crisis. Occupational health teams are faced not only with the colossal task of advising public safety, but also assessing risk for thousands of hospital staff, and making decisions about who should or shouldn’t come in to work to help fight COVID-19.

These decisions are critical to the herculean efforts of the NHS; the health service needs to ensure that everyone who can be working, is working, and that preventable infection to hospital staff is avoided.

Earlier this spring, NHS 111 had 1.7 million queries in 15 days, with call volumes soaring 400% since the start of the coronavirus crisis in the UK. In some areas, the service is so overstaffed that operators have complained of sitting less than 1 metre apart, while desperate recruitment drives have been required in other areas, with huge numbers of healthcare workers even coming out of retirement to aid the cause. “Managers, floorwalkers and staff have all been thrown into this with very, very basic training”, wrote one call handler in a letter to MPs. It’s clear to all involved that the stretched and increasingly cobbled-together resources of hospitals need to be protected and maximised – so that they can continue to protect the public. But before each member of staff is cleared to aid the cause, a thorough risk assessment must take place.

The difficulties of accurate risk assessment (and mortal consequences of getting it wrong)

Risk assessment for staff is plagued by complications. These difficulties are in evidence, with NHS 111 and hospital staff known to have been “turning up to work with coronavirus symptoms”, as reported in The Guardian. This is a matter of resource management and scheduling: with numbers at breaking point, managers need to be able to plan for which of their team members will be in or out of action, and for how long.

Then there is the matter of conflicts between government guidance and hospital advice. There are subtle differences. These guidelines are often more complex than a simple yes or no question. Take the April 12th letter to regional NHS bosses: “If an individual living in the same household as a member of the NHS family tests negative then the NHS worker can return to work without themselves being tested, as long as they remain symptom-free and the whole household can come out of self-isolation.”

Without consistent, centralised and contextualised advice, damaging decisions could easily be made. Potential symptoms could be brushed off, particularly when some staff are being ‘denied proper sick pay’, according to a complaint made to MPs. Left to their own devices, those with mild symptoms are finding their way to an unappealing GANT chart to review the official guidelines – and judging by the statistics, have instead erred towards calling the inundated 111 helpline or occupational health teams instead. Most importantly, people who should never be exposed to the virus may be inadvertently put at risk.

There’s the sad case of Belly Mujinga, a COVID-19 high-risk individual who was working the concourse at Victoria Station in London—she then contracted (in the line of duty) and subsequently died from coronavirus.


To avoid such tragedies, the NHS is using an automated tool, powered by Rainbird, that saves occupational health teams time, allows managers to more accurately and effectively plan their resources, and helps the NHS safely maximise the availability of staff. Find out more in the full eBook below

Transform your business into a Decision Intelligence powerhouse

Explore how Rainbird can seamlessly integrate human expertise into every decision-making process. Embrace the future of Decision Intelligence powered by explainable AI.