Daily Insight: New forms of error

E-prescribe, but get it right

Trusts across England are being strongly encouraged to take on e-prescribing systems as they are meant to reduce medication errors.

However, the Healthcare Safety Investigation Branch has found the use of “poorly implemented” e-prescribing systems in hospitals is putting patients’ lives at risk.

HSIB launched an investigation after an elderly cancer patient was mistakenly given two strong anticoagulant medications for two weeks in 2018 after an e-prescribing error meant her community pharmacy was not informed that one of the prescriptions was meant to be cancelled.

The pensioner died from her cancer less than three weeks after the error was made – although it is not thought the mistake contributed to her death, according to HSIB.

The case highlights the importance of IT interoperability across NHS services, as well as the issue of poor e-prescribing implementation.

HSIB’s report says that while e-prescribing systems can reduce prescribing errors their introduction creates “new forms of risk and error”.

It is difficult to assess the impact of the issue as there are limited statistics available.

The report said: “In the US, the degree of harm and impact on quality and safety is not well known because of the varied nature of health IT products and the lack of a uniform measurement to assess the impact. The position in the UK is similar.”

In tech we trust

The rise of artificial intelligence promises much to the NHS, but it is yet to revolutionise services in the way some excitable scientists have predicted.

Yet here and there, claims emerge that give a glimpse into how new analytics might one day make a real difference to the way patients are treated.

One example is within population health management. In York, the claim is that AI was deployed to identify patients at risk of returning to hospital – before a team of health nurses provided them with intensive health coaching.

The result, according to those involved, was for those patients selected for involvement, a 30 per cent reduction in unplanned hospital admissions, ditto for emergency department attendances, and a 25 per cent decrease in planned hospital admissions.

The project is part of a randomised control trial which is significant for its size, patient selection, and length. So far that is not published, and these are interim findings, which some will wisely interpret with caution and a sceptical eye.

The activity reductions all refer to the difference between two patient groups, one of which received health coaching, while the other did not.

Its results prompted Vale of York Clinical Commissioning Group to commission the service for up to 1,800 patients.

The CCG also hopes to expand the service into primary care.

Under the plans, GPs would identify patients with long-term conditions who attend their surgery on a regular basis. These patients will then be offered health coaching.

If the technology allows it, the “AI” analytics would later be tasked with identifying the at-risk patients – instead of the GPs.

We await further details with interest.