Online patient feedback, such as posts on social media sites, can provide accurate near real-time representations of the quality of care in NHS hospitals, according to researchers.
The researchers, from the London School of Economics, said they were currently in discussion with the regulators over how their system may be used to help prioritise future inspections.
“It gives us an understanding of aspects of care not captured by existing surveys”
They said they had developed and tested algorithms that reliably read and synthesise thousands of patient comments posted every day on websites such as Twitter, Facebook and NHS Choices.
Remarkably, when tested, they found the synthesised data could effectively predict the outcome of hospital inspections by the Care Quality Commission.
As a result, the study authors suggested their approach could be used to highlight the hospitals most likely to be performing poorly and flag them for inspection.
The researchers said the near real-time information captured by their system offered a much quicker insight into performance than existing surveys and official data collections.
They noted that, following the tragic events at the Mid Staffordshire NHS Foundation Trust in the late 2000s, there had been widespread calls to make better use of patient feedback in addition to traditional performance indicators, which were slow to identify the unacceptable levels of care.
Social media feedback can help identify ‘high risk’ hospitals
Study author Dr Alex Griffiths said: “The use of automated, near real-time patient feedback provides an opportunity not only to spot and rectify declining standards of care before they become too serious, but also to quickly identify improvements in care and learn what is behind them.”
“Aggregating comments from multiple sources allows us to gain insight from different demographics helping to reduce the obvious problem of bias that comes with using a single source of information, such as Twitter,” he said.
“Moreover, it gives us an understanding of aspects of care not captured by existing surveys, such as interactions between staff and carers at multiple points along care pathways, and often at a more granular level,” he added.
The findings from the study have been published in the journal BMJ Quality and Safety.