Sanome

NHS Problem

Every year, about 28,500 patients in England die from Hospital-acquired Infections (HAIs) like sepsis and pneumonia, surpassing some cancer fatalities. These largely preventable infections not only lead to the highest adverse event in healthcare, affecting over 840,000 cases per annum but also cause significant patient re-admissions, frailty and other long-term health issues, as well as contribute to antimicrobial resistance. HAIs cost the NHS £2.7 billion annually and occupy between 5-10% of all hospital beds, mainly due to prolonged hospital stays of >9 days, leading to an estimated extra 7.7 million bed days per annum. The cost is comparable to that of all smoking-related diseases.

The Solution

Sanome’s first product is a clinical co-pilot to support clinicians in identifying patient deteriorations sooner. The first intended use is an early warning system (EWS) for HAIs for hospitalised patients in secondary care. 

Sanome’s multimodal AI model uses routinely collected hospital data from the EHR (clinical observations, prescriptions and medical history). The continuous risk assessment algorithm calculates a patient’s percentage risk score of HAI deterioration within the next 48 hours. The risk score, alongside a “next best action” suggestion, and explainable AI is displayed back into the EHR to support clinical decision-making. Regulated as a class 2a Software-as-a-Medical Device.

Impact

The NHS faces a substantial financial burden from HAIs exceeding £2.7 billion annually, a figure comparable to the costs of smoking-related diseases and representing 1.6% of the NHS’s 2023/24 budget. The primary cost driver is the extended hospital stay, averaging 9.8 days. Annually, HAIs result in over 840,000 cases, occupying >7.1 million bed days. Our research indicates that earlier detection of HAIs could cut the length of stay by 33%, potentially saving up to £1 billion and freeing ~2.3 million bed days each year. A conservative estimate, accounting for potential false positives, suggests a 10% reduction in length of stay, equating to ~£270 million in savings and the liberation of 710,000 bed days, or ~ 2% of all NHS acute beds. This reduction could enable treatment for an additional 136,000 elective care patients, addressing 1.7% of the current backlog. 

Direct costs are significant, with 16 million admissions and a 6.4% HAI incidence rate across NHS trusts. This translates to roughly 5,600 HAI cases per trust annually. With each case costing an average of £4,000, a 10% reduction in length of stay through our product could save over £2 million per trust each year.