Trends.
Predictive behavioural modelling allows our clients to anticipate future trends and make informed decisions with confidence.
Whether it's optimising operations, enhancing customer experiences, mitigating risks, or improving performance, in today’s fast-paced world, staying ahead of the curve is more crucial than ever.
The following case study is a great example of this.
Case Study :
Transforming Crisis Management During COVID.
Regional Integrated Care Board (ICB).
Summary
This case study highlights how SoMoCo’s discovery of a critical bias in COVID-19 testing data transformed the management of hospital bed demand during a national crisis. By replacing biased data with data from wastewater viral loads and using behavioural insights derived from a deep dive into COVID related behaviours in the local population, SoMoCo were able to accurately predict hospital bed demand 10 days in advance.
An accompanying COVID dashboard provided real-time, hyper-local risk assessments, enabling the Integrated Care Board (ICB) to alleviate pressure by discharging patients into lower-risk areas.
This intervention was crucial during the COVID-19 pandemic, significantly enhancing patient safety and preventing healthcare systems from being overwhelmed. It not only set a new standard for public health response but also left a lasting legacy in how health services manage capacity and respond to crises.
Context
In the second year of the COVID-19 pandemic, the emergence of new variants and localised outbreaks led to unpredictable spikes in hospital bed demand.
Facing imminent risk of overwhelmed hospital capacity, our client, a regional Integrated Care Board, urgently needed help to avoid going into crisis and ensure patient safety.
Behavioural Analysis
By linking of multiple data sets we were able to reveal a significant bias in the LFD (community testing) data used by the Integrated Care Board.
This data disproportionately represented more affluent and educated individuals, while underrepresenting disadvantaged and underserved populations.
This crucial anomaly, previously undetected, stemmed from behavioural differences in reporting the results of home tests (LFD).
Strategy
Remove Bias: Find a way to replace the biased self-reported test data with more accurate sources.
Predictive Modelling: Use advanced predictive behavioural modelling to forecast hospital bed demand several days in advance, allowing the ICB to make swift, informed decisions.
Dynamic Dashboard: Use unbiased infection data to create a dynamic COVID dashboard capable of displaying real-time fluctuations in COVID levels at a hyper-local level, enabling resource managers to identify safe locations for additional bed procurement in care facilities and activate home discharge for more stable patients.
Implementation
We brokered an agreement with the UK Health and Security Agency (UKHSA) to replace biased test data with routine sampling of COVID viral loads in wastewater.
Leveraging insights from a prior behavioural audit, we developed a statistical model capable of forecasting hospital bed occupancy 10 days in advance, achieving high accuracy within narrow confidence intervals.
By combining accurate data sources, we were able to calculate COVID risk (viral load concentration) with population vulnerability (immunity levels and behavioural responses). The result was a digital, dynamic COVID dashboard that geo-mapped and ranked small geographic areas (Lower Super Output Areas) and large businesses, in real time.
Areas were ranked according to risk and vulnerability providing a more sophisticated understanding on where infection was present but more importantly where there was less risk of rapid transmission.
Results :
Greatly enhancing patient safety at a time of crisis, the ability to predict bed demand surges 10 days in advance with high accuracy allowed the ICB to proactively manage hospital bed capacity during COVID-19.
The COVID dashboard's real-time ranking of local areas by risk and vulnerability enabled commissioners to secure additional beds in low-risk care homes and select suitable re-ablement locations.
This greatly enhanced patient safety during a national crisis, demonstrating that informed decision-making can significantly improve emergency health outcomes. This commission set a new standard for public health response.
Testimonial :
“As a leader, the ability to say, ‘I know this now! Let's plan for this now!’ gives you the head space to understand what's coming and to act at pace, because a new COVID wave was on you before you knew it. We now knew what we were dealing with. We could see it. It was there. It was evidence-based, and we got that.
You take a lot of comfort from that. Comfort is probably the right word. You’d want this system all day, every day, behind you, to be able to inform and articulate what is happening. ”
DANIEL MCCABE, HEAD OF URGENT AND EMERGENCY CARE - INTEGRATED CARE BOARD (ICB)
Predictive modelling of hospital bed occupancy, accurate 10 days from date of prediction.
Cheshire East's Isoa's at most risk to increase in Covid cases.