Using real-time risk to profile and predict morbidity risk
In the 1990s, when the move away from paper to electronic health records began, the paper process was merely replicated rather than using it as an opportunity to look at data differently and achieve real transformation.
The challenge for us today is that this approach has limited the effectiveness of EHR systems to measure clinical outcomes, risk and avoidable harm. EHR systems were designed around process and procurement, not clinical outcomes. Current discussion in the United States has started to focus minds on the issue.
As a result, EHR data is analysed retrospectively so the default is to look, for example, at mortality as an end measure through structured mortality reviews. Current EHRs are not set up to highlight systemic clinical failings for fear of a backlash from clinicians who are against the use of such systems
In addition, the data captured is often only used within departments, further reducing its impact. Clunky interface issues also make it less likely that relevant patient data is entered accurately such as capturing patient comorbidities. Interoperability is a further challenge that needs to be addressed.
Factor in variability in EHRs ranging from scanned copies of paper notes (i.e. no way of easily extracting health data in a structured way) through to structured pathway management via electronic data entry/patient records as well as decision support tools and measurement becomes near impossible.
Given we can identify the factors that affect risk of deterioration in real time, we believe a rethink is required. After all, patient care moves in real time – EPRs do not. This leaves trusts at risk of failing to recognise deteriorating patients, or with poor operational visibility and suboptimal flow. Insights from data days, weeks, or months later limits the ability to make corrections and find opportunities for improvements. We should start making this insight into risk available in real-time.
Transformational change can only really occur when data is pooled, analysed and insights shared across an organisation in real time. With real-time data available it is possible to measure the risk of morbidity and therefore profile and predict risk. Clinicians can then focus attention on the patients most at risk of harm or those who are likely be more susceptible to complications such as sepsis or acute kidney injury (AKI).
The first step to ensure patient and related data is captured in a data-warehouse and not locked within proprietary EHR systems, the second would be to profile risk on admission as part of a targeted operating care pathway. Data captured by a real-time EHR can then be used generate and monitor patient risk throughout their entire episode and to optimise care and to reduce risk of harm.
This approach will help to improve outcomes and patient experience and CRAB is currently working with several hospitals across Europe to monitor real time risk of morbidity (both in terms of absolute risk and rate of increase in risk) for emergency patients with fracture neck of femur.
