Trust introduces the “suspicion of sepsis” tool to identify the cohort of patients infected by sepsis, achieving positive results

Summary

Sepsis is a severe life-threatening condition. Despite its seriousness, we do not have any consistent method of measuring its scale or outcomes. Till now, we have relied on inaccurate estimates aimed at measuring the ill-defined and fluctuating cohort of sepsis patients. So, we had no consistent means of defining the baseline for sepsis outcomes nor a reliable way of demonstrating improvement.

In response, we have created the Suspicion of Sepsis (SOS) concept and national dashboard. In England, SOS is the admission code in 1.9 million emergency admissions annually and is responsible for 28-38% of emergency admissions. An SOS code confers 3-5 times the mortality of non SOS codes and it is the admitted reason in 60% of patients that die.

[2] Instead of focussing exclusively on sepsis patients, SOS covers the cohort of patients who are admitted into hospital in emergency with infection, which can cause sepsis. It is based on a validated set of 200 ICD10 codes and uses readily obtainable administrative data. By bunching together all infection codes, SOS eliminates the fluctuation previously linked to the measurement of sepsis and creates a reliable measure over time to support local, regional and national improvement in outcomes.

Challenges

Sepsis is the severe end of infection or ‘infection with badness’ and is a massive healthcare problem with high estimated mortality and burden. However, it doesn’t have a gold standard diagnostic test or stable definition. Sources claim there are from 36,000-260,000 cases and 9,000-120,000 deaths in the UK/year. This has stifled ambitions to understand the scale, improve sepsis care, antimicrobial stewardship/resistance initiatives, to operationally define it and test the impact of biomarkers. It fulfils the brief of a wicked problem, and this lack of understanding of the ‘N’ has hampered improvement through an inability to measure progress.

To help combat this, there was a pressing need for a credible, reproducible and easily obtainable intelligence tool relevant to those admitted with infection. Such data could be utilised through the application of quality improvement methodologies in organisations and regions seeking to improve their outcomes. The ambition was to create a dashboard that could be accessed by patients, clinicians, managers, data analysts and researchers to identify the infection population and be meaningful in the context of all of these audiences. At the same time, the team wished to showcase the power and importance of the use of data in the right context.

Outcomes

The outcomes were around validating SOS, its utilisation across England, applying it to national data for outcomes and burden and developing dashboard for all to use in measurement for improvement.

1. Validating SOS and research around it, which is/has been accomplished in 6 ways over the last year.

•BMJ paper[1] was published on SOS within the Oxford region

•A study demonstrating that 93.3% of patients with confirmed sepsis had an SOS code

•A bacteraemia study demonstrating that 86% of patients with a positive significant bacteraemia had an SOS code[3]

•A study analysing National Early Warning Score [NEWS] in SOS diagnoses, demonstrating an AUROC of 0.83-0.87

• A 5 site clinical notes review SOS validation study

•An SOS bow tie analysis [4] to establish the factors associated with bad outcomes

2. Analysis of SOS using Hospital Episode Statistics [HES] data at NHS Digital to ascertain national burden [2], future research and the potential for a national dashboard that has demonstrated strong deterioration signal and value in tracking sepsis outcomes.

3. Analysis by the Department of Health looking at SOS in death certificates [2] that has demonstrated strong signal to infection and sepsis.

4.The creation of the national SOS dashboard [5]Describe and demonstrate how you have enabled others to use your initiatives across other departments and organisations. SOS was presented at key national meetings [9] including Patient Safety congress 2018, National PSC 2017, ICS SOA 2017, NICE. It has had strong support/usage from the National Cross System Sepsis board[6], PHE, DoH, PSCs and front-line clinicians, eg: National- The Sepsis CQUIN, which was focused on improving processes in the early identification and treatment of sepsis, has improved national SOS outcomes[2] Regional The PSCs have also used the SOS with the West of England AHSN[7] (6 acute trusts and a population of 2.4 million) being able to demonstrate substantial, above expected SOS lives saved due to the regional implementation of National Early Warning Scores(NEWS).

SOS was integral to the project and has directly influenced the national mandate for uptake of NEWS across all acute and ambulance trusts. Oxford AHSN which is where the concept and paper was published has created resources to assist others utilise SOS to measure outcomes[8]. Local 20 acute organisations have used SOS to measure for improvement with support from the project lead.

A National SOS Dashboard- A gap was identified between the concept and application, which ICHP & PSMU have addressed with an easily accessible National SOS dashboard launched by Secretary of States for health.

Value

The objectives of the SOS Dashboard are to provide: -National data on SOS and sepsis supporting a truly national conversation -Reliable data to monitor and assess the impact of improvement work on the deteriorating patient and patient safety more generally as SOS causes such a high proportion of emergency admissions and deaths -Cutting-edge visualisation for SOS data at hospital, regional, and national levels.

The value creation is this initiative is multiple fold: The ongoing narrative has demonstrated an emerging national conversation around the measurement of sepsis outcomes using SOS[9] -The diversity and nature of the collaborations which led to the development of the dashboard. An analogy would be that SOS was conceived in Wessex, Oxford derived a ‘mathematical’ formula (SOS) which the teams from Imperial College Health Partners and Patient Safety Measurement Unit have developed into an ‘engineering’ innovation (SOS dashboard).

The diversity of the collaborators, the maturity of SOS concept from theory to problem solving and the new opportunities generated for application to wider areas of enquiry including patient safety, deterioration and mortality review all represent value -Finally, the efficiencies from better reliability in measurement of sepsis improvement interventions represent value e.g. reducing Length of stay.

Involvement

All healthcare professionals and organisations and all local/regional/national policy makers, health economists, educators, researchers, department of health strategists, quality improvers, patients and sepsis enthusiasts.

The development of the dashboard has engaged frontline clinicians in primary, secondary and tertiary care in at least 6 NHS Trusts across multiple regions. There was also national input from NHS England[2,6], NHS Improvement, Patient Safety Collaboratives[7,8], Patient Safety Measurement Unit, CQC, and NHS Resolution via a national workshop[9].

The Intensive Care Society and Office of National Statistics have also contributed into the development of the concept. The dashboard has evolved through various stages of development following input from academics and researchers from UCL, industry and health economists. The Quality Measurement Group of the National Quality Board has also had an input into developing the concept[6]. Patients and sepsis survivors have also had input and understand the simplicity and pragmatism of the SOS approach[7,8] Researchers have understood the value of SOS in defining the at-risk infection population, and are utilising SOS to operationally define sepsis, test biomarkers and ascertain what the high-risk signals are from millions of patient episodes using artificial intelligence and machine learning[4]. The dashboard was short listed for the HSJ Awards 2018.