Learning from hospital deaths

Introduction

High levels of hospital mortality reflect poor healthcare standards. ¹ Some investigations lead to findings of criminality have been crossed. When making his recommendations in the Mid Staffordshire NHS enquiry, Sir Robert Francis QC wrote:

“Where serious harm or death has resulted to a patient as a result of a breach of the fundamental standards, criminal liability should follow and failure to disclose breaches of these standards to the affected patient (or concerned relative) and a regulator should also attract regulatory consequences.” ²

Quantitative assessment

Dr Foster Intelligence publishes individual hospitals’ relative mortality through an annual Hospital Standardised Mortality Ratio which is in-hospital mortality versus expected deaths. The Summary Hospital Mortality Indicator, on the other hand, is published by the Health and Social Care Information Centre and compares the actual number of in-hospital deaths and deaths within 30 days of discharge from hospital with those expected. The Summary Hospital Mortality Indicator counts more mortalities, considers other variables like co-morbidities, and distinguishes between elective and emergency patients,^3–5 but criticisms exist for both systems. First, both extract data from HES-data (Hospital Episode Statistics) and are potentially biased through coding errors. Secondly, a small proportion of deaths is over-estimated and publicised by these statistical tools (when considered alone), to erroneously judge the overall standard of care provided by a hospital. Further, the Summary Hospital Mortality Indicator does not consider the local deprivation factor when calculating expected mortality. This wrongly indicates certain hospitals serving lower socio-economic cohort as outliers or poor performers. “The local authority in England with the highest SMR (Standardised Mortality Ratio) was Manchester (32% above the national level) while South Cambridgeshire had the lowest (26% below the national level).” ^{6 , 7} In spite of these arguments, both these tools generate statistical insight into trusts’ performances. The aim to generate professional curiosity to identify the potential for learning, rather than to “explain away” figures that alert, and indicate the potential for further in-depth qualitative assessment of services. These should be considered in conjunction with other indicators of safety and quality when decisions are being made regarding overall performance of a healthcare organisation.

Qualitative assessment

Prior to 2016, the qualitative assessment of pre-mortem hospital care provided to patients was done through ad hoc retrospective case reviews. In the absence of robust guidelines and standardisation measures, such reviews were subject to inter-rater variability, hindsight and personal bias, and often learning points were missed. Although some standardised tools like Structured Judgement Review, IHI Global Trigger tool and PRISM methodology were available, the process lacked uniformity. ⁴ The review of the Care Quality Commission (CQC) in 2016 highlighted that none of the trusts they contacted could demonstrate a robust mortality review process. ⁸ The Secretary of State for Health reflected on the CQC report and made a range of commitments in his parliamentary speech. ⁹ The National Guidance on Learning from Deaths was published soon after in 2017 and acted as the first official framework to investigate hospital deaths, identify learning points and improve practice through quality improvement process. ¹⁰ This framework states:

“Learning from a review of the care provided to patients who die should be integral to a provider’s clinical governance and quality improvement work. … Trusts should ensure their governance arrangements and processes include, facilitate and give due focus to the review, investigation and reporting of deaths, including those deaths that are determined more likely than not to have resulted from problems in care.”

In the recent Ockenden Report, ¹² which investigated the failures in maternity services at a prominent healthcare organisation in England, the key wishes of patients highlighted were, “questions answered in order that they (patients and relatives) understand what happened during their maternity care. They want the system to learn, so as to ensure that any identified failings from their care are not repeated”. ¹³ This report, yet again, highlights the importance of communication and learning from deaths.

Hospitals learning through their mortalities: a multidisciplinary approach

Learning from deaths involves a multidisciplinary approach; it should include input from senior clinicians, operational team and representatives from governance, coding and risk management, and be overseen by a mortality manager. The team’s function is to reflect on the Hospital Standardised Mortality Ratio and Summary Hospital Mortality Indicator, identify appropriate cases for mortality review, to be distributed to a panel of senior clinical reviewers for in-depth structured judgment review. Qualitative assessment of individual mortalities by independent reviewers not only increases the trusts’ insight into factors contributing to patient harm, but also improves the clinicians’ perspective of safe practice. This exercise directly translates into creating safer and more competent clinicians. Every NHS clinician should thus be involved with the qualitative mortality assessment process to benefit from this exercise, through adequate compensation in their job plans. In addition, the members should meet regularly to discuss mortality reports, address perceived deficiencies in care and recommend areas of improvement, to their peers through representative leads. With board members present, operational issues identified through a mortality review process can be addressed and ratified. Trusts should also publish their annual mortality report for public scrutiny.

Conclusion

Every NHS trust should have a robust mortality assessment structure in place as mortality review is required for all NHS trusts. A robust “learning from deaths” service reflects the trust’s performance and the overall quality of care it aims to provide to its patients.