At the start of this study, we first discussed identifying risk factors and ranking them in terms of risk. These comprised patient factors (e.g. advanced age) and surgical factors (e.g. urgency of surgery). However, it became apparent that a multifactorial approach was more accurate. For example, one might think that advanced age would be an easily identifiable risk factor in its entirety, yet there are 90 year olds who run marathons.
In our study, we identified 6 risk factors that increased the risk of mortality within 30 days of inpatient surgery. The more of these a patient has, the more the risk increases.
Our work included analyses of 10 comorbidities as risk factors. These were selected by a multi-disciplinary Expert Group of the NCEPOD "Knowing the Risk" study and considered to be the most important factors at that time. They are arrhythmia, cancer, documented cirrhosis, congestive cardiac failure, current smoker, diabetes (insulin), diabetes (non insulin), ischaemic heart disease, respiratory disease, and prior TIA/stroke. There are studies that have analysed more risk factors (as described in our paper) but we wanted to adopt a pragmatic approach by producing a risk calculator that was simple and easy to use by clinicians.
There have also been suggestions of quantifying what it means to be "low", "medium" and "high" risk. A group who advised on the Royal College of Surgeons' report "The Higher Risk General Surgical Patient", reached a consensus that high risk patients had a predicted hospital mortality risk of ≥5% and as a minimum, should be admitted to critical care if their estimated mortality risk was ≥10%. Clinical experience suggests that a mortality risk of 1% or lower is "low" risk. Several of my colleagues believe medium risk should be 2-5% with high risk as having a mortality risk greater than 5%.
What are the clinical implications of categorising risk? In the UK, we might not have the resources to request that all high risk patients (greater than 5% mortality risk) be admitted to higher dependency care before and/or after their surgery. However, even if bands of risk are devised, decisions regarding admission to critical care or delay of surgery until the patient is more fit for surgery, should still be made in conjunction with clinical judgement and in the context of organisational facilities and staffing within each hospital, as well as the patient.
Patients may think of their risk differently compared to the clinical team. For example, the severity of their symptoms and any comorbidities they may have, together with the potential improvement in quality of life after their surgery, and life expectancy, may influence whether or not they think the risk is worthwhile. Some patients may find risk banding more easily comprehensible than a percentage mortality calculation. This may in turn facilitate the shared decision making process with their healthcare professionals.
Karen Protopapa
Researcher at NCEPOD
April 2015
References and further information on the above are available in our paper.