Futility in cardiology: the need for a change in perspectives: reply

Abstract

We thank Dr Severino and colleagues for their interest and insightful comments on our article ‘Predicting one-year mortality in heart failure using the ‘Surprise Question’: a prospective pilot study’, in which we described the ability of a simple, convenient and intuitive question to predict 1-year mortality in patients hospitalised with heart failure (HF).1 We chose our mode of assessment based upon simplicity, ease of use and the fact that palliative care teams in the United Kingdom are already familiar with the Surprise Question. The Surprise Question has been validated in cancer and end-stage renal failure, it forms part of the Gold Standards .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . Framework decision tool and features in the National Institute of Health and Clinical Excellence guidance on end-of-life care. Simple tools such as the New York Heart Association (NYHA) classification and Killip class have been widely embraced as they are clinically based and simple, whilst providing valuable insights. More complex tools, such as the Seattle Heart Failure Model, whilst having excellent construct validity, are less frequently used in daily clinical practice due to the additional data collection and time needed to complete. The differing trajectories of HF, and relatively unpredictable prognosis mean that prognostication for individual patients is difficult. Despite this, an all-encompassing, single question was able to identify nearly all patients within the last year of life (sensitivity 0.85), whilst also accurately identifying those unlikely to die (negative predictive value 0.88).1 As the respondents state, there was an over-identification of patients into the ‘not surprised’ category (positive predictive value 0.52). Although this could have economic implications, it is unlikely to lead to adverse patient care or experience, since many patients with severe HF might benefit from an early integration of palliative care into their treatment. Crucially, but somewhat unsurprisingly, the Surprise Question outperformed NYHA class and ejection fraction, which lost their fidelity in this sick, hospitalised population. We also showed that the Surprise Question can be used by doctors, nurses, specialists and non-specialists, providing insights for patients increasingly managed by teams of medical and allied healthcare professionals. The correspondents describe the possibilities of using a novel and more detailed HLM classification system, analogous to the widely-used TNM classification in oncology.2,3 Their proposed HLM classification provides .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . additional insights into the complex pathophysiology of the HF syndrome, and it does seem to outperform NYHA class.4 However, we counter the claim that this assessment is truly ‘holistic’ since focusing on physiology cannot account for frailty, performance status, or the ‘end-of-the-bed’ assessment that the Surprise Question inevitably encompasses. Non-cardiovascular causes account for approximately half of all deaths in patients with HF.5 Therefore, as attractive a more detailed assessment of the cardiovascular system seems, it can never encompass the status of today’s multi-morbid patients.