It is a surgical aphorism that it is more difficult to decide when not to operate than when to operate.
In the study “Learning from Regret”1, surgeons contributing to Queensland’s Audit of Surgical Mortality were asked to reflect on deaths following surgery. They were asked whether they would have done anything differently in retrospect. The aim was to explore surgeon’s reflections on what might have been done differently for patients who died under their care. The secondary aim was to assess for the presence of regret – defined as the presence of personal responsibility and realisation that another decision could have been better.
Post-decision regret was explicitly identified in 16.9% of responses. One of the scenarios specifically referred to was regret surrounding with proceeding to surgery in cases where the likelihood of survival was slim (“the decision entailed the alternative of certain death rather than probable death”). The fields of behavioural economics and psychology give us several reasons why we shouldn’t reflexively regret decision-making in these types of charged scenarios – because regret is the de facto outcome of a decision to operate in this scenario.
Firstly, the fact of operating rather than palliating in this scenario makes us more likely to regret it in retrospect when the outcome is poor by the mere fact of one having acted rather than not acted. Actions are more salient (i.e. easier to recall) than inaction and, therefore, more prone to regret upon reflection from a psychological perspective2. Humans also tend to regret action (acts of commission) more than inaction (acts of omission) because having acted, we then attempt to mitigate the cognitive dissonance that arises between our ideal and actual selves by reducing the dissonance by deriving lessons learned from the perceived error3.
The authors of “Learning from Regret”1 highlighted that surgical decision-making in cases associated with deaths were uncertain, complex and subject to situational pressures. In not operating, the surgeon would have to violate the norms of usual surgical behaviour4 that in the face of certain death without an operation, an operation should be performed.
In fact, in scenarios where treatment is highly likely to be futile, expected utility theory would tell us that our choices are very limited. In classical economics, expected utility theory is the theory of how rational actors make decisions. That is, that rational beings choose the option to maximise the risk: benefit ratio. When applying expected utility theory to medical decision-making, as the net treatment benefits increase, we become more uncomfortable withholding treatment even when the probability of a good outcome is relatively low and the risks are substantial. Therefore, when the counterfactual treatment outcome is certain death, if we apply traditional rational economic decision-making theory, then there are few circumstances in which the certainty of this poor outcome can outweigh the small possibility of a good outcome (especially where survival and not functional outcome is the primary outcome).
Furthermore, we increasingly recognise that medical decision-making employs dual-processing theory5. Rather than employing the neural pathways utilised in rational decision analysis, medical decision-making also draws from type 1 or emotional decision-making. This tends to make medical decision makers more risk-averse. Dual-processing of this nature also means that it is more challenging to apply rational thinking behaviours to decision involving how we must act, than it is to rationally decide how others should act6.
Clearly, aiming to avoid futile treatment that increases suffering at the end of life is an important goal in emergency surgery, however, making these types of decisions will always be difficult because surgeons are human and prone to all of the cognitive biases inherent to this state.