We thank Donaldson [1] for his tremendous interest in our discussion of time-limited trial (TLT) of intensive care [2, 3]. We are grateful that Donaldson concurred with us that “the clinical uncertainty inherent in ICU triage will, in many contexts, make exact prognostication impossible” [1].

While Donaldson questioned whether the patient would be able to provide an adequately informed consent in the context of uncertain survival benefit, we would like to reiterate our stance on TLT: it acknowledges the inherent clinical uncertainties faced by the patient and medical team and intends to prevent invasive interventions in intensive care unit (ICU) from being extended until the appropriateness of such treatment can be more accurately assessed [2]. It involves shared decision-making, requiring the clinicians to communicate with the patient and family, identifying their values, and weighing them against the possible harm associated with invasive organ support [3]. Although we agree with Donaldson that the predicted chance of survival and the proportionality of treatment burden are vital considerations in determining ICU admission, in reality, it is sometimes difficult to accurately predict patient’s outcome based on the initial encounter. Therefore, as advocates of TLT, we intend to empower the patient and family in reaching a mutual consensus with the medical team over what would be regarded as the most reasonable management plan in face of the unpredictable clinical course. This does not imply that we would admit patients to the ICU for TLT despite clear signs of medical futility [3].

We believe that survival should not be taken as “success” automatically; for the same reason, one should not insist that death during a TLT “must be considered a failure” [1]. Although death is not a desirable outcome of life-sustaining treatment, it does not necessarily preclude the successful fulfilment of a dignified death. If it is the patient’s will to receive a trial of intensive-care treatment before a potential final transition to end-of-life care, TLT respects patient’s autonomy that constitutes a fundamental element of perceived dignity. Donaldson’s philosophical inquiry into the goals of clinical medicine has reminded us of the importance of humanistic values in caring for patients at risk of imminent death, in which the moral character of compassion and discernment in the virtue ethics tradition [4] are highly relevant. Offering TLT to patients who opt for proportional (albeit invasive) treatment in the context of a shared agreement is coherent with the ethical intuition of a compassionate doctor.

Moreover, due to the recent surge in coronavirus disease 2019 (COVID-19) infections in our city, ICU physicians have been bombarded with difficult triage decisions for elderly patients with severe COVID-19 pneumonia and multiple comorbidities. We consider that establishing TLT agreements at the time of ICU triage may serve as an alternative to the dichotomous process of admission with unrestricted life-sustaining treatments versus non-admission. While rationing consideration is beyond the scope of an individualistic assessment of patient’s best interests that formed the basis of our discussions in ethical reasoning for TLT, we believe that it is also important to consider the society’s collective interests in optimising utilisation of ICU beds and safeguarding scarce resources in the contingency capacity. Acknowledging the complexity of the decision-making process, a discerning doctor who tries to make both sensitive and sensible judgements would consider thoroughly whether TLT of intensive care is the best option, after weighing the pros and cons from different parties’ perspectives delicately to strike a balance.

Finally, we recognise that the provision of life-sustaining treatment and transition to end-of-life care are greatly influenced by cultural and societal contexts. In a locality where the general public is less ready to accept palliative care due to previously established social norms and structural limitations of the healthcare system [5], TLT may act as a bridge to palliative and end-of-life care in an acute care setting when the expected prognosis is poor (e.g., a young patient with motor neurone disease who developed severe pneumonia) but not immediately obvious to the patients or their surrogates, who may hold unrealistically high expectations of survival with intensive-care treatment despite thorough explanation. TLT respects their autonomy, allowing time for them to comprehend the clinical reality of dying and decide on their preferred care pathway. The fact that different societies may have different popular perceptions and acceptance of intensive-care treatment should be acknowledged. Though controversial, we think that TLT may serve as a sensible compromise in this regard.