There are five sick babies and only four incubators in the neonatal intensive care unit. What do I do? The infants who get the incubators and the ventilation equipment may survive because of it; the child who does not will probably die because of the omission. What would you do? Five tiny hearts flutter desperately inside fragile ribcages; four Perspex boxes offer temperature-controlled, oxygen-supplied environments. How I can decide who gets what? The dilemma has become familiar, but no easier, during recent weeks here in Quetta’s paediatric hospital.

In Britain, discussions about withholding or withdrawing life-sustaining medical treatment in newborns have focused upon those individuals with such extreme pathological conditions that ceasing or not even starting life-sustaining medical intervention might be considered ethically appropriate. The UK’s Royal College of Paediatrics and Child Health has produced a framework for ethical practice which includes examples of ethical non-treatments that include: non-resuscitation of a baby with severe congenital abnormalities incompatible with life; and withdrawal of ventilation from a child proven to have severe birth asphyxia-related brain damage.

But none of the five neonates needing incubators here in Quetta have such severe pathology: they each might have the potential to survive. The western ethical dialogue about withholding treatment in extremely unwell neonates does not aid my decision-making here. The issue in this setting is one of resource allocation. And my recourse is to the moral commitments that I recall vaguely from medical school and revise from an old, but still insightful, BMJ article: respect for autonomy, beneficence, non-maleficence, and justice. Of which the most pertinent here appears to be justice.

But what would justice look like in this situation? What would constitute a fair distribution of resources among the five newborns? Should I assess the infants, select the sickest, and place them in the four incubators? Or take something like a utilitarian consequentialist approach and select the four infants with the greatest prospect of survival? Or draw straws?

Investing treatment efforts in those infants with greatest needs will mean scarce resources are likely to be expended upon children who will subsequently die. And meanwhile the incubator is unavailable for the treatment of others who might have better chances. On the other hand, selecting the children with the best prospects of survival means denying those most in need; thereby sustaining the inverse care law (which states that availability of health care varies inversely according to need) and seeming to contradict the very principle of an intensive care unit. And ultimately, I guess, if we truly wished to do the most good for the greatest number of infants, we might question the role and morality of the neonatal intensive care unit (NICU) in developing countries: as has been pointed out, adequate maternal nutrition, antenatal care, basic childcare and breastfeeding would “undoubtedly have a greater impact on neonatal infant mortality” and at far less cost.

Notably, published accounts that describe western charitable organisations setting up and operating neonatal intensive care units in the developing world have emphasised practical, technical aspects. Ethical considerations are mentioned only inadequately, or not at all, or only to dismiss them, as one observer suggested: “in developing countries with few resources, decision making for vulnerable neonates is relatively simple, as the options are limited.”

The gross misunderstanding inherent in such statements was made apparent by authors here in Pakistan over two decades ago:

“Limited resources, widespread poverty, and the absence of health insurance pose daily ethical problems for Third World physicians, who must balance their roles as individual patient advocates against a desire to provide healthcare to the greatest number of children. Whereas a physician in a developed nation may struggle over whether to resuscitate a malformed infant or a Trisomy 13 infant in the delivery room, his counterpart in Pakistan must agonize over which, among two otherwise healthy neonates with mild IRDS [infant respiratory distress syndrome], should benefit from the last bed in the NICU”

Which returns me to the question of the five sick neonates and the four incubators here in Quetta. As the above authors state: “whereas neonatal technology has grown rapidly in developing countries, means of addressing these thorny issues are in their infancy.” In 2004, one of those authors, Farhat Moazam, was instrumental in establishing Karachi’s Centre for Biomedical Ethics and Culture (CBEC)—the first such centre in Pakistan. Doubts that I might have about the Centre’s rejection of universal ethics are to some extent dispelled by the clear inadequacy of my own ethical ponderings in this setting. The CBEC’s efforts to “indigenise” bioethics in Pakistan in a manner that is “relevant and responsive to local cultural and social realities” may well offer a more appropriate basis for decisions about establishing neonatal intensive care units and the decision-making that occurs within these NICUs.

Andrew Moscrop qualified as a GP last year and will be working with MSF in Pakistan for six months.