The journal Nature reported this week the discovery of three remarkable fossils in Austria. Each belonged to a human baby that died 27,000 years ago and was buried adorned with ochre and ivory. Contrary to the theory that early human beings did not much value infant life, these children were cherished and given ceremonial funerals.

This deep human attachment to babies, even in death, explains much of the emotional response to the Nuffield Council on Bioethics report into neonatal medicine, which was published on the same day. The utter dependence of infants on adults has always aroused protective instincts, and these have made many people uneasy about the recommendation that babies born before 23 weeks should usually be left to die.

Conservative newspapers sought to tap this sentiment with stories of premature babies that did make it, implying that callous experts want such children dead. The council was accused of proposing “blanket” rules that override the clinical judgment of paediatricians and the wishes of parents.

Such coverage, however, has misrepresented the Nuffield position and oversimplified the difficult ethical choices that sometimes have to be made. It is vanishingly rare for infants born before 22 weeks to survive, and 1 per cent of those born before 23 weeks leave hospital. That much is demonstrated by the way several papers used the same case study: Georgina Woods, born at 22 weeks, who is now a healthy five-year-old. There are not many others to be found.

This choice of case was also disingenuous, as Georgina would not automatically have been condemned by the Nuffield guidelines. Only below 22 weeks is it recommended that infants never be treated, unless as part of an approved research study. While normal practice between 22 and 23 weeks would be no treatment, the council makes it plain that parents can request it after being appraised of the risks, and doctors should use their professional discretion. There is no blanket prohibition at all.

What has been recommended is an inclination towards non-treatment when the baby has little prospect of survival, and the gestational ages Nuffield has set are founded on the best available science. The logic is also profoundly humanitarian.

The interventions that occasionally save extremely premature babies cause great pain and generally achieve nothing more than prolonging death. If such infants were routinely treated, a handful might survive, but at an unacceptable cost. For each rare case in which a life is saved, dozens of babies will endure futile suffering.

Like it or not, there are also resource implications. While it would be nice for the NHS to spare no expense saving life, this is impossible. Tough decisions have to be made about cost-effectiveness and, in this instance, treatment is better directed at slightly older premature infants whose chances are much greater. After 23 weeks, 11 per cent survive, and after 25 weeks, 44 per cent live, two thirds of them without serious disability.

If some criticism of this nuanced report reflects discomfort that such dilemmas must be considered at all, there is another agenda in play. Anti-abortion campaigners have long sought to tie the falling gestational age at which a few babies can survive to their demands to cut the 24-week limit for terminations, and many had high hopes for the Nuffield report. It refused, however, to make this link, drawing a sharp line between the legal status of a foetus and a baby that has been delivered. Disappointed activists are talking down the Nuffield conclusions, and talking up the viability of infants born at ages when terminations are occasionally still performed. The upper limit for abortion may be worthy of debate but that issue should not hijack this balanced consideration of a different ethical matter.

Mark Henderson is Science Editor of The Times