Annie Janvier’s comment on my recent post made we wonder what are the ethical limits of pain research in the newborn.

Let me start with an adult analogy, if a researcher wanted to investigate the efficacy of a new analgesic agent, say for a fairly serious pain, such as the pain of a broken limb in the emergency room, would it be ethically acceptable to compare the analgesic to placebo? We know that placebos, especially elaborate magical placebos such as homeopathy and acupuncture, do have modest effects against pain. But I posit that for a serious painful event, comparison with placebo would be unacceptable.

Even though analgesia has not been proven to improve long term outcomes, reduce brain injury, or increase survival. Generally speaking in order to be ethical, research should compare to the current optimal therapy, or if there is no intervention generally considered the current optimal, to one of the acceptable adequate alternatives. When we are talking about analgesia, an appropriate comparison analgesic is essential.

And yet I still see articles about heelstick which have an untreated, or placebo group. To follow the adult analogy above, there is no way this should be acceptable, especially for children who cannot agree (and should not be asked to agree) to experience some pain for the good of science. One example is this very recent publication (Asmerom Y, Slater L, Boskovic DS, et al : Oral Sucrose for Heel Lance Increases Adenosine Triphosphate Use and Oxidative Stress in Preterm Neonates. The Journal of pediatrics 2013, 163(1):29-35.e21.)  I did not blog on this at the time, and I am still not going to discuss it in detail, because I am disturbed that anyone would randomize newborn babies to have more pain. The fact that the babies who received sucrose also had a very small increase in their hypoxanthine levels, compared to controls who had a slight decrease, in my mind is unimportant. Pain hurts, and if you can interrupt it, which we can, with sucrose, then studies looking at mechanisms and potential adverse effects should compare only to controls being actively treated with other interventions.

In this instance, breast milk is effective as an intervention, as is kangaroo care, neither of which are going to be withheld from preterm babies, even if they increase hypoxanthine concentrations! Either of them could have been used in this study. If you want to investigate effects of sucrose on some outcome as an analgesic in the newborn, it is unethical to do so with untreated controls. Use of breast milk, or kangaroo care would be an ethically acceptable comparison.