Hylke Salverda

9 The first sign of the drawbacks of administering such a high fraction of oxygen in the air infants breathed appeared in 1940, when paediatrician Clifford noticed a new eye condition, later called retrolental fibroplasia. This new condition was meticulously studied by ophthalmologist Terry in the following years and another eleven years of research were needed to link this blindness-causing disease to administering oxygen. From then on, physicians began to realize that too much oxygen was harmful, and a more restrictive approach followed in the mid-1950s. Physicians lowered the oxygen content in supplied breathing air to 40%, and as a result the rate of retinopathy of prematurity - the contemporary name for retrolental fibroplasia - decreased. This change in practice was not based on evidence from research, but on clinical findings from individual paediatricians. Lowering the oxygen content also had a problem: both the rate of hyaline membrane disease, now known as respiratory distress syndrome, and the rate of cerebral palsy went up for preterm infants - keeping the balance between too much and too little was, and still is, difficult. Although a form of a pulse oximeter - a device to measure the oxygen saturation of the blood – was developed in 1935, titrating oxygen on the basis of the oxygen content in the blood only started in the 1960s, when blood gas monitoring became readily available. It would not be pulse oximetry as used today, as this was only developed by researcher Takuo Aoyagi in 1974. These early pulse oximeters were highly inaccurate when patients moved. This poses a particular problem for preterm infants, who cannot be instructed to lie still. Eventually, in 1995, Masimo developed Signal Extraction Technology, which is more resistant to motion, and this is now the basis for guiding how much oxygen to give preterm infants. Technology has become more and more sophisticated. In our unit, infants in need of respiratory support can receive breath volumes as low as 2 millilitres. The breath is given at the exact moment the baby attempts to breathe, and the amount of oxygen is automatically adapted to the infants’ need by measuring the oxygen saturation of the blood. By continuous automatic adjustment of the oxygen content we give, we are better than ever at keeping the fine balance between too much and too little. Despite these improvements many infants still suffer from the complications of prematurity every year. With this thesis, I hope to contribute to a better life for these infants by researching how the devices for automated oxygen control work, how well they work to balance the oxygen saturation in the blood and, most importantly, what the health outcomes are of the preterm infants treated with these devices.