Introduction:
The impact of SpO2 target ranges (TR) has been carefully studied; however, reports suggest a wide variation among infants and centres in maintaining the intended range. Little is known about the effectiveness of different approaches to manual control. Auto-SpO2 controllers are now available which show promise.

Objective:
The aim was to compare two different protocol-driven manual strategies with different response requirements to each other, and a faster automated system (AveaCLiO2, Yorba Linda, CA, USA).

Material and Methods:
In a crossover design, each of the three FiO2/SpO2 approaches was implemented in three randomly assigned consecutive 2.5-hour runs. The two manual strategies (Attentive and Observational) were implemented by a trained operator. The primary endpoints were time in 1) SpO2 TR, 2) <80% SpO2 and 3) >98% SpO2.

Results:
Fifteen studies were completed. All three approaches resulted in good control, with time in the target range >60%. CLiO2 use reflected reduced exposure at the two SpO2 extremes. Post hoc analysis determined that the differences were more marked in the infants with more frequent desaturations. Likewise, in this group, the Attentive strategy performed better than the Observative.

Conclusions:
All three approaches provided excellent control of SpO2 in infants with infrequent desaturations, significantly better than typical routine care. In hard to manage infants with frequent desaturations, faster response appeared to result in better control. The potential of automating the tedious error prone FiO2 adjustment offers significant promise. If manual titration of FiO2 is to remain the usual method of care, additional studies are needed to identify optimal approaches.