I doubt there is a unit in the world where at least once a day a discussion ensues about whether an infant is ready to wean or come off their CPAP. For many years we have made the decision based on a variety of markers. Some people would comment on the work of breathing, others on the FiO2 or what the oxygen saturations are at the moment as we round on the patient. Our unit has been pulling oxygen histograms off the patient monitor for years now to provide a more objective measurement to determine if an infant is ready or not. What is a histogram? It is a bar graph representation of the percentage of time in a 24 hour period that an infant has spent in several different oxygen saturation ranges.

A group in Alabama recently published the following paper Oxygen saturation histograms predict nasal continuous positive airway pressure-weaning success in preterm infants. which attempts to answer the question as to whether this practice has merit.

What did they do?

They looked at 36 babies (24 control and 12 cases) in which controls were babies who successfully weaned off CPAP when on less than or equal to 30% oxygen in the first week of life and compared them to infants who failed and had to go back on. Success was defined as remaining off CPAP for 7 consecutive days while failure was having to go back on with in 7 days of discontinuation. All infants were <1250g at birth or less then or equal to 30 weeks gestational age at delivery. Infants were enrolled prospectively in an observational case-control study. During the study goal oxygen saturations were 90-95% and oxygen histograms were monitored q6h by respiratory therapists. Importantly, during the study there was no standard approach to weaning patients off of CPAP but as per many NICUs, discontinuation occurred when FiO2 was low and there were only 1-2 events per day requiring stimulation. The authors controlled for a number of potential factors which could influence success such as GA, BW, Sex, receipt of antenatal steroids, ventilation, caffeine dose, FiO2 prior to weaning and surfactant but found no differences between groups.

What did they find though?

As you might expect there was a difference found and it was in the histograms. The infants who ultimately succeeded in coming off CPAP were better oxygenated in the 24 hours prior to coming off CPAP. Of note, the cases had a median FiO2 of 22% and the controls 21% which was not statistically different.

Looking at the above figure you can see that there were statistically significant differences in the two groups with the babies who successfully weaned off CPAP having significantly higher levels of oxygen saturation in the 95% and above ranges. The authors concluded “The optimal value of oxygen
saturation achievement >95% to predict CPAP-weaning success by Youden index was 31.6% with a sensitivity of 75% and specificity of 75%.”

In other words if you have about 30% of the time spent above 95% in the 24 hours prior to coming off CPAP you have a pretty good chance of success!

Applying the information

Who doesn’t like a study that validates your own practice?! The study is really a beginning though as the study tells us that for babies that are mildly ill (as evidenced by being on room air or 22%) that you can utilize the histogram data to make decisions about when it is best to stop CPAP. What this study though examined is a particular population of small infants who were all taken off CPAP in the first week of life. Would the same principals apply to an older infant or one who is larger at birth? I would like to think so but there are many infants who are on oxygen with BPD who are also weaning off CPAP after many weeks of age. We use histograms in this population as well to guide our weaning but an important measurement that must be taken into account is the FiO2. I can really manipulate a histogram to show anything I want for a baby on oxygen. If it is better from one day to the next is it because the lungs have improved or has the average FiO2 simply been higher in the preceding 24 hours? Conversely if it is worse does the infant have atelectasis or pneumonia or has nursing been more restrictive in FiO2?

Further studies in this area need to create an objective tool that takes into account level of support and mean FiO2 when interpreting the histogram. Failure to do so would lead at times to incorrect decisions if you solely look at a bar graph. As with everything in NICU, the devil is in the details!