The NICU

Posted: 2023-09-13

My son was born about a month and a half early, and had a monthlong NICU stay. The NICU is a very stressful place for parents, even though everyone who works there is compassionate. We felt very well cared for, and were relieved when we left to get to the “normal” newborn time of eating, playing, and (occasionally) sleeping. There will be lots more time with the medical system in our future, but I wanted to write down some relatively fresh thoughts on the NICU.

Empathy

At least for our son, the most urgent concern was difficulty breathing immediately after birth, and it was addressed in a way that felt very familiar. There was a short list of increasingly invasive interventions to try, and when we reached the least invasive one that worked, the problem was resolved and we stopped.

That was a relatively short percentage of the time we were in the NICU. Most of my son’s NICU stay was spent waiting for him to grow out of risky behaviors.

It was very hard. Occasionally my son would just have a bad night (he’d stop breathing or have risky heartbeat patterns) and while a nurse could intervene in the moment, there was nothing anyone could do to help him avoid it the next day. Each time it happened, we just had to wait a little longer. For what it’s worth, while this was scary for my wife and me, my son probably did not notice or care! From his perspective people would occasionally come cuddle him at all hours.

The nurses and child life specialists were amazing for us during that phase of our son’s NICU stay. They helped us deal with uncertainty, handle the stretches when there was nothing to do, and understand what would be different when we went home. We had a lot of attention from some incredible people.

Medicine for infants

The medical team (MDs, PAs, NPs) was also empathetic and compassionate. The ratio of babies to providers is much higher than that of babies to nurses, so it is not really possible for them to give each patient the same amount of attention as the nurses do.

The medical team in our NICU consisted of a resident, a fellow, and an attending physician. All three are doctors, in increasing order of experience. As best I could tell, each day the unit was staffed with one attending, maybe one to two fellows, and three to five residents.¹ All providers rotated, and we had continuity with a single team for roughly a week at a time, except on weekends.

The medical team makes a plan for the patients, but NICU babies can’t really tell you how they’re doing! The resident does a physical exam of their babies once per day, probably no more than a minute or two, which is occasionally checked by the attending, also for no more than two minutes. The doctors discuss a plan once per day, unless something extreme happens.

Since time with the patient is limited, a lot of the plan is based on the outcome of a test or a radiology read. When something “seems off,” more tests may be ordered based on the philosophy of that day’s attending. For example, one night our son had been fussy and his blood oxygen saturation had been fine, but slightly lower than usual. The attending ordered an X-ray to check if his breathing issues were getting worse again. The lungs looked fine on the X-ray, but they saw gas in his gut. We and the nurses knew that - he was gassy that day! But because gas in the gut can be a problem (an obstruction), the X-ray read triggered a fellow to come check on him again several hours later. This sort of incidental followup, including the multi-hour delay,² was very common while our son was in the NICU.

In fact, that same day he was also a little snotty from spit up in his nose. After rounds, we sucked out some boogies, and his blood oxygen saturations went back up to where they normally were.³ We don’t yet know if any of the “odd” test results from that day are meaningful or not: no baby at home would have gotten any tests at all.

Neonatologists have a hard job. The patients can’t talk and the parents want to know everything, so it seems like a medical specialty that is very susceptible to over-testing or map/territory mismatches. For example, our son received three newborn screening blood tests. Generally, babies need only one, but if the first is taken under 24 hours after birth they sometimes need two. No baby really ever needs three, but for largely administrative reasons they did an additional one just before we left the hospital. Even though the first two had no findings, if anything comes back unusual on the third one, we’ll need more medical followup.

I’m not sure what the p-values are on these tests, but it seems like the medical system does not take a Bayesian approach⁴ to test results. Our son got ultrasounds and MRIs, blood electrolyte panels, complete blood counts, blood gas checks, and more. There were very few days in the NICU that he did not get tested in some way. Very few of those tests changed his immediate-term treatment plan. The most worrisome things we had to take home with us were incidental findings or ambiguous test results.

It was clear that all our doctors felt this was the best and safest way to take care of our son. They wanted us to leave the NICU as quickly as safely possible, with useful information to help our son grow up as healthy as possible. They wdanted to take care of us as parents as well as our son as their patient. Nevertheless the approach has meant a lot of followup and repeat tests, often without any treatment, even after we went home. That means a lot of time spent in the medical system.

Practicality

On a more lighthearted note, there are some practical things about how the NICU works that were, if not fun, at least interesting to a systems person like me.

Because of the map/territory risks, the maps are pretty cool. In particular the heart rate/respiratory rate/blood oxygen saturation monitor had a cool visualization where the PDF/histogram of the past 12 hours of readings was laid over the CDF. It’s a rare effective use of separate left and right Y-axes. It also had multiple ways of displaying the raw inputs and calculated rates to time-correlate the three measurements.

Unfortunately, respiratory rate monitoring for infants is not very reliable. It measures chest expansion, which works ok if the infant is asleep and still. That’s not very often. Any time the baby moves the monitor gets confused, and the calculated respiratory rates are way off. This would be fine, except that the rates are important for correlating apnea with decreased heart rate and blood oxygen saturation levels!

Finally, the supply cabinets in the hospital are locked with a code. All the ones I saw had the same combo I use on my luggage. I have heard from friends in medical fields that occasionally JCAHO will come check that equipment is secured correctly, and probably on those days everyone flails around trying to remember a temporarily secure code.

Home

We’re very lucky to have had high quality healthcare. The hard parts we have left are stress, followup, and potential future medical issues, and it could certainly be worse. For now, we’re just enjoying spending time with our son!

--Chris