Are Frenotomies Worth It?

What does the data say about the increasingly common practice of cutting tongue-ties in infants to improve breastfeeding success? Before getting into it, I want to acknowledge where this starts from — namely, from the fact that breastfeeding is often, especially at the start, difficult. Milk can take time to come in, and many infants (and their moms) struggle to get the hang of a good latch. This can lead to painful nipples and to inefficient feeding, and then low weight gain.

There are many baseline solutions to this: lactation consultants can help, as can pediatricians and doulas. Formula supplementation can also be supportive during this period.

The experience of struggling to nurse can be emotionally fraught — I say this from experience — and it is natural to look for solutions. Which, I think, may explain the growth over time of tongue-tie procedures, called frenotomy. This procedure, in which a small cut is made in the underside of the tongue to allow it to move more freely, is billed as a solution for improving breastfeeding success and lessening nipple pain. Over the past decade, use of the procedure has increased. There isn’t enough systematic data to be clear on how much, but individual practice patterns show an increase in young babies (0 to 2 months) in particular. This certainly does not reflect a change in how infants’ mouths are shaped over this period. It may reflect overuse of the procedure or it could reflect an evidence-based change in our understanding of its value.

Several weeks ago, the New York Times published a long and quite scary article on this topic. The premise of the article is that the procedure is overused by people greedy for profit. It paints a dark picture, including some difficult-to-read stories that seem to reflect an almost violent use of the procedure on babies. The article does not do a detailed review of the data, however, on either benefits or harms.

My goal here is to review both the evidence on benefits and what we know about risk.

Evidence on benefits

There are a number of reasons why doctors might perform a frenotomy. It sometimes comes up in the context of speech issues for older children, for example (we will leave these issues for another time). But the most common reason is for infants who are having difficulties breastfeeding. This might be identified by infants having trouble getting enough milk or in cases when breastfeeding is painful for mom. The procedure involves cutting the tissue that connects the bottom of the tongue to the bottom of the mouth.

(There are many nuances here — tongue-ties can be more or less severe, attached further to the front or to the back of the mouth, and there are also lip-ties. The already limited evidence does not typically distinguish among these.)

Big-picture summary: We have little good evidence on the benefits of frenotomy.

It’s possible to imagine a good study to evaluate this. Such a study would recruit a large sample of infants with breastfeeding issues. They would be randomly allocated to either frenotomy or to general breastfeeding support. They would be followed over time, and outcomes would include continuation of breastfeeding, pain reported by mom, and some objective measurement of breastfeeding efficacy.

There is a recent trial that tried to do something quite close to this, called FROSTTIE. Unfortunately, it ran into two problems. First, it was started right before the COVID-19 pandemic, and ultimately recruitment stalled and the authors were unable to come close to the planned sample size.

Second, and more problematic, a large share of the control group — the group not allocated to the frenotomy — got one anyway: 74% of this group opted for the procedure. In a trial like this, the randomization is not enforced; people are randomized into one group, but they can choose differently. The hope is that enough people stick with their assigned group that you have differences on average. Here, since such a large share of the control group actually did have the procedure, it is hard to learn much from comparing the two groups. The authors didn’t find any evidence of any differences in breastfeeding outcomes, but the limitations of the trial make that hard to draw conclusions from.

The failure of this trial was disappointing, because the data that comes before it is unconvincing. There is a long review article, and a review including the small number of randomized studies. If we focus on the randomized trials, the evidence varies. In three randomized trials in which breastfeeding effectiveness was evaluated by an external observer, two found no differences and one found some small differences.

There are studies that show much more positive results. Generally, these rely on maternal reports without randomization. For example: data from Denmark found that 78% of women indicated a moderate to high improvement in symptoms after the procedure. This type of before-after design, especially with self-reports, is unreliable. Without a control group, we do not know if things would have improved otherwise. Self-reports can be driven by a desire to think things have improved, even if the objective reality is the same.

Within the randomized data, there is more consistent evidence of reductions in maternal pain. This is perhaps not surprising, since it also relies exclusively on self-reports.

All of the existing studies have one fundamental drawback: short follow-up, often only a few days. The goal of this procedure, in a broad sense, is to improve the experience of breastfeeding so people are able to do it longer. We have no reliable evidence either way on whether this is true.

(Sidebar: All of this focuses on breastfeeding success, but if you’re struggling here, it’s worth re-reading this chapter from Cribsheet to be clear that many of the benefits of breastfeeding you hear about are overstated.)

Evidence on harms

In the systematic data reviews, harms from this procedure are limited. A comprehensive look at harms from a large number of studies and case reviews noted that most studies found no significant harm or minimal harms. Minor, limited bleeding was the most common. A 2020 study looked for case reports of serious complications from this procedure from 1965 to 2020 and found 34 infants with serious complications. This is a very small share of the infants who have had this procedure. Most complications involved issues with feeding post-surgery.

This isn’t to say that the procedure has no risks. Any surgical procedure entails some small risk, and we can see from the limited case reports that there are examples of more significant complications (like those highlighted by the New York Times). However, in the vast majority of cases, this procedure doesn’t result in harm.

I think this is important to say because I’ve heard from a number of people basically asking, Did I make a huge mistake doing this? Have I done something terrible to my child? Should I be looking for long-term negative consequences? The answer is no. The procedure may not have helped, but there is no reason to beat yourself up.

Summary

What we have here is a procedure with limited systematic evidence to support it, but also limited evidence of harm, and a lot of people who think it works. Breastfeeding often gets easier over time, or with advice about position and latching, so even if this procedure has no impact at all, it is still likely that many people will find breastfeeding improves afterward.

Put simply: the placebo effect is powerful. We know from countless other settings that people who are treated with sugar pills, or sham procedures, can receive significant benefits just from thinking they are being treated. It may be that this is a big part of what is happening here.

If that is true, we get into a philosophical discussion. Is it ethical to perform this procedure if it improves people’s experiences, even if that is just a placebo? Generally in medicine, the answer to this is no. But it’s slippery. Imagine you thought this actually worked for a small share of people and worked due to a placebo effect for a large share, but you didn’t know who was who. Then it might feel perfectly okay to do the procedure widely.

Of course, we like evidence. The placebo effect is interesting, but I’d like to know if it works for real.

This is proving hard.

The fact that many people — both individuals and providers — feel that this works is, paradoxically, making it more difficult to figure out if it actually works. The main problem in the FROSTTIE trial was that people wanted the procedure so much that they were not willing to stick with their assigned group if they weren’t assigned to get it. This is a huge problem for research design. The authors of that trial end their article discussing it by saying we need alternative approaches to evaluate this, which seems spot-on to me. As usual when we talk about breastfeeding, we do not have enough evidence-based support — whether it’s mastitis, breast milk storage, or this issue — we don’t know enough and we deserve more.

In the end, the New York Times is almost certainly right that we’re doing more of these procedures than are necessary. A frenotomy should not be a first-line option — it should be considered only after a serious evaluation of the infant and of the breastfeeding issues. If it is suggested to you, you might consider a second opinion.

Bottom line

There is limited evidence that frenotomy procedures improve breastfeeding efficacy.
There is slightly more evidence on impacts on maternal pain, but this is entirely based on self-reports.
In systematic data, the harms of the procedure are minimal.
It should not be a first-line treatment for breastfeeding problems.