The letter responds directly to a recent study by Kirkham and colleagues, published in the same journal, that followed children with mild sleep-disordered breathing who were managed with watchful waiting rather than immediate surgery. That study identified prognostic factors associated with progression of the condition, and the Turkish authors contend that such findings are at risk of being misinterpreted. If a child carries a feature that predicts worsening over time, the intuitive leap is that removing the obstructing tonsillar and adenoid tissue early would prevent that worsening, and that high-risk children therefore stand to gain the most from adenotonsillectomy. Yılmaz and his colleagues argue that this leap is logically unsound, and they ground their objection in a well-established framework from clinical epidemiology: the distinction between prognostic and predictive markers.

The distinction was crystallized in a widely cited 2015 paper by Kristel Ballman in the Journal of Clinical Oncology, which the letter authors invoke. A prognostic marker tells you something about the likely course of a disease or condition regardless of which treatment the patient receives. A predictive marker, by contrast, tells you something about the differential effect of a specific treatment, identifying patients in whom the therapy works better or worse than it does in others. A tumor biomarker may signal aggressive disease in every patient who carries it, yet only randomized evidence can show whether that same biomarker identifies patients who respond to a particular drug. The two concepts answer different questions, and conflating them can lead to treatment decisions that are not supported by evidence.

Applied to pediatric sleep medicine, the logic becomes clear. If a child with mild sleep-disordered breathing has characteristics associated with a higher likelihood of progression, those characteristics are prognostic. They suggest the child’s symptoms may worsen whether or not surgery is performed. They do not demonstrate that surgery will produce a larger improvement in that child than in a low-risk child. Only a predictive analysis, ideally within a randomized trial that formally tests for an interaction between the risk factor and treatment assignment, can establish differential treatment benefit. Without such an interaction analysis, claiming that high-risk children derive greater benefit from adenotonsillectomy extrapolates beyond what the data can support.

This is not merely a statistical quibble, because the strongest available evidence on the question comes from a landmark randomized clinical trial. In 2023, Redline and colleagues published in JAMA the results of a randomized trial of adenotonsillectomy for snoring and mild sleep apnea in children, a study that grew out of the Pediatric Adenotonsillectomy Trial for Snoring, or PATS, protocol described by Wang and colleagues in BMJ Open in 2020. That trial randomized children with mild sleep-disordered breathing to surgery or to watchful waiting and evaluated outcomes including behavioral measures and symptom resolution. The letter authors point to this randomized evidence as the appropriate standard against which claims of differential benefit must be tested, rather than observational progression data from cohorts of children managed without surgery.

The stakes of the distinction are considerable. Adenotonsillectomy, while generally safe, is a real surgical intervention carried out under general anesthesia, with associated risks of bleeding, infection, pain and the considerable burden that surgery imposes on children and families. If clinicians begin triaging children toward surgery on the basis of prognostic risk alone, families may be counseled toward an operation on the mistaken understanding that their child is precisely the kind of patient who benefits most. Conversely, if genuine predictive markers of surgical benefit do exist, identifying them properly could spare low-benefit children an unnecessary operation while directing those who would truly gain toward timely treatment. Either way, the answer determines how thousands of pediatric consultations unfold each year.

The watchful waiting study that prompted the letter adds an important empirical dimension to the debate. By demonstrating that many children with mild sleep-disordered breathing who are managed without immediate surgery do not progress, and by characterizing which children are more likely to worsen, it provides clinicians with genuinely useful information about the natural history of the condition. Natural history data of this kind are exactly what prognostic research is designed to deliver. The letter authors do not dispute the value of that data; what they dispute is the inferential step of converting prognosis into treatment guidance. Knowing that a subset of children tends to worsen over time does not reveal whether early surgery would have altered that trajectory, because the children in the watchful waiting cohort were, by design, not operated on.

There is also a broader lesson here for how medical evidence travels. Findings from observational cohorts are frequently repurposed, in review articles, guidelines and clinical conversations, as justification for interventions they were never designed to evaluate. The prognostic-predictive framework offers a simple safeguard: before using a risk marker to justify treatment, ask whether the marker was shown to modify the effect of that treatment in a randomized comparison. In oncology, where the framework was formalized, this discipline has reshaped how biomarkers are validated and how targeted therapies are prescribed. Sleep medicine, the letter suggests, should adopt the same rigor, particularly for a condition as common and as variably managed as mild pediatric sleep-disordered breathing.

For now, the practical message for clinicians and families is one of calibrated caution. Mild sleep-disordered breathing in children sits on a spectrum that ranges from benign primary snoring to frank obstructive sleep apnea, and the decision to operate should rest on the best randomized evidence available, on the severity and impact of symptoms, and on shared decision making that weighs the known benefits and burdens of surgery. A child’s risk of progression is relevant information, but it is information about the disease, not about the treatment. As Yılmaz, Özer and Topal emphasize, until studies demonstrate that specific risk factors predict differential benefit from adenotonsillectomy, prognostic risk alone cannot establish that high-risk children have more to gain from the operation. The distinction may sound like semantics, but in the clinic it is the difference between evidence-based surgery and surgery based on an inferential shortcut.

Subject of Research: The distinction between prognostic and predictive risk markers in deciding adenotonsillectomy benefit for children with mild sleep-disordered breathing

Article Title: Prognostic risk does not establish greater adenotonsillectomy benefit in children with mild sleep-disordered breathing

Article References: Yılmaz, E., Özer, N. K., & Topal, E. (2026). Prognostic risk does not establish greater adenotonsillectomy benefit in children with mild sleep-disordered breathing. Journal of Clinical Sleep Medicine, 22(1), Article 176. https://doi.org/10.1007/s44470-026-00196-3

Image Credits: AI Generated

DOI: 10.1007/s44470-026-00196-3

Keywords: adenotonsillectomy, pediatric sleep apnea, sleep-disordered breathing, prognostic markers, predictive markers, watchful waiting, randomized controlled trial, clinical decision making, snoring, biomarkers, evidence-based medicine, Journal of Clinical Sleep Medicine