I knew from the title of this article that we wouldn't have an answer to the question of tonsillectomy for recurrent throat infections. Read on.
Source: Morad A, Sathe NA, Francis DO, et al. Tonsillectomy versus watchful waiting for recurrent throat infection: a systematic review. Pediatrics. 2017; 139:e20163490. doi:10.1542/peds.2016-3490. See AAP Grand Rounds commentary by Dr. Carrie Phillipi (subscription required).
The clue in the title is that this studied is labelled a systematic review, without mentioning meta-analysis. Meta-analysis is the process of applying statistical tests to studies collected in a systemic review to provide a combined summary statistic of the benefits (or lack thereof) of a particular intervention. The subject of tonsillectomy versus watchful waiting should lend itself well to a meta-analysis, but in this case the collected studies were too heterogeneous to be combined for analysis. In short, even though the studies were looking at approximately the same problems and interventions, they were too different to combine in one analysis: you can't really compare apples and oranges.
Still, the study bears at least a glance because it highlights the problems inherent in any systematic review of the literature, plus readers can learn about the Agency for Healthcare Research and Quality's comparative effectiveness review of tonsillectomy in children.
The systematic review researchers screened 9608 literature citations on the subject and ended up with only 7 studies that met the investigators' predetermined criteria for inclusion (which included randomized controlled trials as well as prospective and retrospective cohort studies) and low to moderate risk of bias. Furthermore, they felt the 7 studies were too heterogeneous in how outcomes were reported to combine in a single meta-analysis. What they could report from these studies, however, was a suggestion in the tonsillectomy group of less throat infection episodes, healthcare visits, and school episodes over a 12-month period, but lack of evidence showing persistence of any benefits over longer time periods. There were no demonstrated differences in quality of life analyses in the studies that looked at this outcome. In part, the lack of long-term benefit demonstration may be due to very high drop-out rates of study participants over time. Also, the researchers felt that all of their conclusions had relatively low degrees of strength of evidence, with only a moderate strength of evidence rating for 1 conclusion of less sore throats/throat infections in the less than 12 month follow up period.
As the investigators point out, not all sore throats are created equal. How many of these patients had recurrent streptococcal pharyngitis, which is in itself difficult to diagnose due to high rates of asymptomatic carriage of group A streptococcus in children? How many with recurrent sore throat had the PFAPA (periodic fever, aphthous stomatitis, pharyngitis, cervical adenitis) syndrome, for which tonsillectomy may very well be effective? These are just a few of the "apples and oranges" that complicate study interpretation.
So, what do we do now? First, I tried to estimate a number needed to treat from 1 of the randomized controlled studies included in the review, and came up with the tonsillectomy group having 1.68 fewer episodes of sore throat in the 12 month period following the procedure. The NNT for this number is 60, meaning that 60 children with recurrent throat infections would need to undergo tonsillectomy for 1 additional child to achieve that lower sore throat rate. That's not a very good yield in my book, given that tonsillectomy is not an entirely benign procedure, but others may differ on my assessment.
I liked the authors' wording for how to manage until better information is available, so I'll repeat it here: "...individual decision-making needs to balance the benefits of reducing illness-related outcomes (including missing school and work) with the risks associated with surgery. Caregivers and providers may wish to consider the potential benefits and drawbacks of attempting to manage children's illnesses for a period of time to see if they outgrow the propensity for infection to avoid surgery." Also, clinicians should note that at least 1 well-respected practice guideline has recommended against tonsillectomy solely to reduce the frequency of streptococcal pharyngitis.
