Considerations for Interpreting Childhood Obesity Treatment Trials from the COVID-19 Pandemic Era

Calibration Against Contemporaneous Observational/Surveillance Data

Researchers may report comparison data from sources, including medical records or longitudinal cohorts, which would be particularly useful for contextualizing outcomes from single-arm trials. These data may be difficult to obtain; researchers can partner with local medical centers to use medical records or use data sources such as Oracle/Cerner or the National Health and Nutrition Examination Survey for comparison.

Use of Historical Controls

Studies that were ongoing before and during the pandemic can present results from these two groups. Subanalyses examining group differences based on pandemic timing may be quantitatively and/or qualitatively described to discuss pandemic impacts. Using this method, one study observed children who completed treatment during the pandemic increased their BMI z-score (BMIz), whereas children who completed their treatment before the pandemic decreased their BMIz. ³

Comparison With Untreated Control Groups

In randomized trials, treatment and control groups should exhibit similar effects of the pandemic. One study observed that children in the control group increased their BMIz to a larger extent compared with children in the treatment group 1 year postintervention. ⁶ Studies using treated control groups may find useful comparisons in observation/surveillance data described earlier.

Although no approach is perfect, these methods can support the interpretability of COVID-era study findings. Moving forward, articles should orient the study and results within the pandemic timeline ⁴ and report secondary/sensitivity analyses explicitly modeling pandemic-related effects. Of note, it cannot be assumed that general population weight change is a proxy for what should/could have been expected for weight change for treatment-engaged children.

To address issues of pandemic impacts on outcomes, we recommend (1) an expert committee devises guidelines to assess treatment outcomes from the pandemic; (2) the research community recognizes that trials conducted during the pandemic may have clinically meaningful results even without meeting previously accepted metrics (i.e., reducing BMIz by 0.2–0.25) ⁷ ; (3) dissemination/publication of “null” results. Finally, researchers should upload data into publicly available repositories with detailed timelines to evaluate the impact of the pandemic across studies. We believe these steps will move the field of pediatric obesity treatment toward meaningful change despite the setbacks of the COVID-19 pandemic.

Impact Statement

There is limited guidance on how to interpret results from pediatric obesity treatments during COVID-19 and early studies indicate attenuated response in treatments from this time. This editorial provides considerations regarding the interpretation of results and recommends guidance from experts surrounding treatment analyses and interpretation.