Pediatric Cancer in the Northeast United States—Reply

Abstract

Our article (1) set out to analyze data relating to childhood cancer following Dr. Siegel's Centers for Disease Control and Prevention (CDC) descriptive study in 2018 (2) in which a headline result was that “rates were highest in New Hampshire.” Messmer raises several questions, which seem to distill to: (i) was our conclusion appropriate, that New Hampshire's (NH) incidence of childhood cancer was not statistically significantly higher than the rest of the Northeast; (ii) did we draw conclusions beyond the scope of the analyses; and (iii) how should resources be directed in the goal of preventing childhood cancer?(i) When all races are analyzed together, NH has the highest incidence of all states, but as we reported, it is not a statistical outlier. However, incidence varied across race/ethnicity groups by more than 38% and was highest in White children (Table 1; ref. 1), making race/ethnicity a potentially important confounder. To avoid the criticism that differences in incidence were simply due to differences in population composition, stratification by race/ethnicity was needed. Both overall and stratified measures are potentially useful, depending on the goal. If we want to discuss the burden of childhood cancer, then overall incidence should be the focus. If we want to make unbiased comparisons as a first step in the journey toward causal inferences, stratifying by major confounders is the methodologically correct approach. We note that NH's age-standardized rates do not rank highest in any race/ethnicity stratum, and this is not dependent on the use of Bonferroni (Table 1; ref. 1). Did the high incidence in NH bias our regional comparisons? NH accounted for only 2.5% of cases in the Northeast, and 2.4% of the population, so this seems unlikely. The Tiwari method of modification that we used means that the comparison we reported is between NH and the rest of the Northeast, and not between NH and the Northeast including NH.(ii) Messmer is incorrect in her assertion that we drew conclusions that de-emphasized any single possible cause of the observed patterns because we drew no conclusions about etiology. Our conclusions are limited to the findings relating to our a priori hypotheses that assessed NH's comparison with the Northeast, and the Northeast's comparison with other regions. We presented a balanced discussion of many possible causes but drew no conclusions about them.(iii) Our study was a first step toward identifying the extent of the problem, and we believe our neighbors in the Northeast should be as deeply concerned about childhood cancer as we are in NH. A different kind of study is needed to explore the reasons for these patterns, but etiologic research requires large numbers of participants with detailed individual-level data which are not readily available and this would be difficult to conduct in a single state with a relatively small number of cancers. The NCI is leading the National Childhood Cancer Data Initiative to pool data from many sources and states for research, recognizing that, “due to their rare nature, it has been challenging to collect substantial and vital information on a large scale to study and understand the needs for this unique population of cancer patients.” (3) NH is eligible to participate.NH and neighboring states each have special local concerns about high pediatric cancer incidence, pollution, and other factors. Arguments over whether NH ranks #1 are distracting; clearly the state's very high incidence should be investigated, but our findings open the door to a larger effort, if those involved in government throughout the Northeast act together on their collective responsibility. A Northeastern coalition could pool resources toward the conduct of research that includes local areas of concern in multiple states, because large-scale, rigorously designed and analyzed studies of childhood cancer etiology have the greatest chance of determining causes.The findings and conclusions in this report are those of the authors and do not necessarily represent the official positions of the New Hampshire Department of Environmental Services, the New Hampshire Department of Health and Human Services, or the CDC.See the original Letter to the Editor, p. 149J.R. Rees reports other support from the New Hampshire Department of Health and Human Services during the conduct of the study. M.O. Celaya reports other support from the New Hampshire Department of Health & Human Services during the conduct of the study. No disclosures were reported by the other authors.