False Positives in Cancer Epidemiology

Abstract

Background: A recent attempt to estimate the false-positive rate for cancer epidemiology studies is based on agents in International Agency for Research on Cancer (IARC) category 3 (agent not classifiable as to its carcinogenicity to humans) in the IARC Monographs Program.Methods: The estimation method is critiqued regarding biases caused by its reliance on the IARC classification criteria for assessing carcinogenic potential.Results: The privileged position given to epidemiologic studies by the IARC criteria ensures that the percentage of positive epidemiologic studies for an agent will depend strongly on the IARC category to which the agent is assigned. Because IARC category 3 is composed of agents with the lowest-assessed carcinogenic potential to which the estimation approach in question could be applied, a spuriously low estimated falsepositive rate was necessarily the outcome of this approach.Conclusions: Tendentious estimation approaches like that employed will by necessity produce spuriously low and misleading false positive rates.Impact: The recently reported estimates of the false-positive rate in cancer epidemiology are seriously biased and contribute nothing substantive to the literature on the very real problems related to false-positive findings in epidemiology. Cancer Epidemiol Biomarkers Prev; 22(1); 11-15. Ó2012 AACR.

Mortality from solid tumors in the updated NCI formaldehyde cohort

“…The new NPC death was observed in plant 3 of the NCI study. As has been reported previously, in occupational cohort studies worldwide, only plant 1 of the NCI study shows evidence of increased NPC risk in formaldehyde-exposed workers [Marsh and Youk, 2005;Tarone and McLaughlin, 2005;Bosetti et al, 2008;McLaughlin and Tarone, 2013]. Including the additional NPC death in plant 3, the estimated (based on relative contributions to person-years reported by Beane Freeman et al [2013]) SMR for NPC among formaldehyde-exposed workers in plants 2-10 of the NCI cohort is 0.7 (95% CI, 0.2-2.2) based on 3 NPC deaths.…”

“…The OR for NPC among embalmers was 0.1 (95% CI, 0.01-1.2), even though embalmers were reported to have the highest peak exposure to formaldehyde of any known occupation [Hauptmann et al, 2009]. A scientifically sound summary of all of the epidemiologic evidence regarding occupational formaldehyde exposure and NPC risk would conclude that something in plant 1 of the NCI study or associated with the geographic area containing plant 1 may have led to increased risk of NPC [Marsh et al, 2007], but that occupational exposure to formaldehyde is unlikely to be a risk factor for NPC [McLaughlin and Tarone, 2013 …”

Type 2 diabetes and cancer: umbrella review of meta-analyses of observational studies

“…5 6 Empirical evidence suggests that biased observational associations might be common in the literature on cancer epidemiology. [7][8][9][10][11][12][13] Usual suspected mechanisms that can operate are selective non-publication of "negative" results and selective choice of reporting analyses and outcomes that have stronger effects. 50 51 The extent to which the literature on diabetes and cancer is affected by such biases is difficult to prove definitively.…”

“…6 Biases have been suspected, documented, or debated in observational epidemiology of cancer for multiple putative carcinogens, risk factors, biomarkers, and prognostic factors. [7][8][9][10][11][12][13] To understand the strength of evidence and extent of potential biases in the claimed associations between type 2 diabetes and risk of developing cancer, we performed an umbrella review of the evidence across published meta-analyses or systematic reviews. We summarised the evidence on the incidence of cancer and cancer mortality for each site that has been studied in its association with type 2 diabetes; described the magnitude, direction, and significance of the observed associations; evaluated whether there are hints of biases in this evidence and how they manifest; and identified which are the most robust associations without potential biases.…”

Issues of methods and interpretation in the National Cancer Institute formaldehyde cohort study

“…As noted by McLaughlin and Tarone in a commentary [25] and recent letter to the editor and response [26,27], Beane Freeman et al [6] cite the NCI embalmer case–control study [28] in their discussion as evidence for a possible leukemia risk among FA-exposed workers, but ignore the absence of increased risk of NPC in the embalmer study. The odds ratio for NPC among embalmers was 0.10 (95% CI = 0.01-1.20), despite the fact that embalmers were reported to have the highest peak exposure to FA of any known occupation [28].…”