Combining Decision Rules from Classification Tree Models and Expert Assessment to Estimate Occupational Exposure to Diesel Exhaust for a Case-Control Study.

OBJECTIVES: To efficiently and reproducibly assess occupational diesel exhaust exposure in a Spanish case-control study, we examined the utility of applying decision rules that had been extracted from expert estimates and questionnaire response patterns using classification tree (CT) models from a similar US study. METHODS: First, previously extracted CT decision rules were used to obtain initial ordinal (0-3) estimates of the probability, intensity, and frequency of occupational exposure to diesel exhaust for the 10 182 jobs reported in a Spanish case-control study of bladder cancer. Second, two experts reviewed the CT estimates for 350 jobs randomly selected from strata based on each CT rule's agreement with the expert ratings in the original study [agreement rate, from 0 (no agreement) to 1 (perfect agreement)]. Their agreement with each other and with the CT estimates was calculated using weighted kappa (κ w) and guided our choice of jobs for subsequent expert review. Third, an expert review comprised all jobs with lower confidence (low-to-moderate agreement rates or discordant assignments, n = 931) and a subset of jobs with a moderate to high CT probability rating and with moderately high agreement rates (n = 511). Logistic regression was used to examine the likelihood that an expert provided a different estimate than the CT estimate based on the CT rule agreement rates, the CT ordinal rating, and the availability of a module with diesel-related questions. RESULTS: Agreement between estimates made by two experts and between estimates made by each of the experts and the CT estimates was very high for jobs with estimates that were determined by rules with high CT agreement rates (κ w: 0.81-0.90). For jobs with estimates based on rules with lower agreement rates, moderate agreement was observed between the two experts (κ w: 0.42-0.67) and poor-to-moderate agreement was observed between the experts and the CT estimates (κ w: 0.09-0.57). In total, the expert review of 1442 jobs changed 156 probability estimates, 128 intensity estimates, and 614 frequency estimates. The expert was more likely to provide a different estimate when the CT rule agreement rate was <0.8, when the CT ordinal ratings were low to moderate, or when a module with diesel questions was available. CONCLUSIONS: Our reliability assessment provided important insight into where to prioritize additional expert review; as a result, only 14% of the jobs underwent expert review, substantially reducing the exposure assessment burden. Overall, we found that we could efficiently, reproducibly, and reliably apply CT decision rules from one study to assess exposure in another study.

Validity of expert assigned retrospective estimates of occupational polychlorinated biphenyl exposure.

Assessment of retrospective exposures based on expert judgment in case-control studies is usually of unknown validity because of the difficulty in finding gold standards for comparison. We investigated the relationship between expert-assigned retrospective occupational polychlorinated biphenyl (PCB) exposure estimates and serum PCB concentrations. Analyses were conducted on a subset of cases (n = 94) and controls (n = 96) in the multi-center National Cancer Institute, Surveillance, Epidemiology, and End Results Case-Control Study of non-Hodgkin lymphoma. Based on the subjects' lifetime work histories, an industrial hygienist assigned each job a probability of PCB exposure [<5% (unexposed), 5-<50% (possibly exposed), ≥50% (probably exposed)]. Ordinary least squares regression was used to investigate associations between the probability rating and log-transformed lipid-adjusted serum levels of 14 PCB congeners and total PCBs (ΓPCBs). Compared to unexposed participants (n = 163), those with a probably exposed job (n = 7) had serum levels that were 87% higher for ΓPCBs (95% confidence interval: 1.33-2.62) and 38% of serum level variability was explained by the probability rating. Statistically significant associations between probability ratings and serum levels for 12 of 14 individual congeners were also observed. In summary, the observed contrast in PCB serum levels by probability rating provides support for the occupational PCB exposure assessment.

A case-control study of occupation/industry and renal cell carcinoma risk.

BACKGROUND: The role of occupation in the etiology of renal cell carcinoma (RCC) is unclear. Here, we investigated associations between employment in specific occupations and industries and RCC, and its most common histologic subtype, clear cell RCC (ccRCC). METHODS: Between 2002 and 2007, a population-based case-control study of Caucasians and African Americans (1,217 cases; 1,235 controls) was conducted within the Detroit and Chicago metropolitan areas to investigate risk factors for RCC. As part of this study, occupational histories were ascertained through in-person interviews. We computed odds ratios (ORs) and 95% confidence intervals (CIs) relating occupation and industry to RCC risk using adjusted unconditional logistic regression models. RESULTS: Employment in the agricultural crop production industry for five years or more was associated with RCC (OR = 3.3 [95% CI = 1.0-11.5]) and ccRCC in particular (OR = 6.3 [95% CI = 1.7-23.3], P for trend with duration of employment = 0.0050). Similarly, RCC risk was elevated for employment of five years or longer in non-managerial agricultural and related occupations (ORRCC = 2.1 [95% CI = 1.0-4.5]; ORccRCC = 3.1 [95% CI = 1.4-6.8]). Employment in the dry-cleaning industry was also associated with elevated risk (ORRCC = 2.0 [95% CI = 0.9-4.4], P for trend = 0.093; ORccRCC = 3.0 [95% CI = 1.2-7.4], P for trend = 0.031). Suggestive elevated associations were observed for police/public safety workers, health care workers and technicians, and employment in the electronics, auto repair, and cleaning/janitorial services industries; protective associations were suggested for many white-collar jobs including computer science and administrative occupations as well employment in the business, legislative, and education industries. CONCLUSIONS: Our findings provide support for an elevated risk of RCC in the agricultural and dry-cleaning industries and suggest that these associations may be stronger for the ccRCC subtype. Additional studies are needed to confirm these findings.

Family history and the risk of kidney cancer: a multicenter case-control study in Central Europe.

An elevated familial relative risk may indicate either an important genetic component in etiology or shared environmental exposures within the family. Incidence rates of kidney cancer are particularly high in Central Europe, although no data were available on the familial aggregation or genetic background of kidney cancer in this region. We have, therefore, investigated the role of family history in first-degree relatives in a large multicenter case-control study in Central Europe. A total number of 1,097 cases of kidney cancer and 1,555 controls were recruited from 2000 to 2003 from seven centers in Czech Republic, Poland, Romania, and Russia. The risk of kidney cancer increased with the increasing number of relatives with history of any cancer [odds ratio (OR), 1.15; 95% confidence interval (95% CI), 1.00-1.31 per affected relative], and this association seemed to be more prominent among subjects with young onset (OR, 1.55; 95% CI, 1.09-2.20 per affected relative). Overall, the OR was 1.40 (95% CI, 0.71- 2.76) for the subjects who had at least one first-degree relative with kidney cancer after adjusting for tobacco smoking, body mass index, and medical history of hypertension, and this association was most apparent among subjects with affected siblings (OR, 4.09; 95% CI, 1.09-15.4). Based on the relative risk to siblings in our study population, we estimated that 80% of the kidney cancer cases are likely to occur in 20% of the population with the highest genetic risk, which indicate the importance of further investigation of genetic factors in cancer prevention for kidney cancer.

Uncommon CHEK2 mis-sense variant and reduced risk of tobacco-related cancers: case control study.

CHEK2 is a key cell cycle control gene encoding a pluripotent kinase that can cause arrest or apoptosis in response to unrepaired DNA damage. We report a large case-control study of a non-functional variant that had previously been expected to increase cancer rates. Four thousand and fifteen cancer patients (2250 lung, 811 squamous upper aero-digestive and 954 kidney) and 3052 controls in central Europe were genotyped for the mis-sense variant rs17879961 (replacement of T by C), which changes an amino acid (I157T) in an active site of the gene product. The heterozygous (T/C) genotype was associated with a highly significantly lower incidence of lung cancer than the common T/T genotype [relative risk (RR), T/C versus T/T, 0.44, with 95% confidence interval (CI) 0.31-0.63, P < 0.00001] and with a significantly lower incidence of upper aero-digestive cancer (RR 0.44, CI 0.26-0.73, P = 0.001; P = 0.000001 for lung or upper aero-digestive cancer). Protection was significantly greater for squamous than adenomatous lung cancer (P = 0.001). There was an increase of borderline significance in kidney cancer (RR 1.44, CI 0.99-2.00, P = 0.06). This unexpected halving of tobacco-related cancer (since replicated independently) implies much greater absolute risk reduction in smokers than in non-smokers. The mechanism is unknown: perhaps squamous stem cell apoptosis following smoke exposure causes net harm (e.g. by forcing nearby stem cells to divide before they have repaired their own DNA damage from tobacco smoke). If so, reducing the rate of apoptosis by reducing CHEK2 activity could be protective-although not smoking would be far more so.

Impact of misclassification in genotype-exposure interaction studies: example of N-acetyltransferase 2 (NAT2), smoking, and bladder cancer.

Errors in genotype determination can lead to bias in the estimation of genotype effects and gene-environment interactions and increases in the sample size required for molecular epidemiologic studies. We evaluated the effect of genotype misclassification on odds ratio estimates and sample size requirements for a study of NAT2 acetylation status, smoking, and bladder cancer risk. Errors in the assignment of NAT2 acetylation status by a commonly used 3-single nucleotide polymorphism (SNP) genotyping assay, compared with an 11-SNP assay, were relatively small (sensitivity of 94% and specificity of 100%) and resulted in only slight biases of the interaction parameters. However, use of the 11-SNP assay resulted in a substantial decrease in sample size needs to detect a previously reported NAT2-smoking interaction for bladder cancer: 1,121 cases instead of 1,444 cases, assuming a 1:1 case-control ratio. This example illustrates how reducing genotype misclassification can result in substantial decreases in sample size requirements and possibly substantial decreases in the cost of studies to evaluate interactions.