χ2 Woolf and p value for homogeneity, pooled estimates of prostate cancer risk, and 95% CIs for several groupings of the data
| Pooled | Homogeneity | |||||
|---|---|---|---|---|---|---|
| Grouping | No. studies | No. risk estimates | Rate ratio | 95% CI | χ2 Woolf | p value |
| No. risk estimates = number of estimators of relative risk, (fixed) = fixed effects estimates, otherwise random effects estimates. PMR, proportional mortality ratio; SMR, standardised mortality ratio; SIR, standardised incidence ratio; OR, odds ratio; MOR, mortality odds ratio; HWE, healthy worker effect (significantly lower than expected all-cancer mortality or incidence). *In the study of Aronson et al (1996), the control group consisted of the pool of cancer controls and population controls; pooled rate ratios are in bold when the 95% CI does not include 1. †In this subset analysis, only studies reporting RR estimators for prostate cancer as well as for all cancer were included. | ||||||
| All studies | ||||||
| All studies (cohort, case-control, PMR) | 22 | 25 | 1.13 | 1.04 to 1.22 | 370.727 | <0.001 |
| All studies excluding PMR | 18 | 21 | 1.09 | 1.00 to 1.19 | 271.805 | <0.001 |
| Study design | ||||||
| Cohort | ||||||
| All studies | 11 | 14 | 1.13 | 1.02 to 1.24 | 236.773 | <0.001 |
| SMR | 4 | 4 | 1.37 | 0.81 to 2.32 | 15.156 | 0.002 |
| SIR | 7 | 10 | 1.06 | 0.97 to 1.17 | 189.255 | <0.001 |
| Case-control | ||||||
| All studies | 7 | 7 | 0.98 | 0.71 to 1.37 | 34.948 | <0.001 |
| OR | 6 | 6 | 1.06 | 0.76 to 1.48 | 18.294 | 0.003 |
| MOR | 1 | 1 | 0.7 | 0.6 to 0.9 | – | – |
| Geographic location | ||||||
| Europe | ||||||
| All studies | 9 | 12 | 0.98 | 0.93 to 1.02 | 38.595 | <0.001 |
| Cohort | 6 | 9 | 0.98 | 0.94 to 1.03 | 34.229 | <0.001 |
| SMR | 2 | 2 | 0.93 | 0.60 to 1.42 | 0.104 | 0.747 (fixed) |
| SIR | 4 | 7 | 0.98 | 0.94 to 1.03 | 34.061 | <0.001 |
| Case-control | ||||||
| OR | 3 | 3 | 0.73 | 0.55 to 0.97 | 0.406 | 0.816 (fixed) |
| US/Canada | ||||||
| All studies | 8 | 8 | 1.50 | 1.08 to 2.07 | 88.134 | <0.001 |
| Cohort | 4 | 4 | 1.93 | 1.76 to 2.13 | 4.860 | 0.182 (fixed) |
| SMR | 2 | 2 | 1.89 | 1.13 to 3.16 | 4.052 | 0.044 |
| SIR | 2 | 2 | 1.90 | 1.71 to 2.11 | 0.202 | 0.653 (fixed) |
| Case-control | 4 | 4 | 1.21 | 0.75 to 1.96 | 29.578 | <0.001 |
| OR | 3 | 3 | 1.43 | 1.20 to 1.71 | 2.584 | 0.275 (fixed) |
| MOR | 1 | 1 | 0.7 | 0.6 to 0.9 | – | – |
| Others | ||||||
| SIR | 1 | 1 | 0.61 | 0.37 to 0.95 | – | – |
| Outcome | ||||||
| Incidence | ||||||
| All studies | 13 | 16 | 1.07 | 0.98 to 1.17 | 212.611 | <0.001 |
| SIR | 7 | 10 | 1.06 | 0.97 to 1.17 | 189.255 | <0.001 |
| OR | 6 | 6 | 1.06 | 0.76 to 1.48 | 18.294 | 0.002 |
| Mortality | ||||||
| All studies | 5 | 5 | 1.15 | 0.63 to 2.11 | 56.339 | <0.001 |
| SMR | 4 | 4 | 1.37 | 0.81 to 2.32 | 15.156 | 0.002 |
| MOR | 1 | 1 | 0.7 | 0.6 to 0.9 | – | – |
| Reference population | ||||||
| Cohort studies | ||||||
| National/regional | 11 | 14 | 1.13 | 1.02 to 1.24 | 236.773 | <0.001 |
| PMR | 4 | 4 | 1.23 | 1.09 to 1.38 | 26.945 | <0.001 |
| Case to control | ||||||
| Hospital | 2 | 2 | 0.76 | 0.56 to 1.04 | 0.027 | 0.870 (fixed) |
| Cancer | 1 | 1 | 1.43 | 1.161 to 1.761 | – | – |
| General population | 3 | 3 | 0.95 | 0.47 to 1.95 | 12.507 | 0.0019 |
| Cancer + general population* | 1 | 1 | 1.18 | 0.77 to 1.81 | – | – |
| Healthy worker effect (HWE)† | ||||||
| Cohort studies (SMR+SIR) + HWE | 6 | 9 | 1.15 | 1.01 to 1.31 | 218.445 | <0.001 |
| Cohort studies (SMR+SIR) − HWE | 2 | 2 | 0.86 | 0.58 to 1.28 | 0.547 | 0.460 (fixed) |
| Occupation category | ||||||
| Pesticide applicators | 4 | 4 | 1.64 | 1.13 to 2.38 | 64.697 | <0.001 |
| Farmers | 11 | 14 | 0.97 | 0.92 to 1.03 | 62.381 | <0.001 |
| USA/Canada | 5 | 5 | 1.26 | 0.83 to 1.90 | 31.355 | <0.001 |
| Europe | 6 | 9 | 0.96 | 0.92 to 1.01 | 28.923 | <0.001 |
| Prostate cancer versus all cancer | ||||||
| All cancer | 8 | 11 | 0.80 | 0.77 to 0.84 | 128.275 | <0.001 |
| Prostate cancer | 8 | 11 | 1.12 | 0.99 to 1.27 | 219.413 | <0.001 |