Geographic bias related to geocoding in epidemiologic studies.

(Smoothed maps of the annualized, age-adjusted prostate cancer rates for all males in Virginia 1990–99 are shown in Figure 1 .) Statistical testing for global clustering was highly significant for the entire time period (Tango's Maximum Excess Events Test, p < 0.008) and for 1990–94 and 1995–99 separately (p < 0.001 for both).

also in 132,142 other papers