Table 4 The average sensitivity and PPV (standard deviation in parentheses) over 100 iterations for each data type used for spatial cluster detection under six simulation scenarios
Type of data | Individual-level data | Aggregate-level data | |
|---|---|---|---|
Frequency | Weighted frequency | Crude rate | |
Scenario 1 [true cluster (A) + sampling proportion (1)] | |||
 Sensitivity | 0.9214 (0.0425) | 0.9381 (0.0536) | 0.9435 (0.0055) |
 PPV | 0.8071 (0.1312) | 0.1966 (0.0187) | 0.9980 (0.0200) |
Scenario 2 [true cluster (A) + sampling proportion (2)] | |||
 Sensitivity | 0.9320 (0.0406) | 0.9286 (0.0607) | 0.9440 (0.0000) |
 PPV | 0.7861 (0.1255) | 0.1889 (0.0163) | 0.9937 (0.0251) |
Scenario 3 [true cluster (A) + sampling proportion (3)] | |||
 Sensitivity | 0.9203 (0.0379) | 0.9280 (0.0549) | 0.9435 (0.0055) |
 PPV | 0.7999 (0.1186) | 0.1852 (0.0193) | 0.9994 (0.0059) |
Scenario 4 [true cluster (B) + sampling proportion (1)] | |||
 Sensitivity | 0.9163 (0.0350) | 0.9378 (0.0402) | 0.9192 (0.0218) |
 PPV | 0.8734 (0.0752) | 0.3865 (0.0370) | 0.9797 (0.0240) |
Scenario 5 [true cluster (B) + sampling proportion (2)] | |||
 Sensitivity | 0.9203 (0.0299) | 0.9151 (0.0479) | 0.9189 (0.0202) |
 PPV | 0.8749 (0.0766) | 0.3640 (0.0419) | 0.9795 (0.0221) |
Scenario 6 [true cluster (B) + sampling proportion (3)] | |||
 Sensitivity | 0.9183 (0.0339) | 0.9244 (0.0476) | 0.9286 (0.0252) |
 PPV | 0.8678 (0.0765) | 0.3509 (0.0379) | 0.9708 (0.0255) |