Table 2 |
||
|
Precision of each strategy across databases |
||
|
Actually relevant records retrieved ("diagnosed positive") by each strategy |
Precision: Actually relevant records ("diagnosed positive") as a percentage (number) of potentially relevant records for that strategy |
|
|
|
||
|
1)Thesaurus |
191 |
5.4% (191/3537) |
|
2)Free-text |
172 |
4.9% (172/3451) |
|
3)Broad-based |
187 |
4.7% (187/3912) |
|
Total actually relevant records (duplicates eliminated) |
262 |
- |
|
|
||
|
Shaw et al. BMC Medical Research Methodology 2004 4:5 doi:10.1186/1471-2288-4-5 |
||