From: Validation of de-identified record linkage to ascertain hospital admissions in a cohort study
 | Linkage Variables | Records linked | Records remaining |
---|---|---|---|
Iteration | Medicare8 1 + Medsuf1 2 group | Â | |
1-4 | Medicare8+Medsuf1+Yearbirth+Monthbirth+Daybirth+Sex+COB+ Postcodes1-4 | 1594 | 406 |
5-16 | Repeat steps 1-4 using 3 different variations of Medsuf1 | 8 | 398 |
17 | Repeat step 1 dropping postcode | 23 | 375 |
18-21 | Repeat steps 1-4 dropping COB | 7 | 368 |
22-25 | Repeat steps 1-4 dropping sex | 1 | 367 |
26-29 | Repeat steps 1-4 dropping day of birth | 19 | 348 |
30-33 | Repeat steps 1-4 dropping month of birth | 2 | 346 |
34-37 | Repeat steps 1-4 dropping year of birth | 17 | 329 |
38-100 | Repeat steps 5-16, dropping 1 variable | 0 | 329 |
101-141 | Repeat steps 1-4 dropping two demographic variables | 4 | 325 |
 | Medicare8 group | ||
142-145 | Medicare8+Yearbirth+Monthbirth+Daybirth+Sex+COB+ Postcodes 1-4 | 24 | 301 |
146-166 | Repeat steps 142-145 dropping 1 variable | 2 | 299 |
 | Medsuf1 Group | ||
167-170 | Medsuf1+Yearbirth+Monthbirth+Daybirth+Sex+COB+ Postcodes 1-4 | 39 | 260 |
Total records linked = 1740/2000 = 87% |