Skip to main content

Table 1 Linking hospital and residential aged care events: estimated maximum false match rate for several event-based strategies

From: Event-based record linkage in health and aged care services data: a methodological innovation

 

Estimated maximum false match rate (%)

Linkage strategy

Population 10000 (single sex)

Population 35000 (single sex)

Base linkage strategy(a)

1.68

5.87

-but allowing up to a two day gap when matching a hospital separation to a new admission into residential aged care (3-date admission matching)

2.67

9.36

-but insisting episodes of hospital leave from residential aged care exactly match on both the start and end of the hospital episode

1.24

4.34

-but excluding matching to episodes of social leave from residential aged care as unlikely

1.05

3.66

-but insisting episodes of hospital leave from residential aged care exactly match on both the start and end of the hospital episode and at the same time excluding matching to episodes of social leave from residential aged care

0.61

2.14

Changing assumption to birth dates spread uniformly over 30 years

0.84

2.93

Changing assumption to birth dates spread across 5-year age groups as per aged care admissions over 30 years

1.10

3.86

  1. (a) Base linkage strategy includes the following assumptions to allow estimation:
  2. • 15 years of birth dates, uniform distribution
  3. • single-date (exact end-date) matching only of separations to residential aged care admissions and to residential aged care hospital leave
  4. • allow separations to match to social leave covering the hospital episode (cover matching).
  5. Calculations use national average hospital separation rates (2000–01) and residential aged care admission and residential aged care leave rates (2001–02) [38]. Estimates take into account the distribution of hospital episodes by length of stay.
Back to article page

BMC Health Services Research

ISSN: 1472-6963

Contact us