Skip to main content

Table 1 Probability estimates for linkage analyses

From: Investigating increased admissions to neonatal intensive care in England between 1995 and 2006: data linkage study using Hospital Episode Statistics

Matching variable

Baseline best guesses

Dattani et al. [15] estimate

 

m

u

\(w_{\mathrm {m}}^{\mathrm {a}}\)

\(w_{\text {nm}}^{\mathrm {b}}\)

m

u

\(w_{\mathrm {m}}^{\mathrm {a}}\)

\(w_{\text {nm}}^{\mathrm {b}}\)

Date of birth

0.90

0.00274

5.794

-2.3

0.7405

0.0015

6.202

-1.347

GA at birth

0.80

0.02

3.689

-1.589

0.4941

0.0494

2.3028

-0.6308

Sex

0.999

0.49

0.7123

-6.2344

0.7208

0.0062

4.756

-1.270

Discharge date

0.20

0.002

4.6052

-0.2211

Date of death c

0.20

0.00274

4.2904

-0.2204

0.30

0.002

5.0106

-0.3547

Birth weight

0.60

0.001

6.3969

-0.9153

0.7405

0.0074

4.606

-1.342

Birth order

0.87

0.95

-0.08797

0.95551

0.8153

0.0033

5.510

-1.686

Delivery method c

0.80

0.80

0

0

0.67

0.1

1.902

-1.003

Ethnic category

0.20

0.10

0.6931

-0.1178

0.7308

0.095

2.040

-1.212

Mother’s age at delivery

0.95

0.05

2.944

-2.944

Mother’s date of birth

0.90

0.0001

9.105

-2.302

Postcode

0.90

0.001

6.802

-2.302

0.9291

0.065

2.660

-2.579

Number of previous pregnancies

0.60

0.90

-0.4055

1.3863

Number of babies

0.95

0.95

0

0

0.8153

0.0033

5.510

-1.686

  1. Probability estimates for linkage analyses between Hospital Episode Statistics and EPICure data based on best guesses and prior knowledge (adapted from data linkage performed by Dattani et al between Hospital Episode Statistics (HES) and NHS Numbers 4 Babies data sets) [15]
  2. a w m= weight if pairs match
  3. b w nm= weight if pairs do not match
  4. cDate of death and delivery method were both modified using an adjusted best guess for the second linkage analysis performed using estimates from Dattani et al.