Skip to main content

Table 5 Heuristics to identify date of birth, with entirely ficitious examples, as they would appear in source records and CRIS output

From: Development and evaluation of a de-identification procedure for a case register sourced from mental health electronic records

 

Algorithm to identify date of birth (Number date) <beginning > <day|month|year > <one_date_delimiter > <day|month|year > <one_date_delimiter > <day|month|year > <end>

 

Source record

<beginning>

<day|month| year>

<date_delimiter>

< day|month|year>

<date_delimiter>

<day|month|year>

<end>

CRIS output

Dob: 01/01/2001

:

01

/

01

/

2001

(space)

Dob: ZZZZZ

1st of January 2001

(Space)

1st

(space) of (space)

January

(space)

2001

(space)

ZZZZZ

…born in Jan 1st 01…

(space)

Jan

(space)

1st

(space)

01

(space)

…born in ZZZZZ…

…01-01-’01…

(space)

01

-

01

-‘

01

(space)

…ZZZZZ…

…01 Jan 2001

(space)

01

(space)

Jan

(space)

2001

(space)

…ZZZZZ…

Dob: 01//01/2001

:

01

/

None identified owing to typographical error in the source record

None

None

None

Dob: 01//01/2001