Regular Expression | REGEX for ICD9 codes

Question

I am using Python to extract ICD9 codes. And am using the below regular expression

icdRegex = recomp('V\d{2}\.\d{1,2}|\d{3}\.\d{1,2}|E\d{3}\.\d')

It captures pattern similar to 137.98 or V35.62

Everything works fine except the expression also captures patient weights as ICD9 code.

Now what I observed is, the weight is almost always appears as ex: 110.67 kg or kgs or lb or lbs

How do I separate ICD9 from weight !?

You might use a negative lookahead. [See this question](http://stackoverflow.com/questions/12210807/what-does-mean) — HamZa, Dec 02 '14 at 23:27
I did try most of the stuff here on stackoverflow, but it wouldn't work. May be I am a little too dumb for regex. — WeShall, Dec 02 '14 at 23:32

score 1 · Accepted Answer · edited Dec 02 '14 at 23:55

1

Add a negative lookahead assertion like the follwing:

(V\d{2}\.\d{1,2}|\d{3}\.\d{1,2}|E\d{3}\.\d)\b(?!\s?(?:lb|kg)s?)

edited Dec 02 '14 at 23:55

HamZa

answered Dec 02 '14 at 23:47

chapelo

Something's missing... It still captures fat weights as ICD9 code – WeShall Dec 02 '14 at 23:56

score 1 · Answer 2 · edited Dec 03 '14 at 00:25

1

Here is HamZa's expression for everyone:

icdRegex = recomp("\b(?:V\d{2}\.\d{1,2}|\d{3}\.\d{1,2}|E\d{3}\.\d)\b(?!\s*(?:kg|lb)s?\b)")

Thanks HamZa & Chapelo for helping out. Appreciate it.

edited Dec 03 '14 at 00:25

HamZa

answered Dec 03 '14 at 00:15

WeShall

Guys... for the REGEX we wrote almost 2 month's back, failed on one condition. For numbers starting with 0 it is capturing wrong pattern. For example: if ICD9 code is 032.9, the expression will return it as 329. Any fix for codes starting with 0 !? – WeShall Jan 27 '15 at 21:29
Further refinement to this thread [here](http://stackoverflow.com/questions/28200337/how-to-make-regex-ignore-a-pattern-following-a-specific-group) – WeShall Jan 29 '15 at 17:15

2 Answers2