I know! Let's Translate our data format into XML! This will be the bees knees!

reaganomics

And here I thought government meddling was
the main problem with health care costs rising. Take a look at
what the brainiacs at HL7 (a data format standards organization for the
health care industry) did to convert their old EDI standard to
XML. Someone needs to be shot!

Here we see a sample message in the syntax of the standard encoding rules.<o:p></o:p>

MSH|^~\&|LAB|767543|ADT|767543|19900314130405||ADT^A04|XX3657|P|2.3.1<CR><o:p></o:p>

EVN|A01|19980327101314|19980327095000|I||19980327095000<CR><o:p></o:p>

PID|1||123456789ABCDEF|123456789ABCDEF|PATIENT^BOB^S||19590520|M||<o:p></o:p>

<st1:Street w:st="on"><st1:address w:st="on">612345 MAIN STREET</st1:address></st1:Street>^^ANYTOWN^CA^91234||714-555-1212|<o:p></o:p>

714-555-1212||||123456789ABCDEF|||U<CR><o:p></o:p>

PD1||||WELBY<CR><o:p></o:p>

PV1|1|0||NEW|||SPOCK<CR><o:p></o:p>

Here we see that sample message in the syntax of the XML encoding rules.<o:p></o:p>

<!DOCTYPE ADT_A03 SYSTEM "hl7_v231.dtd"><o:p></o:p>

<ADT_A03><o:p></o:p>

<MSH.1>|</MSH.1><o:p></o:p>

<MSH.2>^~\&</MSH.2><o:p></o:p>

<MSH.3><HD.1>LAB</HD.1></MSH.3><o:p></o:p>

<MSH.4><HD.1>767543</HD.1></MSH.4><o:p></o:p>

<MSH.5><HD.1>ADT</HD.1></MSH.5><o:p></o:p>

<MSH.6><HD.1>767543</HD.1></MSH.6><o:p></o:p>

<MSH.7>19900314130405</MSH.7><o:p></o:p>

<MSH.9><o:p></o:p>

<CM_MSG_TYPE.1>ADT</CM_MSG_TYPE.1><o:p></o:p>

<CM_MSG_TYPE.2>A04</CM_MSG_TYPE.2><o:p></o:p>

</MSH.9><o:p></o:p>

<MSH.10>XX3657</MSH.10><o:p></o:p>

<MSH.11><PT.1>P</PT.1></MSH.11><o:p></o:p>

<MSH.12><VID.1>2.3.1</VID.1></MSH.12><o:p></o:p>

</MSH><o:p></o:p>

<EVN.1>A01</EVN.1><o:p></o:p>

<EVN.2>19980327101314</EVN.2><o:p></o:p>

<EVN.3>19980327095000</EVN.3><o:p></o:p>

<EVN.4>I</EVN.4><o:p></o:p>

<EVN.6>19980327095000</EVN.6><o:p></o:p>

</EVN><o:p></o:p>

<PID.1>1</PID.1><o:p></o:p>

<PID.3.LST><o:p></o:p>

<PID.3><CX.1>123456789ABCDEF</CX.1></PID.3><o:p></o:p>

19<o:p></o:p>

</PID.3.LST><o:p></o:p>

<PID.4.LST><o:p></o:p>

<PID.4><CX.1>123456789ABCDEF</CX.1></PID.4><o:p></o:p>

</PID.4.LST><o:p></o:p>

<PID.5.LST><o:p></o:p>

<PID.5><o:p></o:p>

<XPN.1>PATIENT</XPN.1><o:p></o:p>

<XPN.2>BOB</XPN.2><o:p></o:p>

<XPN.3>S</XPN.3><o:p></o:p>

</PID.5><o:p></o:p>

</PID.5.LST><o:p></o:p>

<PID.7>19590520</PID.7><o:p></o:p>

<PID.8>M</PID.8><o:p></o:p>

<PID.10.LST><o:p></o:p>

<PID.10><CE.1>6</CE.1></PID.10><o:p></o:p>

</PID.10.LST><o:p></o:p>

<PID.11.LST><o:p></o:p>

<PID.11><o:p></o:p>

<XAD.1><st1:Street w:st="on"><st1:address w:st="on">12345 MAIN STREET</st1:address></st1:Street></XAD.1><o:p></o:p>

<XAD.3>ANYTOWN</XAD.3><o:p></o:p>

<XAD.4>CA</XAD.4><o:p></o:p>

<XAD.5>91234</XAD.5><o:p></o:p>

</PID.11><o:p></o:p>

</PID.11.LST><o:p></o:p>

<PID.13.LST><o:p></o:p>

<PID.13><XTN.1>714-555-1212</XTN.1></PID.13><o:p></o:p>

</PID.13.LST><o:p></o:p>

<PID.14.LST><o:p></o:p>

<PID.14><XTN.1>714-555-1212</XTN.1></PID.14><o:p></o:p>

</PID.14.LST><o:p></o:p>

<PID.18><CX.1>123456789ABCDEF</CX.1></PID.18><o:p></o:p>

<PID.21.LST><o:p></o:p>

<PID.21><CX.1>U</CX.1></PID.21><o:p></o:p>

</PID.21.LST><o:p></o:p>

</PID><o:p></o:p>

<PD1.4.LST><o:p></o:p>

<PD1.4><XCN.1>WELBY</XCN.1></PD1.4><o:p></o:p>

</PD1.4.LST><o:p></o:p>

</PD1><o:p></o:p>

<PV1.1>1</PV1.1><o:p></o:p>

<PV1.2>0</PV1.2><o:p></o:p>

<PV1.4>NEW</PV1.4><o:p></o:p>

<PV1.7.LST><o:p></o:p>

<PV1.7><XCN.1>SPOCK</XCN.1></PV1.7><o:p></o:p>

</PV1.7.LST><o:p></o:p>

</PV1><o:p></o:p>

</ADT_A03><o:p></o:p>

As is always the case with XML when processed with a validating processor, the extra whitespace between<o:p></o:p>

elements (provided to make the message easier for people to read) can be removed in actual message<o:p></o:p>

instances, resulting is shorter messages in situations
when overall message length is a factor.

merreborn

I guess that's marginally better than this classic wtf:

Alex Papadimoulis / Dec 16, 2005

The Flat-File Society Does XML

A little while back, Phil introduced us to a group of developers from The Flat File Society who were forced to do relational database. Phil is still at the client site and was tapped to lend a hand with a communication project the programmers were struggling with. They were able to get their data...

Some people just don't get it.

dhromed

Much clearer now.

It has structure!

Xarium

I happen to be a big fan of XML and I die a little inside whenever I see this kind of abuse. Too many poor developers/managers see it as a panacea simply because it happens to be "the latest cool thing".

No technology can fix a bad design.

Benanov

Missed the boat here...

    I was hoping that they'd actually modify their standard to encapsulate what the fields actually meant into the tags...utilize Dublin Core, etc.

    Instead they just basically changed their delimiter from the old pipe character into fancy new XML tags, creating everyone's favorite WTF.

    Not as good as the people who put flat file data ino a text field of a relational database.

    (Although here at this company there are a few XML files in NTEXT columns. I suppose that counts as a WTF. It makes the entire thing quite pointless to modify without the tools which used to fail very often.)

DZ_Jay

@Xarium said:

No technology can fix a bad design.

I don't know, hammers and steel-toed boots seem to do the trick sometimes.

-dZ.