Thursday, December 1, 2011

Tips from a Road Warrior (18): What’s in a Name?

I was in the departure lounge of an international airport waiting for a delayed flight when I heard a familiar name over the intercom. It sounded vaguely as if it could have been my last name, but it was not quite as I was used to it being pronounced. Just to make sure, I went to the desk and asked whether they called my name and they told me that they indeed had done so several times. It turned out they wanted to notify me that they had put me on another flight because I would have missed my connection otherwise. 

The incident serves as a reminder of the challenge names pose for healthcare imaging and IT systems. How we identify people, obviously plays a pivitol role in the integrity of the information managed by healthcare imaging and IT systems. First of all, they might be entered incorrectly. A common error is that a data-entry person inputs the last and first name in a single field, which either should contain only the last or first name. Another common problem is a name change due to changes in marital status. At worst, the impact could be that a previous record, result or image might not be available or, at best case has to be merged. The standard actually allows for a special "merge" or “update” transaction to take care of this. 

Other name issues occur with truncations. Not that the connectivity standard does not support sufficient characters, but because an input device has certain limits. I have seen a simple data entry device using a small handheld that only had 16 characters maximum as input. Many foreign names can be longer than that. Some cultures also have double last names, for example, Hispanics typically carry their mother's and father’s names. In my native country, the Netherlands, it is common that a woman maintains both her own and her husband’s name. As a matter of fact, my spouse has been stopped at several international security posts because of the difference between her names in her European passport and how it commonly is entered on airline tickets. 

An additional complication with several European names is the presence of “prefixes,” such as “de la” in French, “von” in German, and “van de” or “van 't” in Dutch with several variations. Many US generated healthcare IT systems do not know how to handle this, or require special configuration options to deal with this. A recent post on a user-group highlighted this issue with a popular CT system that could not match names with these prefixes in its worklist. 

A problem also occurs when people have names in a language and corresponding character set that does not have a exact mapping into English. This is the case with several Asian languages as well as Middle Eastern languages such as Arabic and Hebrew. The name Abraham can be spelled Abrahim and there are many spellings of Mohammed. With medical care becoming increasingly global, it is not uncommon for a patient to be screened initially in a clinic in Dubai, for example, and then treated in a US institution. The international name issue could be easier if the software processing the patient information supports multiple international character sets. This is relatively easy to check in the interface specification. In some cases the name is the only differentiator, as can be the case with identical twins who have the same sex, and birth date, as well as address. 

In conclusion, names are tricky and can cause problems with identification and matching of the right patient records, especially in countries that lack a universal patient identifier (such as the US), making patients identification an ongoing challenge.