With the introduction of the new generation Patient Monitor (code named PicoVue) the backbone of the localization aspects were the Unicode implementation and the leverage of the well established nls (native language support) processes used for the high end monitor (code named Merlin). See as reference the paper presented at the 11-th Unicode conference in San Jose (*** Unicode implementation in the PicoVue patient monitor***).

The HP monitors are now localized into 17 languages - 11 west-european, 3 asian and 3 east-european ones. Constraints on budgets and resources, as well as the goal to release the localized monitor close after the English release made it mandatory to improve the process of language verification and regression testing.

The purpose of this paper is to give insight views to some elements of the language verification process with special focus on patient safety aspects to keep in mind and benefits in term of effectiveness considering that the number of supported languages will even increase.

The improvements are in following areas:
o Establish the NlsAssist Tool as a mean to allow easy and fast language verification of patient safety critical text strings (like patient alarms) in the target monitor. This is only possible due to the Unicode based monolithic text catalog and tools to access it.
o Support the translation process by giving the translators in addition to the SW to localize also specific Screenshots to illustrate the context of the text strings to translate.
o Reduce to a minimum the language regressions test (and thus save project time and resources). At the same time put stronger emphasis on the language verification aspects (coverage, string context, medical meaning, etc).
o Costs and benefits with the improved approach.