BUILDING BLOCKS VENDOR CASE STUDY #1: ADT Message Validation with a Hospital Partner

OZ Systems

BACKGROUND

OZ Systems (OZ) partnered with the Virginia Department of Health (VDH) – Division of Consolidated Laboratory Services (DCLS) to improve newborn screening timeliness by implementing Telepathy™ Newborn Screening (NBS) at Virginia hospitals. A select group of hospitals purchased the Newborn Admission Notification Information (NANI) tool to automate the near to real time transmission of demographic data from hospital electronic health records (EHRs) to the Telepathy™ NBS application. NANI provides a baseline denominator of all hospital births, and reduces the amount of data entry on the newborn screening card. NANI is an Integrated the Healthcare Enterprise (IHE) profile that is based on Health Level 7 (HL7) Version 2 Admission, Discharge, and Transfer (ADT) messages.

PROBLEM

Newborn screening cards are often missing data elements or are illegible. Handwriting discrepancies can lead to misspelling upon data entry and a longer time to follow-up on children if the data are illegible. While every baby in a hospital has an electronic health record, their data are not always recorded on the newborn screening card. The lab often spends time tracking down the correct information from the hospital due to missing data elements or illegibility. 

SOLUTION

The OZ solution, Telepathy™ NBS eliminates handwriting on the card, by populating data elements from the EHR, and sends electronic order messages to the Laboratory Information System (LIMS). To begin this process, OZ reached out to hospitals who were interested in the pilot project and contracted with them to implement NANI as part of the Telepathy™ NBS solution. While ADT messages are commonly used in hospitals, NANI requires message validation to identify minor modifications and ensure that the information is transmitted in the standard format. Initially OZ worked with DCLS to identify the required data elements for the program and those data elements that were required but could be left empty. 

At each facility, OZ started with a kick-off to review the overall project and identify a project team. The facility received documentation outlining technical specifications and requirements. The hospital modified the facility’s ADT messages to be compliant with the documentation. 

OZ worked closely with facilities to identify the required data elements. Content testing occurred and issues were identified and resolved. The hospital had the ability to validate its own messages for compliance with the NANI documentation by using a proprietary online NANI validation tool. This allowed the hospital to have control over initial validation to determine changes. The OZ team further validated the messages to ensure they populate appropriately into the Telepathy™ NBS application. OZ Systems used free tools to validate content and connectivity such as SmartHL7 Viewer and SmartHL7 Sender (http://smarthl7.com/).   

There were three control messages developed, an A01, an A03 and an A08. These messages are tested throughout the process. The team created test scenarios to ensure all data elements are appropriately validated including, but not limited to a healthy baby who was admitted, updated and discharged within the well-baby nursery; a sick baby who is admitted, moved the a higher level of care and discharged home; and a deceased baby who is born but passes away prior to discharge from the facility. Within these three scenarios robust demographic data elements are captured for patient and next of kin based on the requirements specifically identified by DCLS.

The most common findings through message validation were:

  • Data elements such as relationship type, race and ethnicity required translation tables to ensure the use of HL7 standard codes. For example, relationship type should be sent as the code “MTH” to indicate Mother, however, some hospitals used M, Mother and Mom.
  • Birth Weight required a conversion from kilograms to grams.
  • Time of Birth required special testing and algorithms in cases where EHRs did not send time in their date/time of birth field.

Lessons learned showed that when developing timelines, the team needs to consider competing EHR related projects and the availability of facility staff, such as network engineers for VPNs and ADT testers. The understanding and expertise of hospital staff influence the speed of a NANI implementation.