Methods: Leveraging expertise at the Korle Bu Hospital National Diabetes Center, NYU diabetes care and research experts worked collaboratively with Ghanaian diabetes experts to conduct a needs assessment for EHR development by observation of the clinical setting, medical records area, and laboratory data repository. Currently used data-gathering patient forms were reviewed, as were EHR elements used in diabetes management in Kenya and the US. Specifications for the EHR were provided to the Ghanaian software developer.
Results:The laboratory database contained 3,886 of 5,000 patients. Clinic data consisted primarily of demographics, physiologic data from laboratory assays, and clinical anthropometric measurements. The IT team connected all hospital and NatiNDC laboratory auto-analyzers to a main data warehouse supported by a server with extensive back-up system. Programming of the EHR is now underway and we will report on implementation.
Conclusions: Current paper-based data gathering has not resulted in 100% of patient data being entered into the NDC database. An EHR will provide a mechanism for direct entry of clinical data thereby capturing all diabetes patients receiving care. The software platform, used on clinician laptops and whereby clinical visit data is exported and merged with laboratory data warehoused in a local server, is the first step in the EHR implementation. Once the process is further refined, e-connectivity established at the hospital, and NDC staff trained to implement the EHR, this open-source web-based tool can be extended to other diabetes clinics throughout Ghana/regionally.