Development, Implementation and Preliminary Results of an Electronic Reminder for HIV Screening Using a Service Oriented Architecture.

30% of the Argentinian population and 58% of Plan de Salud HIBA patients are unaware of their HIV status. The Ministry of Health and US Preventive Service Task recommends physicians to assess HIV infection in persons aged 15 to 65. An HIV screening reminder integrated in an electronic health record (EHR) was created using FHIR to represent clinical information and CDS-Hooks to represent the exchange of information with a CDS service. The tool had a 1% intervention rate, and 67.4% acceptance rate. The number of HIV screening tests requested during the weeks after the CDSS implementation and in the same period in 2017 were obtained. 575 orders were requested in the 2017 period and 893 in the 2018. 89 (almost 10%) of these came from the electronic tool. The preliminary results indicate that this non disruptive, action oriented reminder can contribute to increased HIV screening orders.

Statistical Process Control: A Quality Tool for a Venous Thromboembolic Disease Registry.

OBJECTIVE: We aim to describe Statistical Control Process as a quality tool for the Institutional Registry of Venous Thromboembolic Disease (IRTD), a registry developed in a community-care tertiary hospital in Buenos Aires, Argentina. METHODS: The IRTD is a prospective cohort. The process of data acquisition began with the creation of a computerized alert generated whenever physicians requested imaging or laboratory study to diagnose venous thromboembolism, which defined eligible patients. The process then followed a structured methodology for patient's inclusion, evaluation, and posterior data entry. To control this process, process performance indicators were designed to be measured monthly. These included the number of eligible patients, the number of included patients, median time to patient's evaluation, and percentage of patients lost to evaluation. Control charts were graphed for each indicator. RESULTS: The registry was evaluated in 93 months, where 25,757 patients were reported and 6,798 patients met inclusion criteria. The median time to evaluation was 20 hours (SD, 12) and 7.7% of the total was lost to evaluation. Each indicator presented trends over time, caused by structural changes and improvement cycles, and therefore the central limit suffered inflexions. CONCLUSION: Statistical process control through process performance indicators allowed us to control the performance of the registry over time to detect systematic problems. We postulate that this approach could be reproduced for other clinical registries.