APHL 2022 POSTER ABSTRACTS
Utilization of a Project Management Planning Model for Client Enrollment in an Electronic Test Ordering and Reporting System ( ETOR ) in the New Jersey Public Health Laboratory
S Mikorski , H Hergert , D Woell , J Ochal , D Rivera , S Malaviya , M Rostash , M Ellethy , J HaydukKramer , W Underwood , A Oyelade , T Frez , A Shah , S Patel , R Eng , A Pryamkova , A Gross , R Finney , R Siderits and T Kirn , New Jersey Public Health and Environmental Laboratories
Orchard Outreach , an Orchard electronic test ordering and reporting ( ETOR ) LIMS component , was purchased and installed using Centers for Disease Control and Prevention ( CDC ) Epidemiology and Laboratory Capacity ( ELC ) Enhanced Detection funding . Fully implemented , Orchard Outreach will enable elimination of all order forms for Virology and Microbiology and an antiquated fax / email reporting process . On October 6 , 2021 , a goal was set for enrollment of all clinical laboratories , COVID testing sites , STD clinics , TB clinics , veterinarians and local health departments by January 1 , 2022 . The Outreach Program ( Manager and Program Specialist ) were recruited to coordinate the project . A Project Management Planning ( PMP ) approach was employed to reach the goal of 100 % enrollment in three months for all clients excepting rabies submitters . This approach consisted of eight specific elements : Stakeholder Needs , Project Objectives , Deliverables and Due Dates , Project Schedule , Roles and Responsibilities , Project Budget , Communication Plan , Tracking and Management Tools . In the initial planning , four phases with defined deliverables and due dates were identified : Preparation , Planning , Implementation , Evaluation . By identifying phases , setting priorities and defining roles and responsibilities , the rollout was managed successfully .
Over the course of the three month rollout , the team conducted 12 + live group trainings , 10 + individual facility trainings and onboarded 1,300 + new users representing 91 % of clinical laboratories , 82 % of STD Clinics and 86 % of TB Clinics . Efforts are now underway to enroll the outliers within these client groups . Beginning January 2022 , the team will utilize the PMP model with the assistance of a Rutgers Masters in Health Administration ( MHA ) intern to enroll and train all rabies clients . This project management model appears promising for short term projects that impact overall operations and may be of more widespread value for effecting operational change .
Presenter : Hannah Hergert , New Jersey Public Health and Environmental Laboratories , hannah . hergert @ doh . nj . gov
Data Storage Implementation on SQL Server for COVID-19 Lab Data
A Rahat , M Alexander , A Devito , S Suleiman , J Wang and S Hughes , New York City Department of Health and Mental Hygiene
One major challenge in public health informatics is to integrate data from various sources , which often do not have matching identifiers . The goal was to create an efficient centralized storage implementation for the laboratory data that could be accessed by data scientists and other laboratory informatics personnel with minimal time spent on matching data . Additionally , the data needs to be accurate and up to date . A key focus at NYC Public Health Laboratory ( PHL ) was maintaining updated pangolin lineage results associated with the sequenced samples which needs to be linked to its corresponding metadata . The data we received included COVID-19 patient metadata from the Electronic Clinical Laboratory Reporting System ( ECLRS ) as well as SARS-CoV-2 laboratory sequence data from internal and external laboratories . We automated raw patient metadata input to the SQL Server using python scripts and created SQL queries to filter and merge relevant data depending on the needs of different analysis procedures . The data storage implementation contains global SARS-CoV-2 sequencing data from Global Initiative on Sharing Avian Influenza Data ( GISAID ), data submitted by external laboratories , and local data from NYC PHL and affiliated lab sequencing results . To match this data accurately , we cleaned the data by removing incorrect or duplicate values , then matched based on a combination of sequence ID , accession number , and fuzzy matching using patient identifiers . After matching the data from various sources , data scientists were able to filter relevant data using methods created in the R programming language , bash scripting , and SQL queries . As of January 10 , 2022 , there were 51,689 matched records from NYC PHL and NYC Pandemic Response Laboratory ( PRL ). The pangolin lineage results reported , however , are fixed values at the time of data submission . Using the SQL database we developed , we were able to retrieve the sequences , reanalyze with the latest pangolin tool , link to the clinical records , and share up-to-date lineage results to agency laboratory and epidemiology staff to perform COVID-19 surveillance work . We highlight the advantages of automating the data integration and share the challenges in data matching . The implementation of a centralized storage system on SQL Server has automated data matching and eliminated manual data cleaning while providing all users with the most updated data .
Presenter : Ahmed Rahat , New York City Department of Health and Mental Hygiene , arahat @ health . nyc . gov
A Collaborative Model for Streamlining Public Health Data
J Napoli , S Premutico , O Joshi and E Malaj , Palantir Technologies Inc .
At pandemic outset , the US Government had an urgent need for high-quality national-level data for data-driven decision-making . To minimize burden on already overwhelmed public health institutions , it required a simple approach for organizations to share granular data . The US Government solved these issues by standing up a robust data platform within days , enabling HHS / CDC to make decisions based on up-to-date data . Known as Protect , the solution has offered insight into what is required of a technology platform to transform a decentralized lab network into a coordinated effort to detect public health threats . Protect serves as both a central , lowburden , and flexible reporting mechanism and a secure analytics and decision-making platform . It is backed by open APIs and flexible data sync , pipelining , and protection capabilities . These facilitate the generation of high-quality , usable data without burdening data sharing institutions — and enable rapid integration of new data ( e . g ., new tests ). Impacts include :
• 1,900 + hospital labs can report data that they previously had no means of reporting
• Nearly 7,000 hospitals can report data at the national level , as compared to less than 1,000 hospitals at pandemic outset
• The CDC can widely share transparent , near real-time data on
Informatics