Improving HIV Surveillance Data by Using the ATra Black Box System to Assist Regional Deduplication Activities

Ocampo, Joanne Michelle F.; Hamp, Auntre; Rhodes, Anne; Smart, J C; Pemmaraju, Raghu; Poschman, Karalee; Hess, Kristen L.; Bhattacharjee, Reshma; Flynn, Colin; Anderson, Bridget J.; Dowling, James E.; Maccormack, Fred; Doshi, Rupali; Lum, Garret; Maddox, Lorene M.; Moncur, Brenda; Barnhart, John; Maxwell, Jason; Aurand, Sahithi Boggavarapu; Hogan, Vicki; Wills, David; Prowell, Stacy; Kassaye, Seble; Karn, Helen E.; Laffoon, Benjamin T.; Collmann, Jeff

doi:10.1097/qai.0000000000002090

Cited by 8 publications

(7 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The ATra Black Box has provided substantial assistance in identifying the current residence and care status of PWDH through deduplication, which, in turn, improves the quality of HIV surveillance data in NHSS. 13…”

Section: Lessons Learnedmentioning

confidence: 99%

Enhancing the ATra Black Box Matching Algorithm: Use of All Names for Deduplication Across Jurisdictions

et al. 2022

Self Cite

View full text Add to dashboard Cite

Objectives: Achieving accurate, timely, and complete HIV surveillance data is complicated in the United States by migration and care seeking across jurisdictional boundaries. To address these issues, public health entities use the ATra Black Box—a secure, electronic, privacy-assuring system developed by Georgetown University—to identify and confirm potential duplicate case records, exchange data, and perform other analytics to improve the quality of data in the Enhanced HIV/AIDS Reporting System (eHARS). We aimed to evaluate the ability of 2 ATra software algorithms to identify potential duplicate case-pairs across 6 jurisdictions for people living with diagnosed HIV. Methods: We implemented 2 matching algorithms for identifying potential duplicate case-pairs in ATra software. The Single Name Matching Algorithm examines only 1 name for a person, whereas the All Names Matching Algorithm examines all names in eHARS for a person. Six public health jurisdictions used the algorithms. We compared outputs for the overall number of potential matches and changes in matching level. Results: The All Names Matching Algorithm found more matches than the Single Name Matching Algorithm and increased levels of match. The All Names Matching Algorithm identified 9070 (4.5%) more duplicate matches than the Single Name Matching Algorithm (n = 198 828) and increased the total number of matches at the exact through high levels by 15.4% (from 167 156 to 192 932; n = 25 776). Conclusions: HIV data quality across multiple jurisdictions can be improved by using all known first and last names of people living with diagnosed HIV that match with eHARS rather than using only 1 first and last name.

show abstract

Section: Lessons Learnedmentioning

confidence: 99%

Enhancing the ATra Black Box Matching Algorithm: Use of All Names for Deduplication Across Jurisdictions

et al. 2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…These algorithms were chosen because they have been recently cited as matching algorithms used to conduct record linkage involving HIV surveillance data. Notably, the Ocampo algorithms have been used by the CDC to match interstate HIV surveillance data [15]. The Stenger algorithm was obtained directly from the PHSKC HIV/STD program, where it has been implemented for several record linkage projects involving HIV surveillance data.…”

Section: Matching Algorithmsmentioning

confidence: 99%

“…Thus, Data to Care strategies that rely entirely on HIV surveillance data involve time-consuming individual case investigations to determine whether persons are truly out of care, although that information is often readily available in other data sources, such as Ryan White-funded care programs, sexually transmitted disease (STD) surveillance, electronic health records, or HIV surveillance systems in other jurisdictions. The Centers for Disease Control and Prevention (CDC) is supporting efforts to match surveillance data between jurisdictions through programs such as the black box system, in which HIV surveillance data from multiple jurisdictions are matched to identify PLWH who have moved from one jurisdiction to another [15,16]. In addition, several health departments are seeking to improve real-time record linkage between HIV and STD surveillance to provide HIV care relinkage services as part of STD partner services [12,17].…”

Section: Introductionmentioning

confidence: 99%

Comparing Methods for Record Linkage for Public Health Action: Matching Algorithm Validation Study

Avoundjian¹,

Dombrowski²,

Golden³

et al. 2020

JMIR Public Health Surveill

View full text Add to dashboard Cite

Background Many public health departments use record linkage between surveillance data and external data sources to inform public health interventions. However, little guidance is available to inform these activities, and many health departments rely on deterministic algorithms that may miss many true matches. In the context of public health action, these missed matches lead to missed opportunities to deliver interventions and may exacerbate existing health inequities. Objective This study aimed to compare the performance of record linkage algorithms commonly used in public health practice. Methods We compared five deterministic (exact, Stenger, Ocampo 1, Ocampo 2, and Bosh) and two probabilistic record linkage algorithms (fastLink and beta record linkage [BRL]) using simulations and a real-world scenario. We simulated pairs of datasets with varying numbers of errors per record and the number of matching records between the two datasets (ie, overlap). We matched the datasets using each algorithm and calculated their recall (ie, sensitivity, the proportion of true matches identified by the algorithm) and precision (ie, positive predictive value, the proportion of matches identified by the algorithm that were true matches). We estimated the average computation time by performing a match with each algorithm 20 times while varying the size of the datasets being matched. In a real-world scenario, HIV and sexually transmitted disease surveillance data from King County, Washington, were matched to identify people living with HIV who had a syphilis diagnosis in 2017. We calculated the recall and precision of each algorithm compared with a composite standard based on the agreement in matching decisions across all the algorithms and manual review. Results In simulations, BRL and fastLink maintained a high recall at nearly all data quality levels, while being comparable with deterministic algorithms in terms of precision. Deterministic algorithms typically failed to identify matches in scenarios with low data quality. All the deterministic algorithms had a shorter average computation time than the probabilistic algorithms. BRL had the slowest overall computation time (14 min when both datasets contained 2000 records). In the real-world scenario, BRL had the lowest trade-off between recall (309/309, 100.0%) and precision (309/312, 99.0%). Conclusions Probabilistic record linkage algorithms maximize the number of true matches identified, reducing gaps in the coverage of interventions and maximizing the reach of public health action.

show abstract

“…Our multidisciplinary team has implemented systems using Georgetown University’s AvesTerra framework for privacy-assured technology for HIV surveillance [ 21 - 23 ]. We then sought to design a user-friendly system to efficiently track symptoms associated with COVID-19 infection to complement existing and evolving contact tracing approaches.…”

Section: Introductionmentioning

confidence: 99%

Rapid Deployment of a Free, Privacy-Assured COVID-19 Symptom Tracker for Public Safety During Reopening: System Development and Feasibility Study

Kassaye¹,

Spence²,

Lau³

et al. 2020

JMIR Public Health Surveill

View full text Add to dashboard Cite

Background Since the emergence of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the number of cases of coronavirus disease (COVID-19) in the United States has exponentially increased. Identifying and monitoring individuals with COVID-19 and individuals who have been exposed to the disease is critical to prevent transmission. Traditional contact tracing mechanisms are not structured on the scale needed to address this pandemic. As businesses reopen, institutions and agencies not traditionally engaged in disease prevention are being tasked with ensuring public safety. Systems to support organizations facing these new challenges are critically needed. Most currently available symptom trackers use a direct-to-consumer approach and use personal identifiers, which raises privacy concerns. Objective Our aim was to develop a monitoring and reporting system for COVID-19 to support institutions conducting monitoring activities without compromising privacy. Methods Our multidisciplinary team designed a symptom tracking system after consultation with experts. The system was designed in the Georgetown University AvesTerra knowledge management environment, which supports data integration and synthesis to identify actionable events and maintain privacy. We conducted a beta test for functionality among consenting Georgetown University medical students. Results The symptom tracker system was designed based on guiding principles developed during peer consultations. Institutions are provided access to the system through an efficient onboarding process that uses clickwrap technology to document agreement to limited terms of use to rapidly enable free access. Institutions provide their constituents with a unique identifier to enter data through a web-based user interface to collect vetted symptoms as well as clinical and epidemiologic data. The website also provides individuals with educational information through links to the COVID-19 prevention recommendations from the US Centers for Disease Control and Prevention. Safety features include instructions for people with new or worsening symptoms to seek care. No personal identifiers are collected in the system. The reporter mechanism safeguards data access so that institutions can only access their own data, and it provides institutions with on-demand access to the data entered by their constituents, organized in summary reports that highlight actionable data. Development of the system began on March 15, 2020, and it was launched on March 20, 2020. In the beta test, 48 Georgetown University School of Medicine students or their social contacts entered data into the system from March 31 to April 5, 2020. One of the 48 users (2%) reported active COVID-19 infection and had no symptoms by the end of the monitoring period. No other participants reported symptoms. Only data with the unique entity identifier for our beta test were generated in our summary reports. Conclusions This system harnesses insights into privacy and data sharing to avoid regulatory and legal hurdles to rapid adaption by entities tasked with maintaining public safety. Our pilot study demonstrated feasibility and ease of use. Refinements based on feedback from early adapters included release of a Spanish language version. These systems provide technological advances to complement the traditional contact tracing and digital tracing applications being implemented to limit SARS-CoV-2 transmission during reopening.

show abstract

Improving HIV Surveillance Data by Using the ATra Black Box System to Assist Regional Deduplication Activities

Cited by 8 publications

References 5 publications

Enhancing the ATra Black Box Matching Algorithm: Use of All Names for Deduplication Across Jurisdictions

Enhancing the ATra Black Box Matching Algorithm: Use of All Names for Deduplication Across Jurisdictions

Comparing Methods for Record Linkage for Public Health Action: Matching Algorithm Validation Study

Rapid Deployment of a Free, Privacy-Assured COVID-19 Symptom Tracker for Public Safety During Reopening: System Development and Feasibility Study

Contact Info

Product

Resources

About