TY - JOUR AU - Dixon, Brian E. AU - Duke, Jon AU - Grannis, Shaun PY - 2017/05/02 Y2 - 2024/03/29 TI - Measuring and Improving the Quality of Data Used for Syndromic Surveillance JF - Online Journal of Public Health Informatics JA - OJPHI VL - 9 IS - 1 SE - Data sources, standards, exchange, visualization, and quality DO - 10.5210/ojphi.v9i1.7623 UR - https://ojphi.org/ojs/index.php/ojphi/article/view/7623 SP - AB - <div style="left: 89.1818px; top: 362.433px; font-size: 14.0379px; font-family: sans-serif; transform: scaleX(1.08186);" data-canvas-width="63.1985303030303">Objective</div><div style="left: 104.045px; top: 377.248px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.99858);" data-canvas-width="377.6961477272726">To extend an open source platform for measuring the quality</div><div style="left: 89.1818px; top: 393.763px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.956432);" data-canvas-width="390.9254446969696">of electronic health data by adding functions useful for syndromic</div><div style="left: 89.1818px; top: 410.279px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.00291);" data-canvas-width="71.34049999999999">surveillance.</div><div style="left: 89.1818px; top: 441.706px; font-size: 14.0379px; font-family: sans-serif; transform: scaleX(1.11811);" data-canvas-width="81.88294696969696">Introduction</div><div style="left: 104.045px; top: 456.521px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.948304);" data-canvas-width="376.523984848485">Nearly all of the myriad activities (or use cases) in clinical and</div><div style="left: 89.1818px; top: 473.036px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.01253);" data-canvas-width="392.82898106060594">public health (e.g., patient care, surveillance, community health</div><div style="left: 89.1818px; top: 489.551px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.98962);" data-canvas-width="389.5637704545454">assessment, policy) involve generating, collecting, storing, analyzing,</div><div style="left: 89.1818px; top: 506.066px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.973009);" data-canvas-width="391.717181060606">or sharing data about individual patients or populations. Effective</div><div style="left: 89.1818px; top: 522.582px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.999993);" data-canvas-width="389.8304901515151">clinical and public health practice in the twenty-first century requires</div><div style="left: 89.1818px; top: 539.097px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.999397);" data-canvas-width="392.51312878787917">access to data from an increasing array of information systems,</div><div style="left: 89.1818px; top: 555.612px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.02895);" data-canvas-width="390.9675583333334">including but not limited to electronic health records. However, the</div><div style="left: 89.1818px; top: 572.127px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.957687);" data-canvas-width="391.50240151515146">quality of data in electronic health record systems can be poor or</div><div style="left: 89.1818px; top: 588.642px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.02967);" data-canvas-width="390.68960833333347">“unfit for use.” Therefore measuring and monitoring data quality is</div><div style="left: 89.1818px; top: 605.157px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.962813);" data-canvas-width="391.22304772727296">an essential activity for clinical and public health professionals as</div><div style="left: 89.1818px; top: 621.672px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.00205);" data-canvas-width="109.52353030303031">well as researchers.</div><div style="left: 89.1818px; top: 653.1px; font-size: 14.0379px; font-family: sans-serif; transform: scaleX(1.07334);" data-canvas-width="57.70971969696969">Methods</div><div style="left: 104.045px; top: 667.915px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.03008);" data-canvas-width="273.79478787878793">Using the Health Data Stewardship Framework</div><div style="left: 377.929px; top: 668.286px; font-size: 8.42273px; font-family: serif;">1</div><div style="left: 382.184px; top: 667.915px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.05231);" data-canvas-width="97.90016666666666">, we will extend</div><div style="left: 89.1818px; top: 684.43px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.991946);" data-canvas-width="391.9179227272726">Automated Characterization of Health Information at Large-scale</div><div style="left: 89.1818px; top: 700.945px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.975252);" data-canvas-width="391.4013287878786">Longitudinal Evidence Systems (ACHILLES), a software package</div><div style="left: 89.1818px; top: 717.46px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.974355);" data-canvas-width="391.3985212121212">published open-source by the Observational Health Data Sciences</div><div style="left: 89.1818px; top: 733.976px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.0253);" data-canvas-width="391.01528712121205">and Informatics collaborative (OHDSI; www.ohdsi.org) to measure</div><div style="left: 89.1818px; top: 750.491px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.0014);" data-canvas-width="389.9624462121212">the quality of data electronically reported from disparate information</div><div style="left: 89.1818px; top: 767.006px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.00067);" data-canvas-width="392.33063636363653">systems. Our extensions will focus on analysis of data reported</div><div style="left: 89.1818px; top: 783.521px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.995844);" data-canvas-width="389.7069568181817">electronically to public health agencies for disease surveillance. Next</div><div style="left: 89.1818px; top: 800.036px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.950913);" data-canvas-width="391.0152871212121">we will apply the ACHILLES extensions to explore the quality of</div><div style="left: 89.1818px; top: 816.551px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.99061);" data-canvas-width="391.6708560606063">data captured from multiple real-world health systems, hospitals,</div><div style="left: 89.1818px; top: 833.066px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.03972);" data-canvas-width="390.81454545454557">laboratories, and clinics. We will further demonstrate the extended</div><div style="left: 89.1818px; top: 849.582px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.969812);" data-canvas-width="391.43221212121205">software to public health professionals, gathering feedback on the</div><div style="left: 89.1818px; top: 866.097px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.975509);" data-canvas-width="391.9109037878791">ability of the methods and software tool to support public health</div><div style="left: 89.1818px; top: 882.612px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.00407);" data-canvas-width="390.03123181818165">agencies’ efforts to routinely monitor the quality of data received for</div><div style="left: 89.1818px; top: 899.127px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.00201);" data-canvas-width="261.1747348484849">surveillance of disease prevalence and burden.</div><div style="left: 89.1818px; top: 930.554px; font-size: 14.0379px; font-family: sans-serif; transform: scaleX(1.08498);" data-canvas-width="50.70481818181818">Results</div><div style="left: 104.045px; top: 945.369px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.985123);" data-canvas-width="377.15428560606045">To date we have mapped key surveillance data fields into the</div><div style="left: 89.1818px; top: 961.885px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.02839);" data-canvas-width="390.8735045454541">OHDSI common data model, and we have transformed 111 million</div><div style="left: 89.1818px; top: 978.4px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.0082);" data-canvas-width="390.2895287878788">syndromic surveillance message segments pertaining to 16.4 million</div><div style="left: 89.1818px; top: 994.915px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.993825);" data-canvas-width="391.89826969697003">emergency department encounters representing 6 million patients</div><div style="left: 89.1818px; top: 1011.43px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.01107);" data-canvas-width="390.3569106060606">for importation into ACHILLES. Using these data, we are exploring</div><div style="left: 89.1818px; top: 1027.95px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.03803);" data-canvas-width="393.5350863636359">the existing 167 metrics across 16 categories available within</div><div style="left: 89.1818px; top: 1044.46px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.970639);" data-canvas-width="391.5557454545454">ACHILLES, including a person (e.g., number of unique persons);</div><div style="left: 89.1818px; top: 1060.98px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.02569);" data-canvas-width="390.82296818181806">and observation period (e.g., Distribution of age at first observation</div><div style="left: 89.1818px; top: 1077.49px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.997896);" data-canvas-width="391.92353787878767">period). Syndromic surveillance (SS), however, is driven largely</div><div style="left: 89.1818px; top: 1094.01px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.00016);" data-canvas-width="392.3292325757575">by monitoring patient stated chief complaints (non-standard free</div><div style="left: 89.1818px; top: 1110.52px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.998495);" data-canvas-width="392.35871212121214">text clinical data) in addition to coded diagnoses. Consequently,</div><div style="left: 89.1818px; top: 1127.04px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.997556);" data-canvas-width="389.77854999999994">ACHILLES must be extended to maximally support use in analyzing</div><div style="left: 89.1818px; top: 1143.55px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.00184);" data-canvas-width="67.07298484848484">SS datasets.</div><div style="left: 89.1818px; top: 1174.98px; font-size: 14.0379px; font-family: sans-serif; transform: scaleX(1.10288);" data-canvas-width="84.24131060606061">Conclusions</div><div style="left: 104.045px; top: 1189.79px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.01499);" data-canvas-width="375.5623901515151">This work remains a work-in-progress. Over the coming year, we</div><div style="left: 89.1818px; top: 1206.31px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.983787);" data-canvas-width="389.4486598484846">will not only explore existing ACHILLES constructs using real-world</div><div style="left: 505.364px; top: 360.733px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.968834);" data-canvas-width="391.4574803030303">public health data but also introduce new functionality to explore</div><div style="left: 505.364px; top: 377.248px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.974312);" data-canvas-width="391.44624999999996">1) patient demographics; 2) facility and location (e.g., emergency</div><div style="left: 505.364px; top: 393.763px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.03254);" data-canvas-width="390.92123333333325">department where care was delivered); and 3) clinical observations</div><div style="left: 505.364px; top: 410.279px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.01122);" data-canvas-width="390.3639295454545">(e.g., chief complaint). The design and methods for examining these</div><div style="left: 505.364px; top: 426.794px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.02489);" data-canvas-width="390.6868007575758">aspects of surveillance data will be included on the poster, and they</div><div style="left: 505.364px; top: 443.309px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.98932);" data-canvas-width="389.5778083333333">will be made freely available for distribution with a future instance of</div><div style="left: 505.364px; top: 459.824px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.03472);" data-canvas-width="391.0391515151514">the ACHILLES software. We ultimately envision these tools being</div><div style="left: 505.364px; top: 476.339px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.00714);" data-canvas-width="390.0986136363636">available for use on platforms such as the CDC’s Biosense – open to</div><div style="left: 505.364px; top: 492.854px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.987374);" data-canvas-width="389.51884924242427">all local and state health agencies as a one-stop portal for surveillance</div><div style="left: 505.364px; top: 509.369px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.029);" data-canvas-width="390.84964015151496">data analysis – or research environments where they can be used to</div><div style="left: 505.364px; top: 525.885px; font-size: 14.0379px; font-family: serif; transform: scaleX(0.973193);" data-canvas-width="391.7424492424241">examine and improve the quality of data output from informatics</div><div style="left: 505.364px; top: 542.4px; font-size: 14.0379px; font-family: serif; transform: scaleX(1.00245);" data-canvas-width="47.967431818181815">systems.</div> ER -