ORIGINAL ARTICLE Electronic health record surveillance algorithms facilitate the detection of transfusion-related pulmonary complications Leanne Clifford, Amandeep Singh, Gregory A. Wilson, Pearl Toy, Ognjen Gajic, Michael Malinchoc, Vitaly Herasevich, Jyotishman Pathak, and Daryl J. Kor BACKGROUND: Transfusion-related acute lung injury (TRALI) and transfusion-associated circulatory overload (TACO) are leading causes of transfusion-related mor- tality. Notably, poor syndrome recognition and underre- porting likely result in an underestimate of their true attributable burden. We aimed to develop accurate electronic health record–based screening algorithms for improved detection of TRALI/transfused acute lung injury (ALI) and TACO. STUDY DESIGN AND METHODS: This was a retro- spective observational study. The study cohort, identi- fied from a previous National Institutes of Health– sponsored prospective investigation, included 223 transfused patients with TRALI, transfused ALI, TACO, or complication-free controls. Optimal case detection algorithms were identified using classification and regression tree (CART) analyses. Algorithm perfor- mance was evaluated with sensitivities, specificities, likelihood ratios, and overall misclassification rates. RESULTS: For TRALI/transfused ALI detection, CART analysis achieved a sensitivity and specificity of 83.9% (95% confidence interval [CI], 74.4%-90.4%) and 89.7% (95% CI, 80.3%-95.2%), respectively. For TACO, the sensitivity and specificity were 86.5% (95% CI, 73.6%- 94.0%) and 92.3% (95% CI, 83.4%-96.8%), respec- tively. Reduced PaO2/FiO2 ratios and the acquisition of posttransfusion chest radiographs were the primary determinants of case versus control status for both syn- dromes. Of true-positive cases identified using the screening algorithms (TRALI/transfused ALI, n = 78; TACO, n = 45), only 11 (14.1%) and five (11.1%) were reported to the blood bank by physicians, respectively. CONCLUSIONS: Electronic screening algorithms have shown good sensitivity and specificity for identifying patients with TRALI/transfused ALI and TACO at our institution. This supports the notion that active elec- tronic surveillance may improve case identification, thereby providing a more accurate understanding of TRALI/transfused ALI and TACO epidemiology. B lood product transfusions have long been rec- ognized as an important risk factor for acute lung injury (ALI), with transfusion-related acute lung injury (TRALI) consistently accounting for the greatest number of transfusion-related fatalities in the developed world. 1,2 More recently, a second transfusion- related respiratory complication termed transfusion- associated circulatory overload (TACO) has surpassed ABBREVIATIONS: ABG = arterial blood gas; ALI = acute lung injury; BNP = B-type natriuretic peptide; CART = classification and regression tree; CVP = central venous pressure; CXR = chest radiograph; EHR(s) = electronic health record(s); HR = heart rate; ICU = intensive care unit; NLR = negative likelihood ratio; PCWP = pulmonary capillary wedge pressure; P : F ratio = par- tial pressure of arterial oxygen to fraction of inspired oxygen ratio; PLR = positive likelihood ratio; RR = respiratory rate; SBP = systolic blood pressure; TACO = transfusion-associated circula- tory overload; TRALI = transfusion-related acute lung injury. From the Mayo Clinic, Rochester, Minnesota; and the University of California at San Francisco, San Francisco, California. Address reprint requests to: Daryl J. Kor, MD, Department of Anesthesiology, Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905; e-mail: kor.daryl@mayo.edu. Supported by the Department of Anesthesia, Mayo Clinic, Rochester, MN. Attribution for work: LC—study design, data acquisition, data analysis, data interpretation, and manuscript preparation; AS—data acquisition and manuscript revision; GAW—data acquisition and manuscript revision; PT—data acquisition, interpretation, and manuscript revision; OG—data acquisition, data interpretation, and manuscript revision; MM—data analy- sis, data interpretation, and manuscript revision; VH—data acquisition and manuscript revision; JP—data acquisition and manuscript revision; DJL—study conception, study design, data analysis, data interpretation, and manuscript revision. There are no disclaimers. Received for publication May 23, 2012; revision received July 9, 2012, and accepted July 19, 2012. doi: 10.1111/j.1537-2995.2012.03886.x TRANSFUSION **;**:**-**. Volume **, ** ** TRANSFUSION 1