Validation of the PLASMIC score for predicting ADAMTS13 activity

Abstract

BackgroundThrombotic thrombocytopenic purpura (TTP) is a life-threatening thrombotic microangiopathy (TMA) that requires prompt plasma exchange. Clinical prediction tools may facilitate decision-making in institutions with delayed turnaround time or limited access to ADAMTS13 assays. The PLASMIC score and Bentley score have been shown to predict severe ADAMTS13 deficiency with excellent sensitivity and specificity. ObjectivesTo validate the PLASMIC score using a population of suspected TTP, and evaluate its discriminatory power in predicting severe ADAMTS13 deficiency in comparison with Bentley score and clinical gestalt. MethodsAdults presenting with suspected TTP in Alberta, Canada between 2008 and 2018 with available ADAMTS13 results were included. The sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were calculated for PLASMIC score, Bentley score and clinical gestalt. Receiver operator characteristics analysis assessed the performance of the scoring systems. ResultsAmong 163 individuals with suspected TTP, ADAMTS13 activity was available in 117 (72%). Severe ADAMTS13 deficiency <10% was present in 62 (53%). High-risk PLASMIC score (≥6) predicted severe ADAMTS13 deficiency with a sensitivity of 81.7%, specificity 71.4%, PPV 75.4% and NPV 78.4% (c-statistic 0.80). Intermediate-high risk Bentley score (≥20) had a lower sensitivity (59.5%) and higher specificity (93.9%) with similar c-statistic (0.77). Clinical gestalt had similar sensitivity as PLASMIC score but very low specificity (16.1%). ConclusionsBoth PLASMIC and Bentley scores had good discriminatory power in identifying severe ADAMTS13 deficiency in a Canadian TMA population compared to clinical gestalt. Integration into institutional clinical pathways may help supplement clinical judgment and reduce costs.