Validation of an algorithm identifying incident primary immune thrombocytopenia in the French national health insurance database.

OBJECTIVES: To evaluate the accuracy of an algorithm identifying newly diagnosed immune thrombocytopenia (ITP) patients in the French national health insurance database (SNIIRAM).

METHODS: The source of data was the SNIIRAM of Midi-Pyrenees region (southwest of France, three million inhabitants). Data of patients with at least one ITP code (D69.3 code of the International Classification of Disease, version 10) were extracted between January 1, 2012, and December 31, 2014. We used an algorithm that identifies newly diagnosed primary ITPs. Medical charts of incident ITPs were reviewed. Positive predictive values (PPVs) of identification of true, incident, and primary ITP cases were estimated.

RESULTS: Of the 168 patients selected, 161 were true ITP cases yielding a PPV of 95.8% (95% confidence interval-95% CI: 92.8-98.8). Among them, 128 were truly incident according to symptom onset date and 134 according to the diagnosis date yielding PPVs of 79.5% (95% CI: 73.2-85.7) and 83.2% (95% CI: 77.4-89.0), respectively. Median time between estimated diagnosis date by the algorithm and true diagnosis date was 0 days (interquartile range: 0 to 15).

CONCLUSIONS: This study showed a very good PPV of this algorithm identifying incident primary ITP patients in the SNIIRAM.