Isolated subsegmental pulmonary embolism identification based on international classification of diseases (ICD)-10 codes and imaging reports

Abstract

Background

Isolated subsegmental pulmonary embolism (issPE) is a commonly encountered diagnosis. Although the International Classification of Diseases (ICD)-10 codes are used for research, their validity for identifying issPE is unknown. Moreover, issPE diagnosis is challenging, and the findings from radiology reports may conflict with those from expert radiologists.

Methods

Based on prespecified criteria, 1734 medical records of adult patients hospitalized within the Mass General Brigham health system (2016–2021) were selected in three equal groups: (1) patients with principal discharge diagnosis codes for PE, (2) patients with secondary discharge diagnosis codes for PE, and (3) patients with no PE codes. The accuracy of ICD-10 codes for issPE was verified by two independent physicians and weighted by total hospitalizations. In a randomly selected sample of 70 patients, the accuracy of initial radiology reports was determined through a blinded re-evaluation by two expert radiologists.

Results

In weighted estimates, ICD-10 codes in primary or secondary discharge positions, compared with chart reviews, showed a low sensitivity (7.0 %) and positive predictive value (25.2 %). Evaluation by two expert radiologists noted that initial radiology reports were sensitive (97.1 %) for issPE but had a low specificity (40.0 %). Two (3.6 %) out of 55 patients with initial issPE reports did not have PE, while 19 (34.5 %) had more proximal PE.

Conclusions

ICD-10 codes for issPE have poor sensitivity and positive predictive value and should not be used for research or quality improvement. Radiology reports for issPE may be inaccurate regarding the location or, less often, the presence of PE.