Abnormal preprocedural international normalized ratio and platelet counts are not associated with increased bleeding...

The objective of our study was to identify differences in hemorrhagic complications after ultrasound-guided thoracentesis on the basis of patient coagulation parameters.

The records of consecutive patients who underwent ultrasound-guided thoracentesis between January 1, 2008 and April 30, 2010 were reviewed to document the international normalized ratio (INR) and platelet count obtained within 72 hours before thoracentesis and to identify bleeding complications that occurred after the procedure. The observed complication rates and 95% CIs for differences in complication rates were calculated.

There were 1076 procedures performed during the study period with no hemorrhagic complications identified (0% complication rate; 95% CI, 0.00-0.34%). INR values before thoracentesis were available for 822 procedures: INR exceeded 2.0 in 139 cases (17%), 2.5 in 59 cases (7%), and 3.0 in 32 cases (4%). The 95% CI for the 0% difference in complications observed between two groups of patients determined by specific INR values was -0.008 to 0.014 (INR, 1.5), -0.007 to 0.026 (INR, 2.0), -0.007 to 0.061 (INR, 2.5), and -0.009 to 0.11 (INR, 3.0). Platelet values before thoracentesis were available for 953 procedures; the platelet count was less than 100,000/μL for 148 procedures (16%), less than 50,000/μL for 58 procedures (6%), and less than 25,000/μL for 12 procedures (1%). The 95% CI for the 0% difference in complications between two groups of patients determined by a platelet count threshold of 50,000/μL was -0.007 to 0.062.

The risk of bleeding after ultrasound-guided thoracentesis performed by radiologists is low even if the preprocedural INR and platelet count are abnormal. An approach in which no coagulation testing or correction is performed before thoracentesis may be justified.