Predictive model for MCA and ICA cerebral blood flow velocities(CBFV)changes following aneurysmal subarachnoid hemorrhage(SAH)
Objectives: Current TCD criteria for VSP is based on CBFV alone but factors such as age and day after SAH also affect CBFV, making the current criteria unreliable. Our objective is to describe MCA and ICA CBFV trend taking into account clinically significant variables. Methods: Patients with the diagnosis of aneurysmal SAH between 1991 and 1999 were identified using a prospective TCD database. Linear and non-linear regression analysis were performed. We included the following variables in the model: 1)age, 2)age2=(Age-38)2, 3)PSAHD:post SAH day, 4)SAHC=0 if SAH day<12 and =SAH-12 if SAH>12 day , 5)CSPASM: clinical spasm (0,1 for absent, present respectively). Results: We identified 399 TCD examinations that included the vessels of interest. Curve fit analysis demonstrated that the quadratic model for both MCA(R2=0.14,P<0.0001) and ICA(R2=0.093,p<0.0001) fit the data best. (1)MCA CBFV Quadratic model equation: Y=52.6+1.3 Age - 0.017 Age2 + 2.65 PSAHD - 5.05 SAHC + 14.5 CSPASM (2)ICA CBFV quadratic model equation: Y=96.37–1.12 Age + 0.007 Age2 + 2.97 PSAHD-8.08 SAHC+ 12.95 CSPASM When controlling for other variables in the model, we report: 1)A daily increase of 2.65 cm/s(MCA)and 2.97 cm/s(ICA) up to day 12, a daily decrease by 2.4 cm/s(MCA) and 5.1 cm/s(ICA)after day 12. 2)Effect of age on MCA CBFV is an increase of 1.3 cm/s per year of age until 38 and a 3.4% decrease after 38 at any given day after SAH. 3)Effect of age on ICA CBFV is a decrease of 1.1 cm/s per year of age until 38 and a 1.4% decrease after 38 at any given day after SAH. 4)When clinical vasospasm is present the CBFV increases by 14.5 cm/s(MCA) and 12.95 cm/s(ICA). Conclusion: our model suggests that the TCD criteria for VSP may be dependent on multiple variables. Prospective evaluation may lead to improved TCD criteria for the diagnosis of VSP.