Is measurement of central venous pressure required to estimate systemic vascular resistance? A retrospective cohort study

BACKGROUND: The clinical range of central venous pressure (CVP) (typically 5 to 15 mmHg) is much less than the range of mean arterial blood pressure (60 to 120 mmHg), suggesting that CVP may have little impact on estimation of systemic vascular resistance (SVR). The accuracy and feasibility of using an arbitrary CVP rather than actual CVP for the estimation of SVR during intraoperative period is not known. METHODS: Using vital records obtained from patients who underwent neurological and cardiac surgery, the present study retrospectively calculated SVR using fixed values of CVP (0, 5, 10, 15, and 20 mmHg) and randomly changing values of CVP (5 to 15 mmHg) and compared these calculated SVRs with actual SVR, calculated using actual CVP. Differences between actual SVR and SVRs based on fixed and random CVPs were quantified as root mean square error (RMSE) and mean absolute percentage error (MAPE). Bland-Altman analysis and four-quadrant plot analysis were performed. RESULTS: A total of 34 patients are included, including 18 who underwent neurosurgery and 16 who underwent cardiac surgery; 501,380 s (139.3 h) of data was analyzed. The SVR derived from a fixed CVP of 10 mmHg (SVRf10) showed the highest accuracy (RMSE: 115 and 104 [dynes/sec/cm(− 5)] and MAPE: 6.3 and 5.7% in neurological and cardiac surgery, respectively). The 95% limits of agreement between SVRf10 and actual SVR were − 208.5 (95% confidence interval [CI], − 306.3 to − 148.1) and 242.2 (95% CI, 181.8 to 340.0) dynes/sec/cm(− 5) in neurosurgery and − 268.1 (95% CI, − 367.5 to − 207.7) and 163.2 (95% CI, 102.9 to 262.6) dynes/sec/cm(− 5) in cardiac surgery. All the SVRs derived from the fixed CVPs (regardless of its absolute value) showed excellent trending ability (concordance rate > 0.99). CONCLUSIONS: SVR can be estimated from a fixed value of CVP without causing significant deviation or a loss of trending ability. However, caution is needed when using point estimates of SVR when the actual CVP is expected to be out of the typical clinical range. TRIAL REGISTRATION: This study was registered Clinical Research Information Service, a clinical trial registry in South Korea (KCT0006187).