- 作者：Robert H. Thiele MD ; ; ; rht7w@virginia.edu ; Douglas A. Colquhoun MBChB ; MSc ; James Patrie MSc ; Sarah H. Nie BS ; Julie L. Huffmyer MD
- 关键词：pulse oximetry ; plethysmography ; cardiac output ; stroke volume ; waveform analysis
- 刊名：Journal of Cardiothoracic and Vascular Anesthesia
- 出版年：2011
- 出版时间：December 2011
- 年：2011
- 卷：25
- 期：6
- 页码：1044-1050
- 全文大小：1026 K
Objective
To assess the relation between photoplethysmographically-derived parameters and invasively-determined hemodynamic variables.Design
After induction of anesthesia and placement of a Swan-Ganz CCOmbo catheter, a Nonin OEM III probe was placed on each patient's earlobe. Photoplethysmographic signals were recorded in conjunction with cardiac output. Photoplethysmographic metrics (amplitude of absorbance waveform, maximal slope of absorbance waveform, area under the curve, and width) were calculated offline and compared with invasively determined hemodynamic variables. Subject-specific associations between each dependent and independent variable pair were summarized on a per-subject basis by the nonparametric Spearman rank correlation coefficient. The bias-corrected accelerated bootstrap resampling procedure of Efron and Tibshirani was used to obtain a 95 % confidence interval for the median subject-specific correlation coefficient, and Wilcoxon sign-rank tests were conducted to test the null hypothesis that the median of the subject-specific correlation coefficients were equal to 0.
Setting
University hospital.
Participants
Eighteen patients undergoing coronary artery bypass graft surgery.
Interventions
Placement of a Swan-Ganz CCOmbo catheter and a Nonin OEM III pulse oximetry probe.
Measurements and main results
There was a positive, statistically significant correlation between stroke volume and width (median correlation coefficient, 0.29; confidence interval, 0.01-0.46; p = 0.034). The concordance between changes in stroke volume and changes in width was 53 % . No other correlations achieved statistical significance.
Conclusions
This study was unable to reproduce the results of prior studies. Only stroke volume and photoplethysmographic width were correlated in this study; however, the correlation and concordance (based on analysis of a 4-quadrant plot) were too weak to be clinically useful. Future studies in patients undergoing low-to-moderate risk surgery may result in improved correlations and clinical utility.