Subramanian, Arun

Introduction: Pulmonary hypertension (PH) is an independent marker of surgical morbidity. Right ventricular systolic pressure (RVSP) measured by transthoracic echocardiography (TTE), correlates poorly with pulmonary artery systolic pressure (PASP) measured by right heart catheterization, especially in PH. The correlation between TTE derived RVSP and intraoperative PASP, while under general anesthesia, is not known. In this study, the association between preoperative RVSP and intraoperative PASP was tested. Methods: All consenting adult surgical patients requiring intraoperative pulmonary pressure monitoring in a 2 year period were retrospectively identified. Patients who had TTE estimated RVSP ≥ 40 mmHg, measured within a week prior to the index surgery, were included. RVSP was calculated from tricuspid regurgitant velocity using the modified Bernoulli equation. For intraoperative PASP, the median of the first 10 minutes after PAC placement was used. CVP, arterial systolic pressure, minimum alveolar concentration of anesthetic (MAC) and basic demographic data were also collected. Agreement between RVSP and PASP was tested using the concordance correlation coefficient. Multivariate linear regression and Bland-Altman plot were used to further refine the analysis. Results: 129 patients met inclusion criteria. Median age (IQR) was 73 (62 - 79) years; TTE was performed a median (IQR) of 2 (1 - 3) days prior to surgery; mean (SD) RVSP was 56.1 (±12.4) mmHg and mean (SD) PASP was 42.1 (±14.9) mmHg. Concordance correlation coefficient between RVSP and PASP was only 0.13 (95% CI 0.02 - 0.24). On multivariate linear regression, PASP was significantly associated with RVSP and CVP (p < 0.01), but not anesthetic MAC, age or proximity of measurements. Bland-Altman bias was -13.9 mmHg; 95% limits of agreement were -48.0 to +20.1 mmHg. Conclusion: Preoperative RVSP measured by TTE correlates poorly with intraoperative PASP measured by PAC during general anesthesia.

Objective: The application of computer models in continuous patient activity monitoring using video cameras is complicated by the capture of images of varying qualities due to poor lighting conditions and lower image resolutions. Insufficient literature has assessed the effects of image resolution, color depth, noise level, and low light on the inference of eye opening and closing and body landmarks from digital images. Method: This study systematically assessed the effects of varying image resolutions (from 100 × 100 pixels to 20 × 20 pixels at an interval of 10 pixels), lighting conditions (from 42 to 2 lux with an interval of 2 lux), color-depths (from 16.7 M colors to 8 M, 1 M, 512 K, 216 K, 64 K, 8 K, 1 K, 729, 512, 343, 216, 125, 64, 27, and 8 colors), and noise levels on the accuracy and model performance in eye dimension estimation and body keypoint localization using the Dlib library and OpenPose with images from the Closed Eyes in the Wild and the COCO datasets, as well as photographs of the face captured at different light intensities. Results: The model accuracy and rate of model failure remained acceptable at an image resolution of 60 × 60 pixels, a color depth of 343 colors, a light intensity of 14 lux, and a Gaussian noise level of 4% (i.e., 4% of pixels replaced by Gaussian noise). Conclusions: The Dlib and OpenPose models failed to detect eye dimensions and body keypoints only at low image resolutions, lighting conditions, and color depths. Clinical Impact: Our established baseline threshold values will be useful for future work in the application of computer vision in continuous patient monitoring.

Background: It is generally perceived that sample sizes of randomized clinical trials (RCTs) have increased over the years, particularly in specialties such as cardiology, with a robust evidence base. The aim of this study was to analyze temporal trends in sample sizes of RCTs in cardiology journals compared to other specialties. Methods: Abstracts of RCTs involving humans from PubMed for 1970-2013 were analyzed using a digital search algorithm. Sample sizes of studies were extracted from each abstract. Date of publication and journal name were collected. Journals from several medical subspecialties were selected for comparison, using the journal impact factor as a measure of clinical relevance. Sample sizes of studies in 1990 were compared to 2010 for each of the journals using the Mann-Whitney U test. Graphical comparisons of sample size trends are presented. Results: 272,054 abstracts of human RCTs were identified. Median sample sizes for the years 1990 and 2010 is shown in table 1. The median sample size for all RCTs published in Circulation was 99 subjects per study in 1990, increasing to 630 subjects per study in 2010 (p < 0.01). All cardiology journals had a significant increase in study sample size over the 20 year period, as did the multispecialty journals (JAMA, NEJM, Lancet). In contrast, only a few non-cardiology specialty journals published studies with increasing sample sizes (table 1). Figure 1 shows the sample size trend for 1970-2013. Conclusions: Our study demonstrates a dramatic temporal trend of increasing sample sizes in RCTs in cardiology compared to other specialties. Since sample size is estimated based on the effect size studied, one explanation for this observation is that the more obvious larger effects have been previously elucidated, leaving only smaller associations to be studied. This requires increasing resources, highlighting the importance of alternate study designs and collaborative registries to develop a cost effective evidence base.

Abstract Background Transfusion-associated circulatory overload remains underappreciated in the perioperative environment. The authors aimed to characterize risk factors for perioperative transfusion-associated circulatory overload and better understand its impact on patient-important outcomes. Methods In this case–control study, 163 adults undergoing noncardiac surgery who developed perioperative transfusion-associated circulatory overload were matched with 726 transfused controls who did not develop respiratory complications. Univariate and multivariable logistic regression analyses were used to evaluate potential risk factors for transfusion-associated circulatory overload. The need for postoperative mechanical ventilation, lengths of intensive care unit and hospital stay, and mortality were compared. Results For this cohort, the mean age was 71 yr and 56% were men. Multivariable analysis revealed the following independent predictors of transfusion-associated circulatory overload: emergency surgery, chronic kidney disease, left ventricular dysfunction, previous β-adrenergic receptor antagonist use, isolated fresh frozen plasma transfusion (vs. isolated erythrocyte transfusion), mixed product transfusion (vs. isolated erythrocyte transfusion), and increasing intraoperative fluid administration. Patients who developed transfusion-associated circulatory overload were more likely to require postoperative mechanical ventilation (73 vs. 33%; P < 0.001) and experienced prolonged intensive care unit (11.1 vs. 6.5 days; P < 0.001) and hospital lengths of stay (19.9 vs. 9.6 days; P < 0.001). Survival was significantly reduced (P < 0.001) in transfusion recipients who developed transfusion-associated circulatory overload (1-yr survival 72 vs. 84%). Conclusions Perioperative transfusion-associated circulatory overload was associated with a protracted hospital course and increased mortality. Efforts to minimize the incidence of transfusion-associated circulatory overload should focus on the judicious use of intraoperative blood transfusions and nonsanguineous fluid therapies, particularly in patients with chronic kidney disease, left ventricular dysfunction, chronic β-blocker therapy, and those requiring emergency surgery.

Abstract Background: Transfusion-related acute lung injury (TRALI) is the leading cause of transfusion-related death in the United States; however, it remains poorly characterized in surgical populations. To better inform perioperative transfusion practice, and to help mitigate perioperative TRALI, the authors aimed to better define its epidemiology before and after TRALI mitigation strategies were introduced. Methods: This retrospective cohort study examined outcomes of adult patients undergoing noncardiac surgery with general anesthesia who received intraoperative transfusions during 2004 (n = 1,817) and 2011 (n = 1,562). The demographics and clinical characteristics of transfusion recipients, blood transfusion descriptors, and combined TRALI/possible TRALI incidence rates were evaluated. Univariate analyses were used to compare associations between patient characteristics, transfusion details, and TRALI mitigation strategies with TRALI/possible TRALI incidence rates in a before-and-after study design. Results: The incidence of TRALI/possible TRALI was 1.3% (23 of 1,613) in 2004 versus 1.4% (22 of 1,562) in 2011 (P = 0.72), with comparable overall rates in males versus females (1.4% [23 of 1,613] vs. 1.2% [22 of 1,766]) (P = 0.65). Overall, thoracic (3.0% [4 of 133]), vascular (2.7% [10 of 375]), and transplant surgeries (2.2% [4 of 178]) carried the highest rates of TRALI/possible TRALI. Obstetric and gynecologic surgical patients had no TRALI episodes. TRALI/possible TRALI incidence increased with larger volumes of blood product transfused (P < 0.001). Conclusions: Perioperative TRALI/possible TRALI is more common than previously reported and its risk increases with greater volumes of blood component therapies. No significant reduction in the combined incidence of TRALI/possible TRALI occurred between 2004 and 2011, despite the introduction of TRALI mitigation strategies. Future efforts to identify specific risk factors for TRALI/possible TRALI in surgical populations may reduce the burden of this life-threatening complication.

Abstract Background: Transfusion-associated circulatory overload (TACO) is a leading cause of transfusion-related fatalities, but its incidence and associated patient and transfusion characteristics are poorly understood. To inform surgical transfusion practice and to begin mitigating perioperative TACO, the authors aimed to define its epidemiology. Methods: In this retrospective cohort study, the medical records of adult patients undergoing noncardiac surgery with general anesthesia during 2004 or 2011 and receiving intraoperative transfusions were screened using an electronic algorithm for identification of TACO. Those patients who were screened as high probability for TACO underwent rigorous manual review. Univariate and multivariate analyses evaluated associations between patient and transfusion characteristics with TACO rates in a before-and-after study design. Results: A total of 2,162 and 1,908 patients met study criteria for 2004 and 2011, respectively. The incidence of TACO was 5.5% (119 of 2,162) in 2004 versus 3.0% (57 of 1,908) in 2011 ( P < 0.001), with comparable rates for men (4.8% [98 of 2,023]) and women (3.8% [78 of 2,047]) ( P = 0.09). Overall, vascular (12.1% [60 of 497]), transplant (8.8% [17 of 193]), and thoracic surgeries (7.2% [10 of 138]) carried the highest TACO rates. Obstetric and gynecologic patients had the lowest rate (1.4% [4 of 295]). The incidence of TACO increased with volume transfused, advancing age, and total intraoperative fluid balance (all P < 0.001). Conclusions: The incidence of perioperative TACO is similar to previous estimates in nonsurgical populations. There was a reduction in TACO rate between 2004 and 2011, with incidence patterns remaining comparable in subgroup analyses. Future efforts exploring risk factors for TACO may guide preventive or therapeutic interventions, helping to further mitigate this transfusion complication.