Ravi S. Tripathi

The most influential articles in critical care medicine

Abstract Purpose The study aimed to examine query strategies that would provide an exhaustive search method to retrieve the most referenced articles within specific categories of critical care. Material and Methods A comprehensive list of the most cited critical care medicine articles was generated by searching the Science Citation Index Expanded data set using general critical care terms keywords such as “critical care,” critical care journal titles, and keywords for subsubjects of critical care. Results The final database included 1187 articles published between 1905 and 2006. The most cited article was referenced 4909 times. The most productive search term was intensive care. However, this term only retrieved 25% of the top 100 articles. Furthermore, 662 of the top 1000 articles could not be found using any of the basic critical care search terms. Sepsis, acute lung injury, and mechanical ventilation were the most common areas of focus for the articles retrieved. Conclusion Retrieving frequently cited, influential articles in critical care requires using multiple search terms and manuscript sources. Periodic compilations of most cited articles may be useful for critical care practitioners and researches to keep abreast of important information.

Pulse oximetry saturation to fraction inspired oxygen ratio as a measure of hypoxia under general anesthesia and the influence of positive end-expiratory pressure.

PURPOSE In ICU patients with acute lung injury, the pulse oximetry saturation (Spo(2)) to fraction of inspired oxygen (Fio(2)) (S/F) ratio is a reliable surrogate measure for the P/F (Pao(2)/Fio(2)) ratio. Our goal was to determine the correlation of the S/F to the P/F in a large sample of patients undergoing general anesthesia and the influence of positive end-expiratory pressure (PEEP) on this measure. METHODS We studied adult general anesthetics performed with arterial blood gas analysis. Intraoperative data were collected from an anesthesia information system. The S/F ratios corresponding to P/F ratios of 300 were determined. RESULTS A total of 4439 values were collected. Linear correlation between S/F and P/F was identified (r = 0.46; P < .01) with a P/F of 300 corresponding to an S/F of 206. The correlation was stronger in patients with 5 to 9 cm PEEP (r = 0.52; P < .01), more than 9 cm H(2)O PEEP (r = 0.68; P < .01), and a P/F ratio of 300 or less (r = 0.61; P < .01). CONCLUSION The S/F correlates with the P/F in our cohort of patients undergoing general anesthesia, especially those ventilated with PEEP more than 9 cm H(2)O and/or with P/F less than 300. It has use as a noninvasive measure to screen for increased pulmonary dysfunction and to trend oxygenation during a general anesthetic.

Echocardiographic assessment for ventricular assist device placement

While many factors depend on successful implantation and outcome of left ventricular assist devices (LVAD), echocardiography remains an integral part and is vital to the success of this process. Transesophageal echocardiography (TEE) allows interrogation of all the cardiac structures and great vessels. The pre-implantation TEE exam establishes a baseline and may identify potential problems that need palliation. Among these, most significant are aortic insufficiency (AI), intracardiac thrombi, poor right ventricular (RV) function, and intracardiac shunts. The post-implantation exam allows for adequate de-airing of the heart and successful LVAD initiation. The position and flow profiles of the inflow and outflow cannulas of the LVAD may be assessed. Finally, it assists in the astute management and vigilant identification and correction of a number of complications in the immediate post-implantation period. TEE will continue to remain vital to the successful outcomes LVAD patients.

Who truly benefits from extracorporeal membrane oxygenation: a call to develop predictive models.

Extracorporeal membrane oxygenation (ECMO) has matured considerably from its initial development by John H. Gibbon, MD, in Philadelphia, PA, in 1937.1 Since then, the evolution of perfusion technology (e.g., centrifugal flow pumps and polymethylpentene membrane oxygenators) and cardiac critical care medicine2 have allowed circuits to be miniaturized and have enabled their role in higher acuity of critically ill patients to expand without a decriment in outcomes. Mendiratta and colleagues3 from the University of Arkansas have performed an excellent retrospective review of the Extracorporeal Life Support Organization (ELSO) Registry.4 In their review of 99 elderly (>65 years of age) patients who underwent ECMO as part of cardiopulmonary resuscitation (E-CPR), they were able to elucidate two key findings: 1) preexisting renal insufficiency is a harbinger of a poor outcome and 2) while E-CPR outcomes in the elderly are not stellar, they appear to be better than conventional CPR. These results bring the question of which patient can most optimally benefit from this advanced resuscitative technology to the forefront. In the current study, Mendiratta et al.3 were able to demonstrate a 22.2% survival to discharge in the elderly population. Given the inherent comorbidities of this population, the practitioners who enroll data in the ELSO Registry are to be commended. Conventional survival to discharge with E-CPR has been reported to be 27%.5 In evaluating our own data in the E-CPR population at The Ohio State University, we have been able to achieve 57% successful decannulation in the E-CPR setting, with a 29% survival to discharge (unpublished data). From these data, one could crudely predict a 25–30% survival to discharge for E-CPR. This demonstrates room for improvement while patients are supported with ECMO or the need for development of predictive models which would enable more optimum patient selection. We would argue the latter. Other areas of critical care medicine have benefitted from predictive models (e.g., Sepsis-related Organ Failure Assessment).6 Researchers in the ECMO/E-CPR arena have identified risk factors which can foretell a poor outcome, such as age, preexisting renal insufficiency, disease process, and the need for anticoagulation.3,5,7–9 Although these risk factors may aid in our prognostication to some degree, the initiation of ECMO or its continuation does not currently have a beneficial predictive model (such as that potentially provided by a back propagating artificial neural network).10 With an ever-expanding eye on process improvement, quality, and resource utilization that has evolved with the National Surgical Quality Improvement Program, it is imperative that those practitioners with expertise and belief in the benefits of ECMO develop expanded predictive models of efficacy and survival. As these models evolve, we should call for the expansion of the ELSO Registry to include the data elements to enable their validation nationally and internationally and further improve the outcomes being obtained by ECMO practitioners.

Nonconvective forces: a critical and often ignored component in the echocardiographic assessment of transvalvular pressure gradients.

Echocardiography is routinely used to assess ventricular and valvular function, particularly in patients with known or suspected cardiac disease and who have evidence of hemodynamic compromise. A cornerstone to the use of echocardiographic imaging is not only the qualitative assessment, but also the quantitative Doppler-derived velocity characteristics of intracardiac blood flow. While simplified equations, such as the modified Bernoulli equation, are used to estimate intracardiac pressure gradients based upon Doppler velocity data, these modified equations are based upon assumptions of the varying contributions of the different forces that contribute to blood flow. Unfortunately, the assumptions can result in significant miscalculations in determining a gradient if not completely understood or they are misapplied. We briefly summarize the principles of fluid dynamics that are used clinically with some of the inherent limitations of routine broad application of the simplified Bernoulli equation.

“Life-threatening” Hypoxemia in One-lung Ventilation

To the Editor: In January’s ANESTHESIOLOGY, Rozé et al. describe a thorough and step-wise approach to the management of hypoxemia in one-lung ventilation. Although we agree that “lifethreatening” hypoxemia should be treated with resumption of bipulmonary ventilation, we question the definition of life-threatening hypoxemia based only on an arterial oxygen saturation (SpO2) less than 90%. Physiologically, end-organ injury caused by inadequate oxygen delivery (DO2) is dependent on the product of arterial oxygen content (CaO2) and cardiac output. SpO2 is only a single component of CaO2, along with hemoglobin concentration. Thus, in addition to SpO2, we would like to illustrate the importance of considering hemoglobin concentration and cardiac output before aborting one-lung ventilation. We agree that an SpO2 less than 90% may be tolerated poorly in an anemic patient; however, for patients with a normal or high hemoglobin concentration, oxygen content can be maintained at much lower oxygen saturations. It is well known that polycythemia is a compensatory mechanism for hypoxia in people native to high altitudes, with their hemoglobin concentrations being on average 5 g/dl higher than that of their counterparts residing at sea level. Transfusion of erythrocytes is associated with known complications, but increasing hemoglobin concentration via transfusion is associated with decreased work of breathing and minute ventilation in ventilated patients with chronic obstructive pulmonary disease and increased successful weaning from mechanical ventilation in anemic patients with chronic obstructive lung disease. With regard to overall oxygen delivery, cardiac output is another key factor. Although Rozé et al. comment briefly on the development of right ventricular dysfunction with hypoxic pulmonary vasoconstriction, vigilance should be kept to maintaining a normal cardiac output in the face of decreased oxygen saturation. Increased cardiac output is a known compensatory mechanism to hypoxia with additional beneficial effects separate from increased oxygen delivery, such as decreased deadspace ventilation. In addition, although Rozé et al. discuss the use of vasodilators to treat hypoxic pulmonary vasoconstriction, dobutamine has been shown to increase oxygen delivery significantly more than does prostacyclin. Pharmacologic assistance may be needed to maintain a sufficient cardiac output, and patients with low cardiac output states may indeed require oxygen saturations much greater than 90%. Although in a different population than those with onelung ventilation but having similar physiologic principles, recommendations for management of patients with acute respiratory distress syndrome requiring extracorporeal life support include maintenance of an SpO2 greater than 80% and a arterial oxygen concentration of 40 mmHg, provided oxygen content is adequate (hematocrit more than 40%) and cardiac function is not threatened.* We have applied similar principles to patients with acute respiratory distress syndrome without extracorporeal life support. In the event of satisfactory hemoglobin concentrations and cardiac function, we have accepted an SpO2 between 85% and 90% in patients with severe acute respiratory distress syndrome without finding evidence of end-organ malperfusion. In an extreme case, we cared for an 42-yr-old, previously healthy woman with H1N1 infection with superimposed Pseudomonas pneumonia. This patient did not achieve an SpO2 greater than 90% for more than 10 days despite advanced ventilation maneuvers and pharmacologic therapies (including inhaled nitric oxide); however, oxygen delivery was maintained through cardiac output and oxygen content (hemoglobin goal, more than 12 g/dl). The patient recovered from the acute pathophysiology with no long-term end-organ damage or signs of neurologic impairment. A complete discussion of mechanisms to increase organ oxygenation is beyond the scope of this letter. Essentially, vigilant consideration must be given to a myriad of parameters, including peak and plateau airway pressures, tidal volumes, rate of cycle delivery or flow, inspired oxygen concentrations, acid–base, markers of end-organ perfusion, cardiac output or ventricular function, and hemoglobin concentration, rather than choosing an SpO2 of 90% as an arbitrary point of discontinuation of a surgical procedure, sometimes treating “life-threatening” disease.

1198: DEGREE OF GLYCEMIC CONTROL AS A PREDICTOR OF POST-DISCHARGE FACILITY NEEDS IN CARDIAC ICU PATIENTS

Critical Care Medicine • Volume 46 • Number 1 (Supplement) www.ccmjournal.org Learning Objectives: There has been considerable study in regard to the importance of normoglycemia in the intensive care unit (ICU). Studies have shown a reduction in mortality, morbidity, and length of stay with adequate glycemic control in ICU patients. Our study further suggests that a lack of glycemic control in ICU patients may contribute to an increased need for care in a post discharge facility (PDF). Our study also demonstrates that continuous glucose monitoring (CGM) may allow better management of diabetic and/or patients with persistent hyperglycemia in the ICU and thereby temper the need for post ICU care in a PDF. Methods: This prospective study enrolled 87 patients admitted to the Ross Heart Hospital ICUs at The Ohio State University Wexner Medical Center. Inclusion criteria were a diagnosis of type 1 or type 2 diabetes mellitus, or no pre-existing diabetes diagnosis with initial glucose ≥150mg/dL upon admission to the ICU. Each patient was subjected to CGM, which recorded interstitial glucose values every 5 minutes for the first 72 hours of ICU admission. All information recorded in the patient’s electronic medical record was available for subsequent analysis. Normal glycemic control was defined as CGM values between 70-180mg/dL per hospital guidelines. Results: A Chi-squared test was used to evaluate whether poor glycemic control in the first 72 hours of ICU admission was indicative of post ICU PDF needs. Poor glycemic control was defined as patients with greater than 33% of CGM values > 180mg/dL (group1); and then these patients were reevaluated with a less strict definition of poor glycemic control with greater than 25% of CGM values > 180mg/dL (group2). In group1 6/11 (55%) whose glucose was not well controlled as defined went to a PDF as opposed to 12/76 (16%) who were well controlled. In reevaluation as group2, 7/19 (37%) whose glucose was not well controlled as defined went to a PDF as opposed to 11/68 (16%) who were well controlled. The PDF requirement was significantly increased in patients with poor glycemic control (p = 0.003 and p = 0.049 for group1 and group2, respectively). Conclusions: The degree of glycemic control in a cardiac ICU patient population may be a predictor of PDF care needs, and can contribute to increased costs. CGM may better assist providers in managing glycemic control in ICU patients and contributes to improved quality and safety. Maintaining normoglycemic control within this population is of paramount importance.

1079: CONSERVATIVE MANAGEMENT AS OPPOSED TO ECMO FOR MALIGNANT SADDLE PE PROVED TO BE LIFESAVING

Critical Care Medicine • Volume 46 • Number 1 (Supplement) www.ccmjournal.org Learning Objectives: Pulmonary emboli are common during malignancy. For large saddle PE, extracorporeal membrane oxygenation (ECMO) can be a viable support modality. We describe a positive outcome of a challenging clinical case where we chose conservative medical management due to risk of ECMO. Methods: A 46yo female patient presented with new diagnosis of cancer. She had a PMH of myasthenia gravis and hypothyroidism. She was found to have bilateral pleural effusions and mediastinal mass with extensive mass effect. She developed severe shortness of breath, hypoxemia, hypotension and chest CT showed massive bilateral pulmonary emboli involving large branches bilaterally. Emergent catheter-directed thrombolytic therapy was started. Echocardiogram showed right ventricular strain. An urgent evaluation for ECMO was requested. Upon evaluation of her medical problems, physiologic state and imaging; ECMO was deferred out of extremely high risk for adverse outcome. Our efforts focused on right ventricular inotropic support, catheter-directed thrombolytic therapy, systemic anticoagulation and continuous noninvasive positive pressure ventilation to avoid intubation. The patient was eventually weaned from catheter directed thrombolytic therapy and NIPPV, maintained on systemic anticoagulation and discharged to home. Results: This case highlights the importance of clinical integration of complex comorbidities, including myasthenia gravis, mediastinal mass, malignancy and cardiogenic shock into consideration for mechanical support. ECMO can be a viable option although this case emphasizes the need for thorough planning due to potential harms during and after placement of mechanical support. The patient’s mediastinal mass likely has the highest potential for catastrophe due to compression of cardiac structures and the risk of the inability to ventilate or oxygenate. Venovenous versus veno-arterial cannulation strategies were also considered due to the high risk of cardiovascular collapse. Sedation to place mechanical support holds equally high risk; given that loss of spontaneous respiratory effort, acidosis and/or hypoxemia could result in cardiorespiratory collapse. Overlying is the need for chemotherapy and radiation to treat malignancy; which would be obstructed while on ECMO. This unique case highlights the complexities of initiating mechanical support and the critical role that the intensivist can play; to first do no harm.

Serotonin Syndrome in the Intensive Care Unit Following Orthotopic Heart Transplantation: A Case Report

The diagnosis of serotonin syndrome in the ICU population is challenging due to its relative complexity and clinical un- familiarity. Our report is based on a clinical presentation of a 64 year-old female on chronic SSRI therapy who underwent orthotopic heart transplantation (OHT), followed by hyperthermia and autonomic instability. The symptoms were related to the subsequent vasoplegic syndrome requiring methylene blue. The purported offending agents were discontinued and her symptoms improved over the next several hours. Although serotonin syndrome is uncommon and becomes a difficult diagnosis to make, clinicians need to recognize it early.

1248: INTRAVENOUS NITROGLYCERIN FOR DIGITAL ISCHEMIA AND LIMB PRESERVATION IN AN ECLS PATIENT WITH ARDS

balloon pump and cessation of fluid resuscitation. Despite the late recognition of the cardiomyopathy, after multiple rounds of ACLS for PEA the family transitioned the patient to comfort care. Postmortem revealed no evidence of infection. Discussion: Surviving sepsis guidelines advocate aggressive fluid resuscitation with goals set for MAP, CVP, urine output, and MVO2. These guidelines do not account for patients with acute right or left ventricular dysfunction. Early bedside echocardiography may assist in diagnosis in hypotensive patients, avoid unnecessary procedures, and improve outcomes.