Editorials
presentation. AKI had been defined by RIELE classification using serum creatinine levels and urine output. Consider the observation that serum creatinine can he spuriously depressed in the setting of intravascular volume expansion (precisely the condition where elevated BNP levels are detected) and that perhaps appropriate diuresis on presentation may have resulted in an expected rise of creatinine in the absence of true renal injury. There are several limitations to this study, some of which have been acknowledged by the authors: first, as a single-center study, the universal relevance remains unanswered; second, the absence of serial measurement of BNP prevents us from studying the utility of the trajectory rather than the absolute value of BNP in predicting AKI; third, volume management was not reported in this study. Although appropriate use of diuretics may have explained an elevated serum creatinine not necessarily related to true renal injury, inappropriate administration of TVfiuidsin patients presenting with shock may have precipitated AKI fi-om worsening of cardiorenal syndrome. Indeed, imhalances in the gradient of arterial to venous pressure can significantly impact renal perfusion pressure (10). Lastly, despite the practical use of serum creatinine in defining AKI, there are real limitations in its utility as a surrogate marker of renal function, particularly in the setting of intravascular volume shifts. In conclusion, the results of this article add some interesting findings despite some of its limitations. Like many other earlier biomarkers, BNP may eventually find its role in predicting AKI, and future work studying clinical settings and its utility in coordination with other biomarkers may prove to be fruitful. However, at this time, it would he premature to suggest that
BNP measurement is ready for the spotlight in prime time use with respect to predicting AKI.
REFERENCES 1. Roger VL, Wesfon SA, Gerber Y, et al; Trends in incidence, severity, and outcome of hospifalized myocardial infarcfion. Circulation 2010; 121;863-869 2. Menees DS, Peterson ED, Wang Y, ef al: Door-to-balloon fime and mortality among patients undergoing primary PCI. N EngI J Med 2013;369;901-909 3. Fox CS, Muntner P, Chen AY, et al: Shorf-term outcomes of acufe myocardial infarcfion in patients wifh acufe kidney injury: A reporf from fhe national cardiovascular dafa registry. Circulation 2012; 125:497-504 4. Parikh CR, Coca SG, Wang Y, ef al: Long-ferm prognosis of acufe kidney injury after acufe myocardial infarcfion. Arch Intern Med 2008; 168:987-995 5. Heviiift SM, Dear J, Sfar RA: Discovery of profein biomarkers for renal diseases, i / I m Soc Nephrol 2004; 15;1677-1 689 6. Parikh CR, Devarajan P: Nevi biomarkers of acute kidney injury. Crit Care Med 2008; 36;S159-S165 7 Moltraso M, Cabiafi A, Milazzo V, ef al; B-Type Nafriurefic Peptide and Risk of Acufe Kidney Injury in Patients Hospitalized With Acute Coronary Syndromes. Crit Care Med 2014; 42:619-624 8. Jarai R, Dangas G, Huber K, ef al: B-type nafriuretic pepfide and risk of contrast-induced acufe kidney injury in acute ST-segment-elevation myocardial infarction: A subsfudy from the HORIZCNS-AMI trial. Circ Cardiovasc Interv 201 2; 5:813-820 9. Patel UD, Garg AX, Krumholz HM, ef al; Translafional Research Investigating Biomarker Endpoints in Acufe Kidney Injury (TRIBEAKI) Consorfium; Preoperafive serum brain nafriuretic pepfide and risk of acute kidney injury affer cardiac surgery. Circulation 2012; 125;1347-1355 10. Schrier RW; Fluid administrafion in critically ill patienfs with acufe kidney injury. Clin J Am Soc Nephrol 2010; 5:733-739
Can Administrative Data Be Used to Consistently IVIeasure the Burden of Sepsis?* Olaf L Cremer, MD, PhD Department of Intensive Care Medicine University Medical Center Utrecht Utrecht, The Netherlands
*See also p. 625. Key Words: diagnosis; epidemiology; morfalify; septic shock; severe sepsis Dr. Cremer's instifufion received granf supporf from fhe Cenfer for Translational Molecular Medicine (Projecf MARS [molecular diagnosis and risk stratificafion of sepsis], granf 041-201). Copyright © 2013 by fhe Society of Crifical Care Medicine and Lippincoff Williams & Wilkins DOI: 10.1097/CCM.0000000000000075
Critical Care Medicine
M
ore than a decade has passed since the launch of the first Surviving Sepsis Campaign initiative in 2002, and issued guidelines have now reached their third version (1). In parallel with this collaborative effort to raise awareness and improve the treatment of sepsis by introducing care bundles, published case fatality rates have been slowly declining (2). At the same time, numerous authors have reported that—due to an aging population—the prevalence of sepsis in the United States is rising (2-4). The majority of these reports have used administrative databases to retrospectively assess the prevalence, variations in care, and outcomes of sepsis over time. This approach is attractive, because administrative data are readily available at low cost and are able to span multiple years and healthcare settings. Presumably, administrative data also better reñect real-world treatment than data collected in prospective trials. However, to avoid erroneous conclusions, users must be aware of the vvww.ccmjournal.org
747
Editorials
inherent limitations in using such databases. Because administrative data are derived from claims submitted by clinicians to receive payment, there may be varying incentives to select diagnoses that are associated with highest reimbursement— for instance, to more frequently select severe sepsis in favor of simple pneumonia as the principal diagnosis (5). Alternatively, it is also possible that coding professionals have simply developed an increased awareness of sepsis following the 2002 initiative. As a result, criteria for coding severe sepsis and septic shock may be subject to "inflation." In this issue of Critical Care Medicine, Stevenson et al (6) aimed to validate the use of administrative data to detect temporal changes in sepsis mortality through straightforward comparison with trends derived firom prospective interventional studies. Using the control arms of 36 multicenter randomized clinical trials, they found that 28-day mortality of patients presenting with severe sepsis or septic shock had decreased from 47% during the years 1991-1995 to 29% during the years 2006-2009 (i.e., annual percent change of 3.0%). Subsequently, they used a nationally representative administrative database—the Nationwide Inpatient Sample—and found that hospital mortality of patients who had received International Classification of Diseases, 9th Revision, Clinical Modifications (ICD-9-CM) codes indicating a likely diagnosis of severe sepsis had decreased by a similar amount (i.e., annual percent change of 4.7% and 3.5% using either an "Angus" or "Martin" algorithm, respectively). They conclude that this finding supports the use of administrative data for epidemiological monitoring of severe sepsis. But is mere similarity of trending mortality enough to be appeased? Interventional trials and administrative registries likely represent very different subsamples of the sepsis population. Both may provide valuable outcome data, but they should be considered complimentary rather than alternative sources of information. Whereas trials favor confirmed cases of community-acquired sepsis in patients without complicated comorbidities, ICD-9-CM-based registries include unconfirmed cases in patients who may have been primarily admitted to the hospital for completely unrelated conditions. Despite the use of international consensus definitions (7), severe sepsis and septic shock remain complicated syndromes that are very difficult to accurately categorize (8). Furthermore, even for clinicians at the bedside, it is often not possible to make a definite diagnosis of infection with any confidence (9). For a medical coding professional without access to real-time information, things must be even worse. ICD-9-CM coding criteria for severe sepsis not only depend on the presence of an assured infectious focus but also on the availability of key clinical findings in the medical record, such as body temperature, WBC count, heart rate, respiratory rate, and markers of organ dysfunction (10)—all of which will vary over the course of the hospital admission. In addition, because specific codes for severe sepsis and septic shock were not in existence before 2002, Stevenson et al (6) had to rely on algorithms that cleverly searched for combinations of codes that could mark the presence of both infection 748
w\/vvv.ccnnjournal.org
and acute organ failure. However, this method is not without its problems. Based on the previously reported sensitivity and positive predictive values for the Angus and Martin algorithms that were used, it can be expected that 2-29% of identified patients did not meet criteria for severe sepsis, whereas 50-83% of true cases remained undetected (11). Furthermore, neither algorithm required infection to be recorded as the principal diagnosis. Sepsis may affect patients with a large array of comorbidities. This is particularly true for nosocomial sepsis, which—by definition—develops in patients in whom other acute illnesses or scheduled surgery form the principal reason for hospital admission. Hence, any observed trend in mortality cannot be simply attributed to altered sepsis care. In the absence of effective new drugs for severe sepsis, it is attractive to think that the decreasing case fatality rates that have been observed are driven by greater awareness of the problem, possibly resulting in earlier administration of antibiotics, increased use of early goal-directed resuscitation, and other general improvements in sepsis management. However, we must also consider the possibility that declining mortality may be explained by completely unrelated advances in the care for various underlying pathologies that are frequently present in patients in whom sepsis develops. Indeed, previous studies have demonstrated decreasing trends in risk-adjusted hospital mortality from numerous conditions, including acute myocardial infarction, congestive heart failure, and stroke (12). Although it is encouraging that Stevenson et al (6) found a steady decline of sepsis-associated mortality by using both trial and administrative data, we still do not know whether this applies to all categories of patients, whether it improved sepsis care or advances in the management of underlying diseases that are to be held responsible, or to what extent the observed improvements in hospital and 28-day mortality translate into meaningful long-term outcomes. Numerous studies have suggested that patients who survive to hospital discharge after sepsis remain at increased risk for death in the following months and years (13). Therefore, let us not rest assured. We still need large, ongoing, prospective, dedicated data registries to study unselected sepsis cases and provide sufficient follow-up.
REFERENCES 1. Dellinger RP, Levy MM, Rhodes A, et al; Surviving Sepsis Campaign Guidelines Committee including the Pédiatrie Subgroup: Surviving sepsis campaign: International guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 2013; 41:580-637 2. Gaieski DF, Edwards JM, Kalian MJ, et al: Benchmarking the incidence and mortality of severe sepsis in the United States. Crit Care /Wed 2013; 41:1167-1174 3. Bateman BT, Schmidt U, Berman MF, et al: Temporal trends in the epidemiology of severe postoperative sepsis after elective surgery: A large, nationwide sample. Anesthesiology 2010; 11 2:91 7-925 4. Dombrovskiy VY, Martin AA, Sunderram J, et al: Rapid increase in hospitalization and mortality rates for severe sepsis in the United States: A trend analysis from 1993 to 2003. Crit Care Med 2007; 35:1244-1250 5. Sarrazin MS, Rosenthal GE: Finding pure and simple truths with administrative data. J/A/W/l 2012; 307:1433-1435 March 2014 • Volume 42 • Number 3
Editorials 6. Stevenson EK, Rubenstein AR, Radin GT, et al: Two Decades of Mortality Trends Among Patients With Severe Sepsis: A Comparative Meta-Analysis. Crit Care Med 2014; 42:625-631 7. Levy MM, Fink MP, Marshall JO, et al; SCCM/ESICM/ACCP/ATS/ SIS: 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med 2003; 31:1250-1256 8. Klein Klouwenberg PM, Ong DS, Bonten MJ, et al: Classification of sepsis, severe sepsis and septic shock: The impact of minor variations in data capture and definition of SIRS criteria. Intensive Care Med 201 2; 38:811-819 9. Klein Klouwenberg PMC, Ong DSY, Bos LDJ, et al: Interobserver agreement of centers for disease control and prevention criteria for classifying infections in critically ill patients. Crit Care Med 2013; 41:2373-2378
10. Centers for Disease Control and Prevention: Classification of diseases, functioning, and disability. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) [Internet]. Available at: http:// wviw.cdc.gov/nchs/icd/icd9cm.htm. Accessed September 21, 2013 11. Iwashyna TJ, Odden A, Rohde J, et al: Identifying patients with severe sepsis using administrative claims: Patient-level validation of the angus implementation of the International Consensus Conference Definition of Severe Sepsis. Med Care 2012 Sep 18. [Epub ahead of print] 12. Hines A, Stranges E, Andrews RM: Trends in hospital risk-adjusted mortality for select diagnoses by patient subgroups, 2000-2007 [Internet]. Agency for Healthcare Research and Quality. Available at: http://www.hcup-us.ahrq.gov/reports/statbriefs/sb98.jsp. Accessed September 21, 2013 13. Yende S, Angus DC: Long-term outcomes from sepsis. Curr Infect Dis Rep 2007; 9:382-386
Traumatic Spinal Cord Injury: Learn From the Brain!' Christos Lazaridis, MD Division of Neurocritical Care Department of Neurology Baylor College of Medicine Houston, TX Charles M. Andrews, MD Department of Neurosciences; and Division of Emergency Medicine Department of Medicine Medical College of South Carolina Charleston, SC f'^i
'^he critical care approach to traumatic brain injury I (TBI) has evolved around two central themes. The - i .. first being intracranial pressure (ICP), with a wealth of data associating high pressures with worse neurologic outcomes (1); the second theme is the idea of secondary injury mechanisms and particularly ischemia (2). These pathophysiologic insights gave rise to ICP-guided and cerebral perfusion pressure (CPP)-guided management strategies. Thresholds of ICP/CPP were established based on observational data and uncontrolled series, were incorporated into clinical guidelines, and have been serving as the foundations of clinical care for patients with TBI. Recent research at both a cellular and clinical level has begun to shed more light on these concepts. Combination strategies of controlling ICP with an adequate CPP have gained ground over indiscriminate CPP augmentation
*See also p. 646. Key Words: cerebral perfusion pressure; neurocritical care; pressure reactivity; spinal cord injury; traumatic brain injury The authors have disclosed that they do not have any potential conflicts of interest. Copyright © 2013 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI:10.1097/CCM.0000000000000077
Critical Care Medicine
(3). The actual thresholds for these measures have come under significant scrutiny in view of recent randomized controlled trials, including Decompressive Craniectomy in Diffuse Traumatic Brain Injury (DECRA) and A Trial of Intracranial-Pressure Monitoring in Traumatic Brain Injury (BEST:TRIP) (4,5). The idea of a universal threshold for all patients irrespectively of their demographics, type of brain injury, disease course, and response to treatments will now be questioned. The opposite conceptual framework is to individualize thresholds and treatments in anticipation of optimizing dynamic physiologic processes. The last Brain Trauma Foundation guidelines had suggested this approach proposing the use of tissue oxygénation, metabolism, and pressure autoregulation in "qualifying" CPP goals (6). Nevertheless, the exact use and validity of these techniques have hindered widespread inclusion into clinical protocols and remain to be elucidated. Why have we examined so much about TBI? For one, many of these core concepts have been the basis of care of patients with spinal cord injuries (SCIs). Critical care management of SCI has also focused on pressure goals and ischemia prevention. Preclinical studies show patterns of primary and secondary injury mechanisms and provide reasoning toward surgical decompression (7). Surgery to decompress the spinal cord is highly debated over appropriate timing and the infiuence on outcomes. From a hemodynamic perspective, management has to deal with the potentially serious consequences of sympathectomy seen in cervical and high thoracic lesions. Furthermore, observational data and clinical dictum have driven the concept of spinal cord perfusion pressure (SCPP) management, the recommendation being to maintain the mean arterial pressure (MAP) more than or equal to 90 mm Hg for 5-7 days. Recent guidelines reaffirm this recommendation as optional, since it is only supported by class III evidence (8). In this issue of Critical Care Medicine, Werndle et al (9) take the first step into measuring spinal cord pressure (intraspinal pressure, ISP) at the level of injury and thus monitoring actual SCPP (10). As mentioned, critical care guidelines of SCI focus on MAP management, as SCPP has remained vvvvw.ccmjournal.org
749
Copyright of Critical Care Medicine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
presentation. AKI had been defined by RIELE classification using serum creatinine levels and urine output. Consider the observation that serum creatinine can he spuriously depressed in the setting of intravascular volume expansion (precisely the condition where elevated BNP levels are detected) and that perhaps appropriate diuresis on presentation may have resulted in an expected rise of creatinine in the absence of true renal injury. There are several limitations to this study, some of which have been acknowledged by the authors: first, as a single-center study, the universal relevance remains unanswered; second, the absence of serial measurement of BNP prevents us from studying the utility of the trajectory rather than the absolute value of BNP in predicting AKI; third, volume management was not reported in this study. Although appropriate use of diuretics may have explained an elevated serum creatinine not necessarily related to true renal injury, inappropriate administration of TVfiuidsin patients presenting with shock may have precipitated AKI fi-om worsening of cardiorenal syndrome. Indeed, imhalances in the gradient of arterial to venous pressure can significantly impact renal perfusion pressure (10). Lastly, despite the practical use of serum creatinine in defining AKI, there are real limitations in its utility as a surrogate marker of renal function, particularly in the setting of intravascular volume shifts. In conclusion, the results of this article add some interesting findings despite some of its limitations. Like many other earlier biomarkers, BNP may eventually find its role in predicting AKI, and future work studying clinical settings and its utility in coordination with other biomarkers may prove to be fruitful. However, at this time, it would he premature to suggest that
BNP measurement is ready for the spotlight in prime time use with respect to predicting AKI.
REFERENCES 1. Roger VL, Wesfon SA, Gerber Y, et al; Trends in incidence, severity, and outcome of hospifalized myocardial infarcfion. Circulation 2010; 121;863-869 2. Menees DS, Peterson ED, Wang Y, ef al: Door-to-balloon fime and mortality among patients undergoing primary PCI. N EngI J Med 2013;369;901-909 3. Fox CS, Muntner P, Chen AY, et al: Shorf-term outcomes of acufe myocardial infarcfion in patients wifh acufe kidney injury: A reporf from fhe national cardiovascular dafa registry. Circulation 2012; 125:497-504 4. Parikh CR, Coca SG, Wang Y, ef al: Long-ferm prognosis of acufe kidney injury after acufe myocardial infarcfion. Arch Intern Med 2008; 168:987-995 5. Heviiift SM, Dear J, Sfar RA: Discovery of profein biomarkers for renal diseases, i / I m Soc Nephrol 2004; 15;1677-1 689 6. Parikh CR, Devarajan P: Nevi biomarkers of acute kidney injury. Crit Care Med 2008; 36;S159-S165 7 Moltraso M, Cabiafi A, Milazzo V, ef al; B-Type Nafriurefic Peptide and Risk of Acufe Kidney Injury in Patients Hospitalized With Acute Coronary Syndromes. Crit Care Med 2014; 42:619-624 8. Jarai R, Dangas G, Huber K, ef al: B-type nafriuretic pepfide and risk of contrast-induced acufe kidney injury in acute ST-segment-elevation myocardial infarction: A subsfudy from the HORIZCNS-AMI trial. Circ Cardiovasc Interv 201 2; 5:813-820 9. Patel UD, Garg AX, Krumholz HM, ef al; Translafional Research Investigating Biomarker Endpoints in Acufe Kidney Injury (TRIBEAKI) Consorfium; Preoperafive serum brain nafriuretic pepfide and risk of acute kidney injury affer cardiac surgery. Circulation 2012; 125;1347-1355 10. Schrier RW; Fluid administrafion in critically ill patienfs with acufe kidney injury. Clin J Am Soc Nephrol 2010; 5:733-739
Can Administrative Data Be Used to Consistently IVIeasure the Burden of Sepsis?* Olaf L Cremer, MD, PhD Department of Intensive Care Medicine University Medical Center Utrecht Utrecht, The Netherlands
*See also p. 625. Key Words: diagnosis; epidemiology; morfalify; septic shock; severe sepsis Dr. Cremer's instifufion received granf supporf from fhe Cenfer for Translational Molecular Medicine (Projecf MARS [molecular diagnosis and risk stratificafion of sepsis], granf 041-201). Copyright © 2013 by fhe Society of Crifical Care Medicine and Lippincoff Williams & Wilkins DOI: 10.1097/CCM.0000000000000075
Critical Care Medicine
M
ore than a decade has passed since the launch of the first Surviving Sepsis Campaign initiative in 2002, and issued guidelines have now reached their third version (1). In parallel with this collaborative effort to raise awareness and improve the treatment of sepsis by introducing care bundles, published case fatality rates have been slowly declining (2). At the same time, numerous authors have reported that—due to an aging population—the prevalence of sepsis in the United States is rising (2-4). The majority of these reports have used administrative databases to retrospectively assess the prevalence, variations in care, and outcomes of sepsis over time. This approach is attractive, because administrative data are readily available at low cost and are able to span multiple years and healthcare settings. Presumably, administrative data also better reñect real-world treatment than data collected in prospective trials. However, to avoid erroneous conclusions, users must be aware of the vvww.ccmjournal.org
747
Editorials
inherent limitations in using such databases. Because administrative data are derived from claims submitted by clinicians to receive payment, there may be varying incentives to select diagnoses that are associated with highest reimbursement— for instance, to more frequently select severe sepsis in favor of simple pneumonia as the principal diagnosis (5). Alternatively, it is also possible that coding professionals have simply developed an increased awareness of sepsis following the 2002 initiative. As a result, criteria for coding severe sepsis and septic shock may be subject to "inflation." In this issue of Critical Care Medicine, Stevenson et al (6) aimed to validate the use of administrative data to detect temporal changes in sepsis mortality through straightforward comparison with trends derived firom prospective interventional studies. Using the control arms of 36 multicenter randomized clinical trials, they found that 28-day mortality of patients presenting with severe sepsis or septic shock had decreased from 47% during the years 1991-1995 to 29% during the years 2006-2009 (i.e., annual percent change of 3.0%). Subsequently, they used a nationally representative administrative database—the Nationwide Inpatient Sample—and found that hospital mortality of patients who had received International Classification of Diseases, 9th Revision, Clinical Modifications (ICD-9-CM) codes indicating a likely diagnosis of severe sepsis had decreased by a similar amount (i.e., annual percent change of 4.7% and 3.5% using either an "Angus" or "Martin" algorithm, respectively). They conclude that this finding supports the use of administrative data for epidemiological monitoring of severe sepsis. But is mere similarity of trending mortality enough to be appeased? Interventional trials and administrative registries likely represent very different subsamples of the sepsis population. Both may provide valuable outcome data, but they should be considered complimentary rather than alternative sources of information. Whereas trials favor confirmed cases of community-acquired sepsis in patients without complicated comorbidities, ICD-9-CM-based registries include unconfirmed cases in patients who may have been primarily admitted to the hospital for completely unrelated conditions. Despite the use of international consensus definitions (7), severe sepsis and septic shock remain complicated syndromes that are very difficult to accurately categorize (8). Furthermore, even for clinicians at the bedside, it is often not possible to make a definite diagnosis of infection with any confidence (9). For a medical coding professional without access to real-time information, things must be even worse. ICD-9-CM coding criteria for severe sepsis not only depend on the presence of an assured infectious focus but also on the availability of key clinical findings in the medical record, such as body temperature, WBC count, heart rate, respiratory rate, and markers of organ dysfunction (10)—all of which will vary over the course of the hospital admission. In addition, because specific codes for severe sepsis and septic shock were not in existence before 2002, Stevenson et al (6) had to rely on algorithms that cleverly searched for combinations of codes that could mark the presence of both infection 748
w\/vvv.ccnnjournal.org
and acute organ failure. However, this method is not without its problems. Based on the previously reported sensitivity and positive predictive values for the Angus and Martin algorithms that were used, it can be expected that 2-29% of identified patients did not meet criteria for severe sepsis, whereas 50-83% of true cases remained undetected (11). Furthermore, neither algorithm required infection to be recorded as the principal diagnosis. Sepsis may affect patients with a large array of comorbidities. This is particularly true for nosocomial sepsis, which—by definition—develops in patients in whom other acute illnesses or scheduled surgery form the principal reason for hospital admission. Hence, any observed trend in mortality cannot be simply attributed to altered sepsis care. In the absence of effective new drugs for severe sepsis, it is attractive to think that the decreasing case fatality rates that have been observed are driven by greater awareness of the problem, possibly resulting in earlier administration of antibiotics, increased use of early goal-directed resuscitation, and other general improvements in sepsis management. However, we must also consider the possibility that declining mortality may be explained by completely unrelated advances in the care for various underlying pathologies that are frequently present in patients in whom sepsis develops. Indeed, previous studies have demonstrated decreasing trends in risk-adjusted hospital mortality from numerous conditions, including acute myocardial infarction, congestive heart failure, and stroke (12). Although it is encouraging that Stevenson et al (6) found a steady decline of sepsis-associated mortality by using both trial and administrative data, we still do not know whether this applies to all categories of patients, whether it improved sepsis care or advances in the management of underlying diseases that are to be held responsible, or to what extent the observed improvements in hospital and 28-day mortality translate into meaningful long-term outcomes. Numerous studies have suggested that patients who survive to hospital discharge after sepsis remain at increased risk for death in the following months and years (13). Therefore, let us not rest assured. We still need large, ongoing, prospective, dedicated data registries to study unselected sepsis cases and provide sufficient follow-up.
REFERENCES 1. Dellinger RP, Levy MM, Rhodes A, et al; Surviving Sepsis Campaign Guidelines Committee including the Pédiatrie Subgroup: Surviving sepsis campaign: International guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 2013; 41:580-637 2. Gaieski DF, Edwards JM, Kalian MJ, et al: Benchmarking the incidence and mortality of severe sepsis in the United States. Crit Care /Wed 2013; 41:1167-1174 3. Bateman BT, Schmidt U, Berman MF, et al: Temporal trends in the epidemiology of severe postoperative sepsis after elective surgery: A large, nationwide sample. Anesthesiology 2010; 11 2:91 7-925 4. Dombrovskiy VY, Martin AA, Sunderram J, et al: Rapid increase in hospitalization and mortality rates for severe sepsis in the United States: A trend analysis from 1993 to 2003. Crit Care Med 2007; 35:1244-1250 5. Sarrazin MS, Rosenthal GE: Finding pure and simple truths with administrative data. J/A/W/l 2012; 307:1433-1435 March 2014 • Volume 42 • Number 3
Editorials 6. Stevenson EK, Rubenstein AR, Radin GT, et al: Two Decades of Mortality Trends Among Patients With Severe Sepsis: A Comparative Meta-Analysis. Crit Care Med 2014; 42:625-631 7. Levy MM, Fink MP, Marshall JO, et al; SCCM/ESICM/ACCP/ATS/ SIS: 2001 SCCM/ESICM/ACCP/ATS/SIS International Sepsis Definitions Conference. Crit Care Med 2003; 31:1250-1256 8. Klein Klouwenberg PM, Ong DS, Bonten MJ, et al: Classification of sepsis, severe sepsis and septic shock: The impact of minor variations in data capture and definition of SIRS criteria. Intensive Care Med 201 2; 38:811-819 9. Klein Klouwenberg PMC, Ong DSY, Bos LDJ, et al: Interobserver agreement of centers for disease control and prevention criteria for classifying infections in critically ill patients. Crit Care Med 2013; 41:2373-2378
10. Centers for Disease Control and Prevention: Classification of diseases, functioning, and disability. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) [Internet]. Available at: http:// wviw.cdc.gov/nchs/icd/icd9cm.htm. Accessed September 21, 2013 11. Iwashyna TJ, Odden A, Rohde J, et al: Identifying patients with severe sepsis using administrative claims: Patient-level validation of the angus implementation of the International Consensus Conference Definition of Severe Sepsis. Med Care 2012 Sep 18. [Epub ahead of print] 12. Hines A, Stranges E, Andrews RM: Trends in hospital risk-adjusted mortality for select diagnoses by patient subgroups, 2000-2007 [Internet]. Agency for Healthcare Research and Quality. Available at: http://www.hcup-us.ahrq.gov/reports/statbriefs/sb98.jsp. Accessed September 21, 2013 13. Yende S, Angus DC: Long-term outcomes from sepsis. Curr Infect Dis Rep 2007; 9:382-386
Traumatic Spinal Cord Injury: Learn From the Brain!' Christos Lazaridis, MD Division of Neurocritical Care Department of Neurology Baylor College of Medicine Houston, TX Charles M. Andrews, MD Department of Neurosciences; and Division of Emergency Medicine Department of Medicine Medical College of South Carolina Charleston, SC f'^i
'^he critical care approach to traumatic brain injury I (TBI) has evolved around two central themes. The - i .. first being intracranial pressure (ICP), with a wealth of data associating high pressures with worse neurologic outcomes (1); the second theme is the idea of secondary injury mechanisms and particularly ischemia (2). These pathophysiologic insights gave rise to ICP-guided and cerebral perfusion pressure (CPP)-guided management strategies. Thresholds of ICP/CPP were established based on observational data and uncontrolled series, were incorporated into clinical guidelines, and have been serving as the foundations of clinical care for patients with TBI. Recent research at both a cellular and clinical level has begun to shed more light on these concepts. Combination strategies of controlling ICP with an adequate CPP have gained ground over indiscriminate CPP augmentation
*See also p. 646. Key Words: cerebral perfusion pressure; neurocritical care; pressure reactivity; spinal cord injury; traumatic brain injury The authors have disclosed that they do not have any potential conflicts of interest. Copyright © 2013 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins DOI:10.1097/CCM.0000000000000077
Critical Care Medicine
(3). The actual thresholds for these measures have come under significant scrutiny in view of recent randomized controlled trials, including Decompressive Craniectomy in Diffuse Traumatic Brain Injury (DECRA) and A Trial of Intracranial-Pressure Monitoring in Traumatic Brain Injury (BEST:TRIP) (4,5). The idea of a universal threshold for all patients irrespectively of their demographics, type of brain injury, disease course, and response to treatments will now be questioned. The opposite conceptual framework is to individualize thresholds and treatments in anticipation of optimizing dynamic physiologic processes. The last Brain Trauma Foundation guidelines had suggested this approach proposing the use of tissue oxygénation, metabolism, and pressure autoregulation in "qualifying" CPP goals (6). Nevertheless, the exact use and validity of these techniques have hindered widespread inclusion into clinical protocols and remain to be elucidated. Why have we examined so much about TBI? For one, many of these core concepts have been the basis of care of patients with spinal cord injuries (SCIs). Critical care management of SCI has also focused on pressure goals and ischemia prevention. Preclinical studies show patterns of primary and secondary injury mechanisms and provide reasoning toward surgical decompression (7). Surgery to decompress the spinal cord is highly debated over appropriate timing and the infiuence on outcomes. From a hemodynamic perspective, management has to deal with the potentially serious consequences of sympathectomy seen in cervical and high thoracic lesions. Furthermore, observational data and clinical dictum have driven the concept of spinal cord perfusion pressure (SCPP) management, the recommendation being to maintain the mean arterial pressure (MAP) more than or equal to 90 mm Hg for 5-7 days. Recent guidelines reaffirm this recommendation as optional, since it is only supported by class III evidence (8). In this issue of Critical Care Medicine, Werndle et al (9) take the first step into measuring spinal cord pressure (intraspinal pressure, ISP) at the level of injury and thus monitoring actual SCPP (10). As mentioned, critical care guidelines of SCI focus on MAP management, as SCPP has remained vvvvw.ccmjournal.org
749
Copyright of Critical Care Medicine is the property of Lippincott Williams & Wilkins and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
