correspondence
As shown in Figure S2B and S2C in the Supplementary Appendix of the article (available at NEJM.org), there was no clear relation between glucose levels and the depth of ketoacidosis. Glucose levels actually were within the normal range (>4 mmol per liter) during most documented events. Ketoacidotic events were generally associated with reduced, but not absent, caloric intake, and events could be stopped only by supplying ample amounts of carbohydrates, in excess of the basal energy requirement. Our observations are in line with those in several other inborn errors of intermediary metabolism: metabolic derangements occur if — and are maintained as long as — the energy requirements are not met.1 The energy requirement is known to be elevated during infections and to decrease with age,1 providing an explanation for both risk factors. We believe that our data are consistent with the concept that the metabolic derangement (i.e., ketoacidosis) is maintained as long as the body senses a low cellular energy status in peripheral tissues, regardless of the plasma glucose values. Under these low-insulin conditions, glucagon levels probably reflect a low cellular energy status. In patients with low insulin levels owing to diabetes mellitus, glucagon excretion correlates with the depth of ketoacidosis,2 and ketoacidosis can be prevented by blocking the excretion of glucagon.3 Schlegel suggests that the profound ketoacidosis observed in our patients may be an indirect effect of disturbances in the production,
rather than utilization, of ketones. Following that line of reasoning, the loss-of-function mutations found in our patients would be expected to result in decreased import of substrates and, thus, decreased ketone formation, This would, in turn, lead to hypoketosis on fasting, not hyperketosis — a situation similar to the biochemical findings in ketogenesis defects and, perhaps, in a defect of ketone export. Even if the mutations would somehow lead to increased import, given the endocrine control of ketogenesis, that would have mild-to-moderate effects on ketones, at best. Peter M. van Hasselt, M.D., Ph.D. University Medical Center Utrecht, the Netherlands
Sacha Ferdinandusse, Ph.D. Academic Medical Center Amsterdam, the Netherlands
Gijs van Haaften, Ph.D. University Medical Center Utrecht, the Netherlands [email protected] Since publication of their article, the authors report no further potential conflict of interest. 1. Baumgartner MR, Hörster F, Dionisi-Vici C, et al. Proposed
guidelines for the diagnosis and management of methylmalonic and propionic acidemia. Orphanet J Rare Dis 2014;9:130. 2. Liljenquist JE, Bomboy JD, Lewis SB, et al. Effects of glucagon on lipolysis and ketogenesis in normal and diabetic men. J Clin Invest 1974;53:190-7. 3. Gerich JE, Lorenzi M, Bier DM, et al. Prevention of human diabetic ketoacidosis by somatostatin: evidence for an essential role of glucagon. N Engl J Med 1975;292:985-9. DOI: 10.1056/NEJMc1415111
Management of Withdrawal Delirium (Delirium Tremens) To the Editor: The review by Schuckit (Nov. 27 issue)1 of the management of delirium tremens offers the reader a concise lesson in pharmacologic treatment of this medical problem. I would caution those who refer to the lorazepam regimens in Table 3 of the article that the doses outlined may provide an amount of propylene glycol in excess of 25 mg per kilogram per day, the amount considered safe by the World Health Organization.2 Lorazepam injection contains only small amounts of propylene glycol, yet when given at high doses continuously or for prolonged periods, the patient exposure may be substantial.
Metabolic acidosis may develop owing to propylene glycol accumulation, particularly in patients with preexisting renal or hepatic insufficiency. Several investigators have determined that high doses of lorazepam (i.e., >0.1 mg per kilogram per hour), particularly when given as continuous infusions for treatment of alcohol withdrawal for periods of 24 to 48 hours, are associated with propylene glycol levels of more than 25 mg per deciliter.3-5 Those of us who treat alcohol withdrawal should remain alert to the possibility of this toxicity when administering high-dose lorazepam.
n engl j med 372;6 nejm.org february 5, 2015
579
The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on February 9, 2015. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
The
n e w e ng l a n d j o u r na l
of
m e dic i n e
calm, awake patient, facilitating extubation and transfer out of the ICU, where a smooth transiSpartanburg, SC tion to oral benzodiazepines was possible. Blood No potential conflict of interest relevant to this letter was re- alcohol levels, even as ethanol was administered ported. in one arm, were undetectable in samples drawn 1. Schuckit MA. Recognition and management of withdrawal from the other, repeatedly, in all cases. I specudelirium (delirium tremens). N Engl J Med 2014;371:2109-13. 2. Public health statement for propylene glycol. Atlanta: Agency late that the unique sensitivity of ethanol recepfor Toxic Substances and Disease Registry (http://www.atsdr.cdc tors to ethanol, relative to benzodiazepines, in .gov/phs/phs.asp?id=1120&tid=240). preventing withdrawal symptoms permitted this 3. Arroliga AC, Shehab N, McCarthy K, Gonzales JP. Relationship of continuous infusion lorazepam to serum propylene gly- result. It was not necessary to have high alcohol col concentration in critically ill adults. Crit Care Med 2004;32: levels, or even measurable alcohol levels, to pre1709-14. vent withdrawal symptoms. 4. Nelsen JL, Haas CE, Habtemariam B, et al. A prospective Regrettably, concerns about “giving alcohol evaluation of propylene glycol clearance and accumulation during continuous-infusion lorazepam in critically ill patients. J Inten- to alcoholics” have limited more extensive apsive Care Med 2008;23:184-94. plication of this therapeutic approach. The ap5. Horinek EL, Kiser TH, Fish DN, MacLaren R. Propylene glycol accumulation in critically ill patients receiving continuous parent condoning of alcohol intake and the intravenous lorazepam infusions. Ann Pharmacother 2009;43: long-term risks of high doses of alcohol have 1964-71. both been cited2 as reasons why alcohol should DOI: 10.1056/NEJMc1415679 not be used in alcohol withdrawal in general; yet in specific, monitored settings, it is a safe, effecTo the Editor: In his discussion of delirium tre- tive, and comfortable treatment option.1 mens, Schuckit lists grand mal convulsions as a Carmel Armon, M.D., M.H.S. symptom of alcohol withdrawal. After an analy- Tel Aviv University Sackler School of Medicine sis of 308 patients with first-time seizure in the Tel Aviv, Israel Harlem Hospital emergency department and 294 [email protected] No potential conflict of interest relevant to this letter was recontrols, all of whom were admitted to the hos- ported. pital, Ng and colleagues concluded that “Alcohol 1. Rosenbaum M, McCarty T. Alcohol prescription by surgeons withdrawal was not associated with the onset of in the prevention and treatment of delirium tremens: historic seizures in this study; 16 percent of first seizures and current practice. Gen Hosp Psychiatry 2002;24:257-9. in drinkers fell outside the conventionally de- 2. Burns MJ. Delirium tremens (DTs) treatment and management. Medscape. August 2014 (http://emedicine.medscape.com/ fined withdrawal period, and the remainder ex- article/166032-treatment). hibited a seemingly random timing after the last DOI: 10.1056/NEJMc1415679 drink.” What is more, “[o]f the 184 seizures [with good data on abstention time], 61.4 percent oc- The Author Replies: I am grateful to Moriarty, curred within 24 hours of the last drink.”1 Finucane, and Armon for their comments, as well as their interest in the important but relaThomas E. Finucane, M.D. tively poorly studied topic of withdrawal delir Johns Hopkins Institute of Medicine Baltimore, MD ium. The length restrictions for Journal brief [email protected] reviews are important, because they allow interNo potential conflict of interest relevant to this letter was reested readers to quickly acquire useful informaported. tion, but the inherent efficiency of those com1. Ng SK, Hauser WA, Brust JC, Susser M. Alcohol consumption and withdrawal in new-onset seizures. N Engl J Med 1988; munications do not allow for a full discussion of 319:666-73. all interesting elements related to the topic. The DOI: 10.1056/NEJMc1415679 issues raised in these letters enable me to briefly comment on three important topics that might To the Editor: An additional therapeutic option be of interest to clinicians who work with these for patients with a severe acute alcohol with- seriously ill patients. drawal syndrome (delirium tremens) is an intraMoriarty reminds us of the serious acidosis venous alcohol drip.1 In rare instances, I have that can develop from prolonged infusions of administered 10% ethanol at a rate of 50 to 100 ml lorazepam as a consequence of the suspension per hour, in an intensive care unit (ICU), to agi- of this benzodiazepine in propylene glycol. tated, intubated patients requiring high, sedat- Similar to what can be observed with ethylene ing doses of benzodiazepines. This resulted in a glycol and salicylate poisonings, the acid–base Veronica P. Moriarty, Pharm.D. Spartanburg Medical Center
580
n engl j med 372;6 nejm.org february 5, 2015
The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on February 9, 2015. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
correspondence
imbalance can complicate treatment of alcohol withdrawal in persons receiving high rates of infusion of lorazepam, especially in the context of renal dysfunction. This is noteworthy for any clinician treating critically ill patients who require intravenous sedatives. Finucane cites a 1988 article that emphasized that the risk of grand mal convulsions increases with higher quantities of alcohol consumed and that alcohol-related seizures can be observed in the absence of obvious withdrawal symptoms. This is an important reminder of the central role that heavy drinking plays in seizure risk and underscores the need for clinicians to consider heavy alcohol intake in the differential diagnosis of convulsions, even in the absence of an alcohol withdrawal syndrome.
Finally, Armon’s letter is consistent with the conclusion that the treatment of alcohol withdrawal syndromes can include any brain-depressant drug, including alcohol. However, it may be a bit more difficult to establish the appropriate dose of alcohol, as compared with other sedative-hypnotic agents. For mild withdrawal, alcohol may be associated with greater respiratory depression and potentially more intoxicationrelated aggression than benzodiazepines. Marc A. Schuckit, M.D. University of California, San Diego
La Jolla, CA [email protected] Since publication of his article, the author reports no further potential conflict of interest. DOI: 10.1056/NEJMc1415679
Case 36-2014: A Woman with Fever, Pharyngitis, and Double Vision To the Editor: We would like to raise an issue about the Case Record by Olson et al. (Nov. 20 issue).1 Antibiotic resistance is a real problem in the 21st century. In this context, there is consensus about the importance of prudent use of antibiotic agents and the implementation of antibiotic stewardship in health care facilities. We think it is debatable whether vancomycin, mero penem, and clindamycin were the right choices for the treatment of an infection with a micro organism that can penetrate the central nervous system. Clindamycin has shown poor penetration into the central nervous system in adults but has not, to our knowledge, been studied in children.2 Finally, after the isolation of Fusobacterium necrophorum from a sterile site, the patient was discharged with intravenous vancomycin and meropenem for the treatment of infection due to anaerobic gram-negative bacilli. Furthermore, the patient was readmitted with fever and a rash, probably caused by the antibiotics. We think that the prudent use of antibiotics must be emphasized, particularly in an academic setting. Sergio Mella, M.D. Alejandro Aguayo, M.D. Maritza Muñoz, Pharm.D. Universidad de Concepción
Concepción, Chile [email protected] No potential conflict of interest relevant to this letter was reported.
1. Case Records of the Massachusetts General Hospital (Case
36-2014). N Engl J Med 2014;371:2018-27.
2. Sullins AK, Abdel-Rahman SM. Pharmacokinetics of anti-
bacterial agents in the CSF of children and adolescents. Paediatr Drugs 2013;15:93-117. DOI: 10.1056/NEJMc1415357
To the Editor: Given that the patient was taking two oral contraceptives, both of which may promote a hypercoagulable state, might those medications have played a role in the development of the Lemierre’s syndrome complication of her infection? If that is the case, then the patient should have been instructed to discontinue those oral contraceptives and to use other forms of birth control. Tristram C. Dammin, M.D. Boston University School of Medicine
Boston, MA [email protected] No potential conflict of interest relevant to this letter was reported. DOI: 10.1056/NEJMc1415357
The Discussant Replies: Mella and colleagues correctly highlight the importance of using the narrowest antibiotic regimen possible. This approach is prudent both to minimize the development of antimicrobial resistance and to decrease adverse effects of medication. In this case, the patient had a life-threatening infection and the potential for gram-positive superinfection, despite negative cultures, particularly given the recent
n engl j med 372;6 nejm.org february 5, 2015
581
The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on February 9, 2015. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
As shown in Figure S2B and S2C in the Supplementary Appendix of the article (available at NEJM.org), there was no clear relation between glucose levels and the depth of ketoacidosis. Glucose levels actually were within the normal range (>4 mmol per liter) during most documented events. Ketoacidotic events were generally associated with reduced, but not absent, caloric intake, and events could be stopped only by supplying ample amounts of carbohydrates, in excess of the basal energy requirement. Our observations are in line with those in several other inborn errors of intermediary metabolism: metabolic derangements occur if — and are maintained as long as — the energy requirements are not met.1 The energy requirement is known to be elevated during infections and to decrease with age,1 providing an explanation for both risk factors. We believe that our data are consistent with the concept that the metabolic derangement (i.e., ketoacidosis) is maintained as long as the body senses a low cellular energy status in peripheral tissues, regardless of the plasma glucose values. Under these low-insulin conditions, glucagon levels probably reflect a low cellular energy status. In patients with low insulin levels owing to diabetes mellitus, glucagon excretion correlates with the depth of ketoacidosis,2 and ketoacidosis can be prevented by blocking the excretion of glucagon.3 Schlegel suggests that the profound ketoacidosis observed in our patients may be an indirect effect of disturbances in the production,
rather than utilization, of ketones. Following that line of reasoning, the loss-of-function mutations found in our patients would be expected to result in decreased import of substrates and, thus, decreased ketone formation, This would, in turn, lead to hypoketosis on fasting, not hyperketosis — a situation similar to the biochemical findings in ketogenesis defects and, perhaps, in a defect of ketone export. Even if the mutations would somehow lead to increased import, given the endocrine control of ketogenesis, that would have mild-to-moderate effects on ketones, at best. Peter M. van Hasselt, M.D., Ph.D. University Medical Center Utrecht, the Netherlands
Sacha Ferdinandusse, Ph.D. Academic Medical Center Amsterdam, the Netherlands
Gijs van Haaften, Ph.D. University Medical Center Utrecht, the Netherlands [email protected] Since publication of their article, the authors report no further potential conflict of interest. 1. Baumgartner MR, Hörster F, Dionisi-Vici C, et al. Proposed
guidelines for the diagnosis and management of methylmalonic and propionic acidemia. Orphanet J Rare Dis 2014;9:130. 2. Liljenquist JE, Bomboy JD, Lewis SB, et al. Effects of glucagon on lipolysis and ketogenesis in normal and diabetic men. J Clin Invest 1974;53:190-7. 3. Gerich JE, Lorenzi M, Bier DM, et al. Prevention of human diabetic ketoacidosis by somatostatin: evidence for an essential role of glucagon. N Engl J Med 1975;292:985-9. DOI: 10.1056/NEJMc1415111
Management of Withdrawal Delirium (Delirium Tremens) To the Editor: The review by Schuckit (Nov. 27 issue)1 of the management of delirium tremens offers the reader a concise lesson in pharmacologic treatment of this medical problem. I would caution those who refer to the lorazepam regimens in Table 3 of the article that the doses outlined may provide an amount of propylene glycol in excess of 25 mg per kilogram per day, the amount considered safe by the World Health Organization.2 Lorazepam injection contains only small amounts of propylene glycol, yet when given at high doses continuously or for prolonged periods, the patient exposure may be substantial.
Metabolic acidosis may develop owing to propylene glycol accumulation, particularly in patients with preexisting renal or hepatic insufficiency. Several investigators have determined that high doses of lorazepam (i.e., >0.1 mg per kilogram per hour), particularly when given as continuous infusions for treatment of alcohol withdrawal for periods of 24 to 48 hours, are associated with propylene glycol levels of more than 25 mg per deciliter.3-5 Those of us who treat alcohol withdrawal should remain alert to the possibility of this toxicity when administering high-dose lorazepam.
n engl j med 372;6 nejm.org february 5, 2015
579
The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on February 9, 2015. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
The
n e w e ng l a n d j o u r na l
of
m e dic i n e
calm, awake patient, facilitating extubation and transfer out of the ICU, where a smooth transiSpartanburg, SC tion to oral benzodiazepines was possible. Blood No potential conflict of interest relevant to this letter was re- alcohol levels, even as ethanol was administered ported. in one arm, were undetectable in samples drawn 1. Schuckit MA. Recognition and management of withdrawal from the other, repeatedly, in all cases. I specudelirium (delirium tremens). N Engl J Med 2014;371:2109-13. 2. Public health statement for propylene glycol. Atlanta: Agency late that the unique sensitivity of ethanol recepfor Toxic Substances and Disease Registry (http://www.atsdr.cdc tors to ethanol, relative to benzodiazepines, in .gov/phs/phs.asp?id=1120&tid=240). preventing withdrawal symptoms permitted this 3. Arroliga AC, Shehab N, McCarthy K, Gonzales JP. Relationship of continuous infusion lorazepam to serum propylene gly- result. It was not necessary to have high alcohol col concentration in critically ill adults. Crit Care Med 2004;32: levels, or even measurable alcohol levels, to pre1709-14. vent withdrawal symptoms. 4. Nelsen JL, Haas CE, Habtemariam B, et al. A prospective Regrettably, concerns about “giving alcohol evaluation of propylene glycol clearance and accumulation during continuous-infusion lorazepam in critically ill patients. J Inten- to alcoholics” have limited more extensive apsive Care Med 2008;23:184-94. plication of this therapeutic approach. The ap5. Horinek EL, Kiser TH, Fish DN, MacLaren R. Propylene glycol accumulation in critically ill patients receiving continuous parent condoning of alcohol intake and the intravenous lorazepam infusions. Ann Pharmacother 2009;43: long-term risks of high doses of alcohol have 1964-71. both been cited2 as reasons why alcohol should DOI: 10.1056/NEJMc1415679 not be used in alcohol withdrawal in general; yet in specific, monitored settings, it is a safe, effecTo the Editor: In his discussion of delirium tre- tive, and comfortable treatment option.1 mens, Schuckit lists grand mal convulsions as a Carmel Armon, M.D., M.H.S. symptom of alcohol withdrawal. After an analy- Tel Aviv University Sackler School of Medicine sis of 308 patients with first-time seizure in the Tel Aviv, Israel Harlem Hospital emergency department and 294 [email protected] No potential conflict of interest relevant to this letter was recontrols, all of whom were admitted to the hos- ported. pital, Ng and colleagues concluded that “Alcohol 1. Rosenbaum M, McCarty T. Alcohol prescription by surgeons withdrawal was not associated with the onset of in the prevention and treatment of delirium tremens: historic seizures in this study; 16 percent of first seizures and current practice. Gen Hosp Psychiatry 2002;24:257-9. in drinkers fell outside the conventionally de- 2. Burns MJ. Delirium tremens (DTs) treatment and management. Medscape. August 2014 (http://emedicine.medscape.com/ fined withdrawal period, and the remainder ex- article/166032-treatment). hibited a seemingly random timing after the last DOI: 10.1056/NEJMc1415679 drink.” What is more, “[o]f the 184 seizures [with good data on abstention time], 61.4 percent oc- The Author Replies: I am grateful to Moriarty, curred within 24 hours of the last drink.”1 Finucane, and Armon for their comments, as well as their interest in the important but relaThomas E. Finucane, M.D. tively poorly studied topic of withdrawal delir Johns Hopkins Institute of Medicine Baltimore, MD ium. The length restrictions for Journal brief [email protected] reviews are important, because they allow interNo potential conflict of interest relevant to this letter was reested readers to quickly acquire useful informaported. tion, but the inherent efficiency of those com1. Ng SK, Hauser WA, Brust JC, Susser M. Alcohol consumption and withdrawal in new-onset seizures. N Engl J Med 1988; munications do not allow for a full discussion of 319:666-73. all interesting elements related to the topic. The DOI: 10.1056/NEJMc1415679 issues raised in these letters enable me to briefly comment on three important topics that might To the Editor: An additional therapeutic option be of interest to clinicians who work with these for patients with a severe acute alcohol with- seriously ill patients. drawal syndrome (delirium tremens) is an intraMoriarty reminds us of the serious acidosis venous alcohol drip.1 In rare instances, I have that can develop from prolonged infusions of administered 10% ethanol at a rate of 50 to 100 ml lorazepam as a consequence of the suspension per hour, in an intensive care unit (ICU), to agi- of this benzodiazepine in propylene glycol. tated, intubated patients requiring high, sedat- Similar to what can be observed with ethylene ing doses of benzodiazepines. This resulted in a glycol and salicylate poisonings, the acid–base Veronica P. Moriarty, Pharm.D. Spartanburg Medical Center
580
n engl j med 372;6 nejm.org february 5, 2015
The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on February 9, 2015. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
correspondence
imbalance can complicate treatment of alcohol withdrawal in persons receiving high rates of infusion of lorazepam, especially in the context of renal dysfunction. This is noteworthy for any clinician treating critically ill patients who require intravenous sedatives. Finucane cites a 1988 article that emphasized that the risk of grand mal convulsions increases with higher quantities of alcohol consumed and that alcohol-related seizures can be observed in the absence of obvious withdrawal symptoms. This is an important reminder of the central role that heavy drinking plays in seizure risk and underscores the need for clinicians to consider heavy alcohol intake in the differential diagnosis of convulsions, even in the absence of an alcohol withdrawal syndrome.
Finally, Armon’s letter is consistent with the conclusion that the treatment of alcohol withdrawal syndromes can include any brain-depressant drug, including alcohol. However, it may be a bit more difficult to establish the appropriate dose of alcohol, as compared with other sedative-hypnotic agents. For mild withdrawal, alcohol may be associated with greater respiratory depression and potentially more intoxicationrelated aggression than benzodiazepines. Marc A. Schuckit, M.D. University of California, San Diego
La Jolla, CA [email protected] Since publication of his article, the author reports no further potential conflict of interest. DOI: 10.1056/NEJMc1415679
Case 36-2014: A Woman with Fever, Pharyngitis, and Double Vision To the Editor: We would like to raise an issue about the Case Record by Olson et al. (Nov. 20 issue).1 Antibiotic resistance is a real problem in the 21st century. In this context, there is consensus about the importance of prudent use of antibiotic agents and the implementation of antibiotic stewardship in health care facilities. We think it is debatable whether vancomycin, mero penem, and clindamycin were the right choices for the treatment of an infection with a micro organism that can penetrate the central nervous system. Clindamycin has shown poor penetration into the central nervous system in adults but has not, to our knowledge, been studied in children.2 Finally, after the isolation of Fusobacterium necrophorum from a sterile site, the patient was discharged with intravenous vancomycin and meropenem for the treatment of infection due to anaerobic gram-negative bacilli. Furthermore, the patient was readmitted with fever and a rash, probably caused by the antibiotics. We think that the prudent use of antibiotics must be emphasized, particularly in an academic setting. Sergio Mella, M.D. Alejandro Aguayo, M.D. Maritza Muñoz, Pharm.D. Universidad de Concepción
Concepción, Chile [email protected] No potential conflict of interest relevant to this letter was reported.
1. Case Records of the Massachusetts General Hospital (Case
36-2014). N Engl J Med 2014;371:2018-27.
2. Sullins AK, Abdel-Rahman SM. Pharmacokinetics of anti-
bacterial agents in the CSF of children and adolescents. Paediatr Drugs 2013;15:93-117. DOI: 10.1056/NEJMc1415357
To the Editor: Given that the patient was taking two oral contraceptives, both of which may promote a hypercoagulable state, might those medications have played a role in the development of the Lemierre’s syndrome complication of her infection? If that is the case, then the patient should have been instructed to discontinue those oral contraceptives and to use other forms of birth control. Tristram C. Dammin, M.D. Boston University School of Medicine
Boston, MA [email protected] No potential conflict of interest relevant to this letter was reported. DOI: 10.1056/NEJMc1415357
The Discussant Replies: Mella and colleagues correctly highlight the importance of using the narrowest antibiotic regimen possible. This approach is prudent both to minimize the development of antimicrobial resistance and to decrease adverse effects of medication. In this case, the patient had a life-threatening infection and the potential for gram-positive superinfection, despite negative cultures, particularly given the recent
n engl j med 372;6 nejm.org february 5, 2015
581
The New England Journal of Medicine Downloaded from nejm.org at UC SHARED JOURNAL COLLECTION on February 9, 2015. For personal use only. No other uses without permission. Copyright © 2015 Massachusetts Medical Society. All rights reserved.
