Letters
5. Freshwater DM, Freshwater MF. Failure by deans of academic medical centers to disclose outside income. Arch Intern Med. 2011;171(6):586-587. 6. Lo B. Serving two masters—conflicts of interest in academic medicine. N Engl J Med. 2010;362(8):669-671.
COMMENT & RESPONSE
Training for Effective Patient Communication To the Editor In their study addressing the effects of communication skills training on patient satisfaction and depression scores, Dr Curtis and colleagues1 argued that the lack of effect may be attributable to inaccurate assessment by patients. They discussed the possibility that patients and family members may require specific training to accurately value the improved communication skills resulting from intensive training of caregivers. I was amazed that Curtis et al1 would suggest that patients did not appreciate the trainees’ improved communication skills because of lack of training. In fact, the patients became significantly more depressed, which Curtis et al 1 downplayed. I believe that the comment by Curtis et al1 that “increasing patients’ awareness … may trigger negative experiences” is an important point. Although optimal communication may be about patients’ awareness and complete transfer of medical facts, and colleagues and professionally trained simulation patients may provide high ratings for such skills, this type of communication may not work for real patients who hear sad news. Physicians can be too explicit, too complete. Patients may find this type of honesty insensitive and difficult to handle. Giving the patients more time, remaining genuine in such moments, and respecting their grief, including its temporary limits, might work better than a course in state-of-the-art communication skills. Yvo Smulders, MD Author Affiliation: Department of Internal Medicine, VU University Medical Center, Amsterdam, the Netherlands. Corresponding Author: Yvo Smulders, MD, VU University Medical Center, PO Box 7057, NH 1007MB Amsterdam, the Netherlands ([email protected]). Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
A convincing chain of evidence is needed to link educational classroom, laboratory, or clinical learning outcomes to downstream clinical results.2 However, Curtis et al1 did not report educational outcome data (ie, acquisition of communication skills among medical residents and nurse practitioner students) in the context of their study. Instead, they cited earlier research 3 with a different learner sample (hematology/oncology fellows) and extrapolated educational outcomes from seasoned clinicians to junior trainees. The study by Curtis et al1 offered no evidence that the trainees’ communication skills actually improved. The effect of an educational intervention is demonstrated from rigorous pretest and posttest evaluations, using measures that yield reliable data that permit valid judgments. Curtis et al1 did not report such data, which are essential to stretch the measurement end point to the bedside and link skill acquisition with improved patient care practices and better outcomes. The high dropout rate (21%) in the intervention group is concerning. Failure to complete the intervention dilutes the training effect and raises questions about participant satisfaction and the curriculum’s efficacy and reproducibility. It also is unclear to us which patient surveys were included in the analysis. We seek clarification about the number of trainees in both groups who were evaluated and how the surveys from the patients of the 48 intervention group trainees who dropped out were managed. Simulation-based education is a powerful tool that positively affects the clinical environment.4 Use of education best practices, including deliberate practice and mastery learning, can amplify this effect. In mastery learning, education outcomes are uniform while training time may vary. Only after reaching a predetermined minimum standard may trainees perform the skill in actual patient care.5 This model allows clinical supervisors to ensure that only competent personnel provide patient care. We agree with Curtis et al 1 that teaching end-of-life communication skills is challenging. However, we believe more rigorous research methods might yield different outcomes. Diane B. Wayne, MD William C. McGaghie, PhD
1. Curtis JR, Back AL, Ford DW, et al. Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial. JAMA. 2013;310(21):2271-2281.
Author Affiliations: Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois (Wayne); Ralph P. Leischner Jr, MD, Institute for Medical Education, Loyola University Chicago Stritch School of Medicine, Maywood, Illinois (McGaghie).
To the Editor In a recent report, Dr Curtis and colleagues1 found that simulation-based education did not improve patientreported quality of communication. We disagree with the authors’ conclusion that “these findings raise questions about skills transfer from simulation training to actual patient care and the adequacy of communication skills assessment.” Instead, we propose a different explanation: the educational intervention was neither powerful nor assessed rigorously, which would prevent transfer of training outcomes to measured patient care practices or patient outcomes.
Corresponding Author: Diane B. Wayne, MD, Northwestern University Feinberg School of Medicine, 251 E Huron St, Chicago, IL 60611 (dwayne @northwestern.edu). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Curtis JR, Back AL, Ford DW, et al. Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial. JAMA. 2013;310(21):2271-2281. 2. Cook DA, West CP. Perspective: reconsidering the focus on “outcomes research” in medical education: a cautionary note. Acad Med. 2013;88(2): 162-167.
jama.com
JAMA April 2, 2014 Volume 311, Number 13
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of Georgia User on 05/29/2015
1355
Letters
3. Back AL, Arnold RM, Baile WF, et al. Efficacy of communication skills training for giving bad news and discussing transitions to palliative care. Arch Intern Med. 2007;167(5):453-460. 4. McGaghie WC, Issenberg SB, Cohen ER, Barsuk JH, Wayne DB. Translational educational research: a necessity for effective health-care improvement. Chest. 2012;142(5):1097-1103. 5. Cohen ER, Barsuk JH, Moazed F, et al. Making July safer: simulation-based mastery learning during intern boot camp. Acad Med. 2013;88(2):233-239.
To the Editor Dr Curtis and colleagues 1 reported that simulation-based communication training compared with usual education did not improve the quality of communication about end-of-life care or the quality of end-oflife care itself, but was associated with a small increase in patients’ depressive symptoms. The authors drew their conclusions based on evaluations of patients with life-limiting illnesses. Although Curtis et al 1 provided data concerning the diagnosis of the patients (eg, cancer, chronic obstructive pulmonary disease, other lung disease, congestive heart failure, end-stage liver disease), I wonder if quality of communication about end-of-life care or quality of end-of-life care improved more with patients with specific diagnoses. Because the intervention was adapted from a workshop associated with improved communication skills for oncology fellows,2 the intervention might be more effective with cancer patients. Could the authors provide additional information on the relationship between the underlying disease of the patients and the results? Masako Sugihara, MD, MA Author Affiliation: Department of Neuropsychiatry, Keio University School of Medicine, Tokyo, Japan. Corresponding Author: Masako Sugihara, MD, MA, Keio University School of Medicine, 35 Shinanomachi, Tokyo 160-8582, Japan (waruko_sugihara@yahoo .co.jp). Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Curtis JR, Back AL, Ford DW, et al. Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial. JAMA. 2013;310(21):2271-2281. 2. Back AL, Arnold RM, Baile WF, et al. Efficacy of communication skills training for giving bad news and discussing transitions to palliative care. Arch Intern Med. 2007;167(5):453-460.
In Reply Dr Smulders takes issue with our comment that patients’ ratings of communications skills may have limitations. We did not intend to suggest that patients were not important evaluators of communication. Patient perceptions are crucial. However, when the conversation involves topics that patients do not want to talk about, or find uncomfortable or sad, patient ratings may have limitations as an outcome measure for several reasons. First, patients may have little experience with other clinicians to help form a judgment. Second, their own emotions raised by the conversation may be more influential in their rat1356
ing than what the clinician said or how he or she said it. Third, some patients may judge optimism preferable to straightforward discussion of prognosis, even when it is unrealistic or leads to misinformation. A recent study found that, for patients with advanced cancer, better understanding of their prognosis was associated with lower ratings of physicians’ communication.1 One method of addressing these patient-level measurement challenges is to ask patients more specific questions about what communication behaviors they observed. Another method may be to prompt patients with suggestions for the specific aspects or tasks of communication that they should evaluate. Drs Wayne and McGaghie state that there were no educational outcome data for this specific intervention. However, educational outcome data for participants in this intervention have been published.2 Our intervention resulted in a statistically significant level of skill acquisition in a pretest and posttest of trainees from this study.2 Regarding the dropout rate, we designed the study intervention to be implemented and assessed in the realworld context of busy internal medicine and nurse practitioner training programs because we thought transferability was important. Despite extraordinar y commitment and work from program directors and staff, most of the trainees who dropped out of the intervention could not participate because of scheduling conflicts or vacations. Nonetheless, this is an important issue to consider for future research. In addition, Wayne and McGaghie ask which patients were included in the analyses and how the surveys for trainees who dropped out were managed. In addressing this reply, we realized that Figure 1 contains a misleading word that may have created confusion. Where the figure states “evaluations received,” it should say “evaluations sought.” The figure shows the evaluations that were requested from patients, family members, and clinicians as well as those returned and used in the analyses. A correction accompanies this letter. Dr Sugihara questions whether outcomes differed across different diseases because the intervention was adapted from one originally designed for oncologists. We found that this intervention resulted in an increase in communication skills among these internal medicine and nurse practitioner trainees.2 We also tested for evidence of a differential effect among patients with cancer and found that patients with cancer did not have significantly different outcomes. Our study used state-of-the-science interventions and measurements yet the results, which we also found disappointing, underscore the importance of further development in the science of both interventions and measurements. As noted in the accompanying Editorial, there is much work to do.3 J. Randall Curtis, MD, MPH Anthony L. Back, MD Ruth A. Engelberg, PhD
JAMA April 2, 2014 Volume 311, Number 13
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of Georgia User on 05/29/2015
jama.com
Letters
Author Affiliations: Division of Pulmonary and Critical Care, University of Washington, Seattle (Curtis, Engelberg); Department of Medicine, University of Washington, Seattle (Back). Corresponding Author: J. Randall Curtis, MD, MPH, University of Washington, 325 Ninth Ave, Seattle, WA 98104 ([email protected]).
We believe that new studies that are large enough to avoid any imbalance in important confounders should be conducted to evaluate the effect of hypothermia induced by techniques that do not rely on fluid administration.
Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Drs Curtis and Engelberg reported receiving grants from the National Institutes of Health. No other disclosures were reported.
Romain Pirracchio, MD, PhD Didier Journois, MD, PhD
1. Weeks JC, Catalano PJ, Cronin A, et al. Patients’ expectations about effects of chemotherapy for advanced cancer. N Engl J Med. 2012;367(17):1616-1625.
Author Affiliations: Department of Anesthesiology and Critical Care Medicine, Hôpital Européen Georges Pompidou, Paris, France.
2. Bays AM, Engelberg RA, Back AL, et al. Interprofessional communication skills training for serious illness: evaluation of a small-group, simulated patient intervention. J Palliat Med. 2014;17(2):159-166.
Corresponding Author: Romain Pirracchio, MD, PhD, Hôpital Européen Georges Pompidou, 20 Rue Leblanc, Paris 75015, France (romainpirracchio @yahoo.fr).
3. Chi J, Verghese A. Improving communication with patients: learning by doing. JAMA. 2013;310(21):2257-2258.
Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Hypothermia for Bacterial Meningitis To the Editor Although we acknowledge the importance and the quality of the study by Dr Mourvillier and colleagues,1 we have some reservations about the authors’ conclusions. They stated that moderate hypothermia does not improve the outcome in patients with severe bacterial meningitis and may even be harmful. For several reasons, we think that this conclusion is not fully supported by the data. First, the treatment evaluated was not moderate hypothermia but early fluid loading using cold saline. The effect of hypothermia alone would have been tested only if both groups had received comparable amounts of fluid. The induction of hypothermia required a median volume of cold saline of 2401 mL; the amount of fluid received in the control group was not reported. Fluid management, especially for chloride solutions, is important in this clinical context because fluid overload may be deleterious at the early stage of a cerebral insult.2 In addition, massive chloride loading may increase the risk of kidney injury.3 Hence, hypothermia using cold fluid administration could have harmed the treatment group for reasons unrelated to any potential detrimental effect of hypothermia. Second, the study had noncomparable groups, especially with respect to the prevalence of septic shock (37% in the treatment group and 20% in the control group). This imbalance could have contributed to the results in the treatment group. The cause of death was systemic in 44% of the patients in the treatment group vs 38% in the control group. In a multivariable analysis, no association was found between hypothermia and mortality (hazard ratio [HR], 1.76; 95% CI, 0.89-3.45), but septic shock was associated with greater mortality (HR, 2.54; 95% CI, 1.31-4.94). Third, the difference in the proportion of patients reaching a Glasgow Outcome Scale (GOS) score of 5 may be explained by the higher mortality rate in the treatment group. However, among survivors, a larger proportion of patients reached a GOS score of 4 in the treatment group (20% in the treatment group vs 12% in the control group) and fewer patients had a GOS score of 3 (14% vs 31%, respectively). Hence, we conjecture that hypothermia might be associated with better neurological outcomes.
1. Mourvillier B, Tubach F, van de Beek D, et al. Induced hypothermia in severe bacterial meningitis: a randomized clinical trial. JAMA. 2013;310(20):2174-2183. 2. Ichai C, Payen J-F, Orban J-C, et al. Half-molar sodium lactate infusion to prevent intracranial hypertensive episodes in severe traumatic brain injured patients: a randomized controlled trial. Intensive Care Med. 2013;39(8):14131422. 3. Yunos NM, Bellomo R, Hegarty C, Story D, Ho L, Bailey M. Association between a chloride-liberal vs chloride-restrictive intravenous fluid administration strategy and kidney injury in critically ill adults. JAMA. 2012;308(15):1566-1572.
To the Editor In the randomized clinical trial by Dr Mourvillier and colleagues,1 the authors found that induced hypothermia in adults with severe community-acquired bacterial meningitis did not improve the outcome and may even be harmful. Furthermore, the authors stated that such results could have important implications for future trials on hypothermia in such patients. The results of this trial are in contrast with experimental studies and recent observations in case series.2,3 Therefore, several important issues regarding the recruitment and methods of this trial should be discussed. First, recruitment at each of the participating centers was low (1 patient per center for 3 years). Although severe bacterial meningitis is a rare disease, we would have expected the number of recruited patients from the 49 centers to be much higher. Therefore, a possibility of selection bias is evident. Second, the baseline characteristics of patients revealed a high proportion with septic shock and who were deeply comatose at admission (median Glasgow Coma Scale score of 7 [interquartile range, 4-8] in the hypothermia group). Furthermore, the presence of septic shock at admission was determined to be a significant predictor for adverse outcome (HR, 2.54; 95% CI, 1.31-4.94). Data regarding the duration of the disease, antibiotic timing according to the disease onset, associated serious comorbidities, and results of blood cultures were not shown. Third, therapeutic hypothermia was arbitrarily conducted for 48 hours without any measurements to show whether hypothermia itself was beneficial or harmful. We agree with the authors that early stopping of the clinical trial precluded firm conclusions about the effects of therapeutic hypothermia in bacterial meningitis. Further trials in-
jama.com
JAMA April 2, 2014 Volume 311, Number 13
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of Georgia User on 05/29/2015
1357
5. Freshwater DM, Freshwater MF. Failure by deans of academic medical centers to disclose outside income. Arch Intern Med. 2011;171(6):586-587. 6. Lo B. Serving two masters—conflicts of interest in academic medicine. N Engl J Med. 2010;362(8):669-671.
COMMENT & RESPONSE
Training for Effective Patient Communication To the Editor In their study addressing the effects of communication skills training on patient satisfaction and depression scores, Dr Curtis and colleagues1 argued that the lack of effect may be attributable to inaccurate assessment by patients. They discussed the possibility that patients and family members may require specific training to accurately value the improved communication skills resulting from intensive training of caregivers. I was amazed that Curtis et al1 would suggest that patients did not appreciate the trainees’ improved communication skills because of lack of training. In fact, the patients became significantly more depressed, which Curtis et al 1 downplayed. I believe that the comment by Curtis et al1 that “increasing patients’ awareness … may trigger negative experiences” is an important point. Although optimal communication may be about patients’ awareness and complete transfer of medical facts, and colleagues and professionally trained simulation patients may provide high ratings for such skills, this type of communication may not work for real patients who hear sad news. Physicians can be too explicit, too complete. Patients may find this type of honesty insensitive and difficult to handle. Giving the patients more time, remaining genuine in such moments, and respecting their grief, including its temporary limits, might work better than a course in state-of-the-art communication skills. Yvo Smulders, MD Author Affiliation: Department of Internal Medicine, VU University Medical Center, Amsterdam, the Netherlands. Corresponding Author: Yvo Smulders, MD, VU University Medical Center, PO Box 7057, NH 1007MB Amsterdam, the Netherlands ([email protected]). Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
A convincing chain of evidence is needed to link educational classroom, laboratory, or clinical learning outcomes to downstream clinical results.2 However, Curtis et al1 did not report educational outcome data (ie, acquisition of communication skills among medical residents and nurse practitioner students) in the context of their study. Instead, they cited earlier research 3 with a different learner sample (hematology/oncology fellows) and extrapolated educational outcomes from seasoned clinicians to junior trainees. The study by Curtis et al1 offered no evidence that the trainees’ communication skills actually improved. The effect of an educational intervention is demonstrated from rigorous pretest and posttest evaluations, using measures that yield reliable data that permit valid judgments. Curtis et al1 did not report such data, which are essential to stretch the measurement end point to the bedside and link skill acquisition with improved patient care practices and better outcomes. The high dropout rate (21%) in the intervention group is concerning. Failure to complete the intervention dilutes the training effect and raises questions about participant satisfaction and the curriculum’s efficacy and reproducibility. It also is unclear to us which patient surveys were included in the analysis. We seek clarification about the number of trainees in both groups who were evaluated and how the surveys from the patients of the 48 intervention group trainees who dropped out were managed. Simulation-based education is a powerful tool that positively affects the clinical environment.4 Use of education best practices, including deliberate practice and mastery learning, can amplify this effect. In mastery learning, education outcomes are uniform while training time may vary. Only after reaching a predetermined minimum standard may trainees perform the skill in actual patient care.5 This model allows clinical supervisors to ensure that only competent personnel provide patient care. We agree with Curtis et al 1 that teaching end-of-life communication skills is challenging. However, we believe more rigorous research methods might yield different outcomes. Diane B. Wayne, MD William C. McGaghie, PhD
1. Curtis JR, Back AL, Ford DW, et al. Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial. JAMA. 2013;310(21):2271-2281.
Author Affiliations: Department of Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois (Wayne); Ralph P. Leischner Jr, MD, Institute for Medical Education, Loyola University Chicago Stritch School of Medicine, Maywood, Illinois (McGaghie).
To the Editor In a recent report, Dr Curtis and colleagues1 found that simulation-based education did not improve patientreported quality of communication. We disagree with the authors’ conclusion that “these findings raise questions about skills transfer from simulation training to actual patient care and the adequacy of communication skills assessment.” Instead, we propose a different explanation: the educational intervention was neither powerful nor assessed rigorously, which would prevent transfer of training outcomes to measured patient care practices or patient outcomes.
Corresponding Author: Diane B. Wayne, MD, Northwestern University Feinberg School of Medicine, 251 E Huron St, Chicago, IL 60611 (dwayne @northwestern.edu). Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Curtis JR, Back AL, Ford DW, et al. Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial. JAMA. 2013;310(21):2271-2281. 2. Cook DA, West CP. Perspective: reconsidering the focus on “outcomes research” in medical education: a cautionary note. Acad Med. 2013;88(2): 162-167.
jama.com
JAMA April 2, 2014 Volume 311, Number 13
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of Georgia User on 05/29/2015
1355
Letters
3. Back AL, Arnold RM, Baile WF, et al. Efficacy of communication skills training for giving bad news and discussing transitions to palliative care. Arch Intern Med. 2007;167(5):453-460. 4. McGaghie WC, Issenberg SB, Cohen ER, Barsuk JH, Wayne DB. Translational educational research: a necessity for effective health-care improvement. Chest. 2012;142(5):1097-1103. 5. Cohen ER, Barsuk JH, Moazed F, et al. Making July safer: simulation-based mastery learning during intern boot camp. Acad Med. 2013;88(2):233-239.
To the Editor Dr Curtis and colleagues 1 reported that simulation-based communication training compared with usual education did not improve the quality of communication about end-of-life care or the quality of end-oflife care itself, but was associated with a small increase in patients’ depressive symptoms. The authors drew their conclusions based on evaluations of patients with life-limiting illnesses. Although Curtis et al 1 provided data concerning the diagnosis of the patients (eg, cancer, chronic obstructive pulmonary disease, other lung disease, congestive heart failure, end-stage liver disease), I wonder if quality of communication about end-of-life care or quality of end-of-life care improved more with patients with specific diagnoses. Because the intervention was adapted from a workshop associated with improved communication skills for oncology fellows,2 the intervention might be more effective with cancer patients. Could the authors provide additional information on the relationship between the underlying disease of the patients and the results? Masako Sugihara, MD, MA Author Affiliation: Department of Neuropsychiatry, Keio University School of Medicine, Tokyo, Japan. Corresponding Author: Masako Sugihara, MD, MA, Keio University School of Medicine, 35 Shinanomachi, Tokyo 160-8582, Japan (waruko_sugihara@yahoo .co.jp). Conflict of Interest Disclosures: The author has completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported. 1. Curtis JR, Back AL, Ford DW, et al. Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial. JAMA. 2013;310(21):2271-2281. 2. Back AL, Arnold RM, Baile WF, et al. Efficacy of communication skills training for giving bad news and discussing transitions to palliative care. Arch Intern Med. 2007;167(5):453-460.
In Reply Dr Smulders takes issue with our comment that patients’ ratings of communications skills may have limitations. We did not intend to suggest that patients were not important evaluators of communication. Patient perceptions are crucial. However, when the conversation involves topics that patients do not want to talk about, or find uncomfortable or sad, patient ratings may have limitations as an outcome measure for several reasons. First, patients may have little experience with other clinicians to help form a judgment. Second, their own emotions raised by the conversation may be more influential in their rat1356
ing than what the clinician said or how he or she said it. Third, some patients may judge optimism preferable to straightforward discussion of prognosis, even when it is unrealistic or leads to misinformation. A recent study found that, for patients with advanced cancer, better understanding of their prognosis was associated with lower ratings of physicians’ communication.1 One method of addressing these patient-level measurement challenges is to ask patients more specific questions about what communication behaviors they observed. Another method may be to prompt patients with suggestions for the specific aspects or tasks of communication that they should evaluate. Drs Wayne and McGaghie state that there were no educational outcome data for this specific intervention. However, educational outcome data for participants in this intervention have been published.2 Our intervention resulted in a statistically significant level of skill acquisition in a pretest and posttest of trainees from this study.2 Regarding the dropout rate, we designed the study intervention to be implemented and assessed in the realworld context of busy internal medicine and nurse practitioner training programs because we thought transferability was important. Despite extraordinar y commitment and work from program directors and staff, most of the trainees who dropped out of the intervention could not participate because of scheduling conflicts or vacations. Nonetheless, this is an important issue to consider for future research. In addition, Wayne and McGaghie ask which patients were included in the analyses and how the surveys for trainees who dropped out were managed. In addressing this reply, we realized that Figure 1 contains a misleading word that may have created confusion. Where the figure states “evaluations received,” it should say “evaluations sought.” The figure shows the evaluations that were requested from patients, family members, and clinicians as well as those returned and used in the analyses. A correction accompanies this letter. Dr Sugihara questions whether outcomes differed across different diseases because the intervention was adapted from one originally designed for oncologists. We found that this intervention resulted in an increase in communication skills among these internal medicine and nurse practitioner trainees.2 We also tested for evidence of a differential effect among patients with cancer and found that patients with cancer did not have significantly different outcomes. Our study used state-of-the-science interventions and measurements yet the results, which we also found disappointing, underscore the importance of further development in the science of both interventions and measurements. As noted in the accompanying Editorial, there is much work to do.3 J. Randall Curtis, MD, MPH Anthony L. Back, MD Ruth A. Engelberg, PhD
JAMA April 2, 2014 Volume 311, Number 13
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of Georgia User on 05/29/2015
jama.com
Letters
Author Affiliations: Division of Pulmonary and Critical Care, University of Washington, Seattle (Curtis, Engelberg); Department of Medicine, University of Washington, Seattle (Back). Corresponding Author: J. Randall Curtis, MD, MPH, University of Washington, 325 Ninth Ave, Seattle, WA 98104 ([email protected]).
We believe that new studies that are large enough to avoid any imbalance in important confounders should be conducted to evaluate the effect of hypothermia induced by techniques that do not rely on fluid administration.
Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Drs Curtis and Engelberg reported receiving grants from the National Institutes of Health. No other disclosures were reported.
Romain Pirracchio, MD, PhD Didier Journois, MD, PhD
1. Weeks JC, Catalano PJ, Cronin A, et al. Patients’ expectations about effects of chemotherapy for advanced cancer. N Engl J Med. 2012;367(17):1616-1625.
Author Affiliations: Department of Anesthesiology and Critical Care Medicine, Hôpital Européen Georges Pompidou, Paris, France.
2. Bays AM, Engelberg RA, Back AL, et al. Interprofessional communication skills training for serious illness: evaluation of a small-group, simulated patient intervention. J Palliat Med. 2014;17(2):159-166.
Corresponding Author: Romain Pirracchio, MD, PhD, Hôpital Européen Georges Pompidou, 20 Rue Leblanc, Paris 75015, France (romainpirracchio @yahoo.fr).
3. Chi J, Verghese A. Improving communication with patients: learning by doing. JAMA. 2013;310(21):2257-2258.
Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.
Hypothermia for Bacterial Meningitis To the Editor Although we acknowledge the importance and the quality of the study by Dr Mourvillier and colleagues,1 we have some reservations about the authors’ conclusions. They stated that moderate hypothermia does not improve the outcome in patients with severe bacterial meningitis and may even be harmful. For several reasons, we think that this conclusion is not fully supported by the data. First, the treatment evaluated was not moderate hypothermia but early fluid loading using cold saline. The effect of hypothermia alone would have been tested only if both groups had received comparable amounts of fluid. The induction of hypothermia required a median volume of cold saline of 2401 mL; the amount of fluid received in the control group was not reported. Fluid management, especially for chloride solutions, is important in this clinical context because fluid overload may be deleterious at the early stage of a cerebral insult.2 In addition, massive chloride loading may increase the risk of kidney injury.3 Hence, hypothermia using cold fluid administration could have harmed the treatment group for reasons unrelated to any potential detrimental effect of hypothermia. Second, the study had noncomparable groups, especially with respect to the prevalence of septic shock (37% in the treatment group and 20% in the control group). This imbalance could have contributed to the results in the treatment group. The cause of death was systemic in 44% of the patients in the treatment group vs 38% in the control group. In a multivariable analysis, no association was found between hypothermia and mortality (hazard ratio [HR], 1.76; 95% CI, 0.89-3.45), but septic shock was associated with greater mortality (HR, 2.54; 95% CI, 1.31-4.94). Third, the difference in the proportion of patients reaching a Glasgow Outcome Scale (GOS) score of 5 may be explained by the higher mortality rate in the treatment group. However, among survivors, a larger proportion of patients reached a GOS score of 4 in the treatment group (20% in the treatment group vs 12% in the control group) and fewer patients had a GOS score of 3 (14% vs 31%, respectively). Hence, we conjecture that hypothermia might be associated with better neurological outcomes.
1. Mourvillier B, Tubach F, van de Beek D, et al. Induced hypothermia in severe bacterial meningitis: a randomized clinical trial. JAMA. 2013;310(20):2174-2183. 2. Ichai C, Payen J-F, Orban J-C, et al. Half-molar sodium lactate infusion to prevent intracranial hypertensive episodes in severe traumatic brain injured patients: a randomized controlled trial. Intensive Care Med. 2013;39(8):14131422. 3. Yunos NM, Bellomo R, Hegarty C, Story D, Ho L, Bailey M. Association between a chloride-liberal vs chloride-restrictive intravenous fluid administration strategy and kidney injury in critically ill adults. JAMA. 2012;308(15):1566-1572.
To the Editor In the randomized clinical trial by Dr Mourvillier and colleagues,1 the authors found that induced hypothermia in adults with severe community-acquired bacterial meningitis did not improve the outcome and may even be harmful. Furthermore, the authors stated that such results could have important implications for future trials on hypothermia in such patients. The results of this trial are in contrast with experimental studies and recent observations in case series.2,3 Therefore, several important issues regarding the recruitment and methods of this trial should be discussed. First, recruitment at each of the participating centers was low (1 patient per center for 3 years). Although severe bacterial meningitis is a rare disease, we would have expected the number of recruited patients from the 49 centers to be much higher. Therefore, a possibility of selection bias is evident. Second, the baseline characteristics of patients revealed a high proportion with septic shock and who were deeply comatose at admission (median Glasgow Coma Scale score of 7 [interquartile range, 4-8] in the hypothermia group). Furthermore, the presence of septic shock at admission was determined to be a significant predictor for adverse outcome (HR, 2.54; 95% CI, 1.31-4.94). Data regarding the duration of the disease, antibiotic timing according to the disease onset, associated serious comorbidities, and results of blood cultures were not shown. Third, therapeutic hypothermia was arbitrarily conducted for 48 hours without any measurements to show whether hypothermia itself was beneficial or harmful. We agree with the authors that early stopping of the clinical trial precluded firm conclusions about the effects of therapeutic hypothermia in bacterial meningitis. Further trials in-
jama.com
JAMA April 2, 2014 Volume 311, Number 13
Copyright 2014 American Medical Association. All rights reserved.
Downloaded From: http://jama.jamanetwork.com/ by a University of Georgia User on 05/29/2015
1357
