User login
Improving our crystal ball: prognostication in neuroscience ICUs
The most difficult decisions in neuroscience intensive care units often involve patients’ ultimate goals of care. Oftentimes, family members of a brain-injured patient with an apparently poor neurologic prognosis must weigh whether their loved one would have preferred prolongation of aggressive ICU and post-ICU care, often with little to no chance for “meaningful” recovery, or death via the institution of comfort measures only. Proper prognostication is crucial to the family when making such decisions. However, the process of formulating and talking about prognosis for our most severely affected patients is subject to physician and family biases, families’ insufficient understanding of projected outcomes, and sometimes clinical nihilism by the physicians.
The process of predicting the outcomes of patients with traumatic brain injury (TBI) serves as an example of these issues. Moderate to severe TBI continues to be a leading cause of death and disability in the United States.1 Most deaths of patients with moderate to severe TBI follow decisions by doctors and families to pursue comfort care only. However, these decisions occur at a disconcertingly highly variable rate at different trauma centers, with the variation seemingly unrelated to patients’ disease severity, age, or previously diagnosed comorbidities.2 These patients are at risk for their care being influenced by a self-fulfilling prophecy: That is, the impression of a poor prognosis communicated by clinicians to a patient’s family, whether correct or incorrect, affects the aggressiveness of the care that a patient receives and determines the patient’s outcome.3
Remedying these issues through a family or health care proxy decision support intervention (“decision aid”) that could improve and standardize the way TBI prognosis is communicated may lead to better informed decisions for these critically ill patients, with potentially less decisional regret and post-ICU stress disorders in families, and decisions more in line with the patient’s values and preferences.4 A recent Cochrane review showed that for a decision aid to be effective and integrated into routine clinical care, it must contain disease-specific data tailored to patients and their families/proxies, and be simple and time efficient for physicians to use.5 Taking these factors into account, researchers at the University of Massachusetts are developing a National Institutes of Health–funded pilot decision aid for goals-of-care decisions in critically-ill TBI patients.
While the field of TBI has tools such as the IMPACT calculator that can be used to estimate a patient’s long-term prognosis based on how patients with similar clinical characteristics in large clinical databases have done, the fundamental uncertainty of prognosis remains a difficult challenge.6 Arguably, this challenge is even more daunting when estimating prognosis for patients with severe ischemic stroke and intracerebral hemorrhage (ICH). The use of ischemic stroke outcome prediction tools is complicated, as many of them are based on population databases with wide variations in whether included patients received intravenous tissue plasminogen activator, endovascular therapy, both, or neither. Furthermore, a recent study comparing the accuracy of the ICH score for predicting 3-month outcome for ICH patients to the subjective predictions of clinicians made within 24 hours of patient admission found that the educated guesses of physicians and nurses overall seemed to correlate with actual outcomes more closely than the ICH score output.7 This finding highlights the challenge of using available outcome “calculators” for individual patients in ICUs.
Ultimately, the decisions made about the goals of care for ICU patients come down not only to what their expected outcomes are, but also whether their surrogate decision makers believe that those outcomes would be acceptable to the patient.8 Potential pitfalls abound with regard to this issue as well. Decision makers are often not made aware of the fact that many times patients with significant disability may nevertheless report a reasonable quality of life. By their very nature, conversations regarding patient prognosis inevitably focus on what future disabilities one might expect; accounting for a patient’s possible adaptation to disability is both easy to overlook and hard to accomplish even when given adequate attention.9 Improvements in the field of neuroprognostication may not only depend on the development of new shared decision making tools for physicians and families but also on increasing awareness of the limitations of prognostic scales and the cognitive biases that may exist when discussing the possibilities of future disability.
References
1. Traumatic Brain Injury Statistics [online]. Available at: http://www.cdc.gov/traumaticbraininjury/statistics.html. Accessed Nov. 1.
3. Neurocrit Care. 2013;19:347-63.
4. Col NF. Chapter 17: Shared Decision Making. In: Communicating Risks and Benefits: An Evidence-Based User’s Guide [online]. Available at http://www.fda.gov/downloads/AboutFDA/ReportsManualsForms/Reports/UCM268069.pdf.
5. Cochrane Database Syst Rev. 2011 Oct 5;(10):CD001431.
6. PLoS Med. 2008 Aug;5(8):e165; discussion e168.
8. Neurocrit Care. 2015;23:131-41.
Dr. Muehlschlegel is associate professor of neurology (neurocritical care), anesthesia/critical care, and surgery at the University of Massachusetts, Worcester. Dr. Hwang is assistant professor of neurology in the division of neurocritical care and emergency neurology at Yale University, New Haven, Conn. Dr. Muehlschlegel reported receiving a grant from the National Institutes of Health for her research in developing a pilot decision aid for goals-of-care decisions in critically-ill TBI patients. Dr. Hwang reported receiving research funding from the American Brain Foundation, the Apple Pickers Foundation, the National Institute on Aging, and the Neurocritical Care Society.
The most difficult decisions in neuroscience intensive care units often involve patients’ ultimate goals of care. Oftentimes, family members of a brain-injured patient with an apparently poor neurologic prognosis must weigh whether their loved one would have preferred prolongation of aggressive ICU and post-ICU care, often with little to no chance for “meaningful” recovery, or death via the institution of comfort measures only. Proper prognostication is crucial to the family when making such decisions. However, the process of formulating and talking about prognosis for our most severely affected patients is subject to physician and family biases, families’ insufficient understanding of projected outcomes, and sometimes clinical nihilism by the physicians.
The process of predicting the outcomes of patients with traumatic brain injury (TBI) serves as an example of these issues. Moderate to severe TBI continues to be a leading cause of death and disability in the United States.1 Most deaths of patients with moderate to severe TBI follow decisions by doctors and families to pursue comfort care only. However, these decisions occur at a disconcertingly highly variable rate at different trauma centers, with the variation seemingly unrelated to patients’ disease severity, age, or previously diagnosed comorbidities.2 These patients are at risk for their care being influenced by a self-fulfilling prophecy: That is, the impression of a poor prognosis communicated by clinicians to a patient’s family, whether correct or incorrect, affects the aggressiveness of the care that a patient receives and determines the patient’s outcome.3
Remedying these issues through a family or health care proxy decision support intervention (“decision aid”) that could improve and standardize the way TBI prognosis is communicated may lead to better informed decisions for these critically ill patients, with potentially less decisional regret and post-ICU stress disorders in families, and decisions more in line with the patient’s values and preferences.4 A recent Cochrane review showed that for a decision aid to be effective and integrated into routine clinical care, it must contain disease-specific data tailored to patients and their families/proxies, and be simple and time efficient for physicians to use.5 Taking these factors into account, researchers at the University of Massachusetts are developing a National Institutes of Health–funded pilot decision aid for goals-of-care decisions in critically-ill TBI patients.
While the field of TBI has tools such as the IMPACT calculator that can be used to estimate a patient’s long-term prognosis based on how patients with similar clinical characteristics in large clinical databases have done, the fundamental uncertainty of prognosis remains a difficult challenge.6 Arguably, this challenge is even more daunting when estimating prognosis for patients with severe ischemic stroke and intracerebral hemorrhage (ICH). The use of ischemic stroke outcome prediction tools is complicated, as many of them are based on population databases with wide variations in whether included patients received intravenous tissue plasminogen activator, endovascular therapy, both, or neither. Furthermore, a recent study comparing the accuracy of the ICH score for predicting 3-month outcome for ICH patients to the subjective predictions of clinicians made within 24 hours of patient admission found that the educated guesses of physicians and nurses overall seemed to correlate with actual outcomes more closely than the ICH score output.7 This finding highlights the challenge of using available outcome “calculators” for individual patients in ICUs.
Ultimately, the decisions made about the goals of care for ICU patients come down not only to what their expected outcomes are, but also whether their surrogate decision makers believe that those outcomes would be acceptable to the patient.8 Potential pitfalls abound with regard to this issue as well. Decision makers are often not made aware of the fact that many times patients with significant disability may nevertheless report a reasonable quality of life. By their very nature, conversations regarding patient prognosis inevitably focus on what future disabilities one might expect; accounting for a patient’s possible adaptation to disability is both easy to overlook and hard to accomplish even when given adequate attention.9 Improvements in the field of neuroprognostication may not only depend on the development of new shared decision making tools for physicians and families but also on increasing awareness of the limitations of prognostic scales and the cognitive biases that may exist when discussing the possibilities of future disability.
References
1. Traumatic Brain Injury Statistics [online]. Available at: http://www.cdc.gov/traumaticbraininjury/statistics.html. Accessed Nov. 1.
3. Neurocrit Care. 2013;19:347-63.
4. Col NF. Chapter 17: Shared Decision Making. In: Communicating Risks and Benefits: An Evidence-Based User’s Guide [online]. Available at http://www.fda.gov/downloads/AboutFDA/ReportsManualsForms/Reports/UCM268069.pdf.
5. Cochrane Database Syst Rev. 2011 Oct 5;(10):CD001431.
6. PLoS Med. 2008 Aug;5(8):e165; discussion e168.
8. Neurocrit Care. 2015;23:131-41.
Dr. Muehlschlegel is associate professor of neurology (neurocritical care), anesthesia/critical care, and surgery at the University of Massachusetts, Worcester. Dr. Hwang is assistant professor of neurology in the division of neurocritical care and emergency neurology at Yale University, New Haven, Conn. Dr. Muehlschlegel reported receiving a grant from the National Institutes of Health for her research in developing a pilot decision aid for goals-of-care decisions in critically-ill TBI patients. Dr. Hwang reported receiving research funding from the American Brain Foundation, the Apple Pickers Foundation, the National Institute on Aging, and the Neurocritical Care Society.
The most difficult decisions in neuroscience intensive care units often involve patients’ ultimate goals of care. Oftentimes, family members of a brain-injured patient with an apparently poor neurologic prognosis must weigh whether their loved one would have preferred prolongation of aggressive ICU and post-ICU care, often with little to no chance for “meaningful” recovery, or death via the institution of comfort measures only. Proper prognostication is crucial to the family when making such decisions. However, the process of formulating and talking about prognosis for our most severely affected patients is subject to physician and family biases, families’ insufficient understanding of projected outcomes, and sometimes clinical nihilism by the physicians.
The process of predicting the outcomes of patients with traumatic brain injury (TBI) serves as an example of these issues. Moderate to severe TBI continues to be a leading cause of death and disability in the United States.1 Most deaths of patients with moderate to severe TBI follow decisions by doctors and families to pursue comfort care only. However, these decisions occur at a disconcertingly highly variable rate at different trauma centers, with the variation seemingly unrelated to patients’ disease severity, age, or previously diagnosed comorbidities.2 These patients are at risk for their care being influenced by a self-fulfilling prophecy: That is, the impression of a poor prognosis communicated by clinicians to a patient’s family, whether correct or incorrect, affects the aggressiveness of the care that a patient receives and determines the patient’s outcome.3
Remedying these issues through a family or health care proxy decision support intervention (“decision aid”) that could improve and standardize the way TBI prognosis is communicated may lead to better informed decisions for these critically ill patients, with potentially less decisional regret and post-ICU stress disorders in families, and decisions more in line with the patient’s values and preferences.4 A recent Cochrane review showed that for a decision aid to be effective and integrated into routine clinical care, it must contain disease-specific data tailored to patients and their families/proxies, and be simple and time efficient for physicians to use.5 Taking these factors into account, researchers at the University of Massachusetts are developing a National Institutes of Health–funded pilot decision aid for goals-of-care decisions in critically-ill TBI patients.
While the field of TBI has tools such as the IMPACT calculator that can be used to estimate a patient’s long-term prognosis based on how patients with similar clinical characteristics in large clinical databases have done, the fundamental uncertainty of prognosis remains a difficult challenge.6 Arguably, this challenge is even more daunting when estimating prognosis for patients with severe ischemic stroke and intracerebral hemorrhage (ICH). The use of ischemic stroke outcome prediction tools is complicated, as many of them are based on population databases with wide variations in whether included patients received intravenous tissue plasminogen activator, endovascular therapy, both, or neither. Furthermore, a recent study comparing the accuracy of the ICH score for predicting 3-month outcome for ICH patients to the subjective predictions of clinicians made within 24 hours of patient admission found that the educated guesses of physicians and nurses overall seemed to correlate with actual outcomes more closely than the ICH score output.7 This finding highlights the challenge of using available outcome “calculators” for individual patients in ICUs.
Ultimately, the decisions made about the goals of care for ICU patients come down not only to what their expected outcomes are, but also whether their surrogate decision makers believe that those outcomes would be acceptable to the patient.8 Potential pitfalls abound with regard to this issue as well. Decision makers are often not made aware of the fact that many times patients with significant disability may nevertheless report a reasonable quality of life. By their very nature, conversations regarding patient prognosis inevitably focus on what future disabilities one might expect; accounting for a patient’s possible adaptation to disability is both easy to overlook and hard to accomplish even when given adequate attention.9 Improvements in the field of neuroprognostication may not only depend on the development of new shared decision making tools for physicians and families but also on increasing awareness of the limitations of prognostic scales and the cognitive biases that may exist when discussing the possibilities of future disability.
References
1. Traumatic Brain Injury Statistics [online]. Available at: http://www.cdc.gov/traumaticbraininjury/statistics.html. Accessed Nov. 1.
3. Neurocrit Care. 2013;19:347-63.
4. Col NF. Chapter 17: Shared Decision Making. In: Communicating Risks and Benefits: An Evidence-Based User’s Guide [online]. Available at http://www.fda.gov/downloads/AboutFDA/ReportsManualsForms/Reports/UCM268069.pdf.
5. Cochrane Database Syst Rev. 2011 Oct 5;(10):CD001431.
6. PLoS Med. 2008 Aug;5(8):e165; discussion e168.
8. Neurocrit Care. 2015;23:131-41.
Dr. Muehlschlegel is associate professor of neurology (neurocritical care), anesthesia/critical care, and surgery at the University of Massachusetts, Worcester. Dr. Hwang is assistant professor of neurology in the division of neurocritical care and emergency neurology at Yale University, New Haven, Conn. Dr. Muehlschlegel reported receiving a grant from the National Institutes of Health for her research in developing a pilot decision aid for goals-of-care decisions in critically-ill TBI patients. Dr. Hwang reported receiving research funding from the American Brain Foundation, the Apple Pickers Foundation, the National Institute on Aging, and the Neurocritical Care Society.
