This website is intended for healthcare professionals only.

Hospital Healthcare Europe
Hospital Pharmacy Europe     Newsletter    Login            

Sedation in neurocritical care

Sedation is one of the main facilities in the intensive care unit (ICU). However, it is well known that no sedation, or minimising it, results in better outcomes, such as decreased length of hospital stay, shorter duration of mechanical ventilation, reduced healthcare costs and a reduced need for additional examinations to determine brain function. It is also related to early mobilisation of the patient, but it is unclear if it decreases delirium. Most of the review studies supporting this evidence are based on populations with non-acute brain injury. Traditionally, these patients were kept under sedation in the early phase to prevent the secondary insult.1,2
Sedation in the ICU has different roles. The drugs used are intravenous, such as propofol and midazolam.2 The objective is to control pain and anxiety, reduce agitation and achieve patient–ventilator synchrony. Nonetheless, there are situations where sedation has specific roles such as in patients with high intracranial pressure (ICP), those requiring muscular relaxation for any reason, and those with a status epilepticus.2
The trend now is to interrupt sedation as soon as possible and avoid it during the day. However, these two new settings have to be analysed and compared with the benefit they may have concerning the risk of exacerbating intracranial hypertension in patients with reduced brain compliance.1
The ideal sedative drug should be able to decrease ICP and maintain an appropriate cerebral perfusion without interfering with its autoregulation. Additionally, it should decrease the cerebral metabolic rate of oxygen and have neuroprotective and anticonvulsant properties. 
In regard to its pharmacological properties, it should have a predictable response, rapid onset and wake-up, and minimum secondary effects.2
There are no management of sedation guidelines in neurocritical patients. Each ICU bases its decisions on its own experiences and follows basic recommendations that are well accepted in the scientific community; however, we need to establish algorithms for the management of sedation these patients. 

Need for sedation 

Sedation in neurocritical patients is essential to relieve pain, anxiety, reduce ICP, decrease oxygen consumption, tolerate therapeutic manoeuvres, and improve adaptation to mechanical ventilation. However, it has its drawbacks. It prevents frequent neurological evaluation, thus decreasing prognostic capacity because clinicians are incapable of detecting changes in the brain state. 
By contrast, sedation is part of management in certain other conditions; targeted temperature management, elevated ICP and refractory status epilepticus.1 It is a first-line treatment when ICP is high, together with other measures. In most cases, elevated ICP develops within 48 hours of the brain insult; however, it can develop earlier than this. 
Prolonged deep sedation might possibly worsen cognitive results after its cessation and contribute to polyneuropathy in critical patients. Unless deep sedation or general anaesthesia is necessary, analgesia must precede sedation. Nowadays, there is a new trend to base sedation on opioids.3
The guidelines for sedoanalgesia in massive cerebral infarction establish the following recommendations:4
  • Analgesia and sedation are recommended if signs of pain, anxiety, or agitation arise (strong recommendation, very low quality of evidence)
  • The lowest possible sedation intensity and earliest possible sedation cessation is recommended, while avoiding physiological instability and discomfort (strong recommendation, very low quality of evidence)
  • The routine use of daily wake-up trials is not recommended. Caution is particularly warranted in patients prone to ICP crisis. Neuromonitoring of at least ICP and CPP is recommended to guide sedation, and daily wake-up trials should be abandoned or postponed at signs of physiological compromise or discomfort.

Sedation assessment

Neurophysiological monitoring should be considered a routine practice for neurocritical patients requiring sedation. Over-sedation increases the risk of infection, delays the removal of mechanical ventilation, and increases the length of stay in ICUs. On the other hand, infrasedation causes agitation, anxiety, and risk of accidents such as self-extubation, pulling out of catheters, discomfort, or ventilator asynchronies5 and has prevented neurologic deterioration.6,7
A number of scales to evaluate arousal, deep of sedation and response to stimuli are available:8
  • Ramsay Scale: evaluates consciousness
  • Richmond Agitation Sedation Scale (RASS): examines cognition 
  • Sedation Agitation Scale (SAS) and Motor Activity Assessment Scale (MAAS): monitor sedation and arousal. 
Their use can reduce the amount of sedatives given to achieve a specific sedation target, decreasing the number of days on mechanical ventilation and cost of hospital stay; however, no validation is available in the neuro-ICU environment.2 The Nociception Coma Scale has emerged as a valid tool to assess pain in patients with disorders of consciousness.9
When it is not possible to use these scales, for example, in patients who require muscular relaxation, then an electrophysiological endpoint must be used instead. One example is bispectral index (BIS) monitoring,10 which has made adjustments of sedation possible. A study showed that BIS values significantly correlated with RASS and SAS scores in patients with acute brain injury.11 In another study, the BIS reliably assessed sedation levels during continuous propofol infusion in the same type of patients.12
BIS was initially developed for monitoring the depth of general anaesthesia in patients without brain pathology. It is thought that the ABI (acute brain injury) may influence the BIS algorithm because of EEG changes related to the pathology itself rather than to the sedative state.The approach to sedation should first consider the severity of acute brain injury and the cerebral physiological state, mainly ICP. Figure 1 shows a possible algorithm for the management of sedation in neurointensive care units.
Figure 1: Algorithm for the management of sedation in the neurointensive care units
The first things to consider are appropriate control of pain, control of agitation and promoting ventilator synchrony. In patients with intracranial hypertension, the targets for sedation and analgesia should be titrated to control ICP and brain tissue oxygen pressure.1

Need to wake up

For the clinical assessment of neurocritical patients, interruption of continuous sedation (IS) is therefore necessary. This is usually short-term, aimed at evaluating the patients and planning further management strategies, including the definitive sedation interruption once the clinical concern and IS does not provoke patients’ distress and metabolic imbalance.2 Withdrawal of sedation and IS by daily wake-up tests might appear beneficial to these patients by allowing clinical neuro-monitoring and timely detection of warning neurological signs.13 Daily IS trials have the potential to reduce mechanical ventilation duration and the need for tracheostomy.14 These potential benefits, however, must be balanced against the risk of further cerebral haemodynamic deterioration when sedation is stopped abruptly.15 Cerebral hypoperfusion and raised ICP might result in an imbalance of energy supply and demand, especially for the injured brain and, therefore, aggravate the risk for metabolic distress and brain tissue hypoxia.16–22 IS might lead to significant ICP elevation and CPP reduction, which are more relevant in the first days after ABI than after 4–5 days.23
Avoidance of IS is recommended in all patients at risk of, or having, elevated ICP. In these patients, sedation should never be stopped abruptly but withdrawn progressively, titrating the sedation dose to ICP targets. In all other ABI patients, withdrawal should proceed as in the general ICU and daily IS in not contraindicated.1
Given that little knowledge is available about the benefits of IS in ABI, it is important to implement multi-modal monitoring in neurocritical care in order to omit IS in those patients who will potentially be harmed by the procedure.2

Sedative drugs available 

The ideal drug for sedation should have a rapid onset and rapid recovery, in order to evaluate the neurological state, as well as a predictable clearance independent of end-organ function (avoiding accumulation). It should be easy to achieve adequate levels of sedation and it would have to reduce ICP by reducing cerebral blood and cerebral vasoconstriction. As it reduces cerebral blood flow, it should decrease the metabolic rate of oxygen consumption at the same time. In addition, cerebral autoregulation should also be maintained, while having minimal cardiovascular depressant effects.24


Propofol has a rapid onset and cessation of sedation. However, it also can unpredictably accumulate after long-term use and cause prolonged sedation.25 Cerebral blood flow and ICP are decreased.26–28 Standard- and high-dose propofol infusion (2mg/kg induction bolus followed by 150–200mg/kg/min infusion) can be used as an anticonvulsivant.29–32 Propofol infusion syndrome should be detected promptly in order to start adequate treatment. The hypotension related to propofol is multifactoral, and severe propofol-associated hypotension occurred in 26.2% of patients in some studies.33–36 Weaning from mechanical ventilation occurs earlier than with midazolam.37


In general, opioids decrease the cerebral metabolic rate of oxygen, cerebral blood flow and ICP, as long as normocapnia is maintained by mechanical ventilation.2 However, Roberts et al found that morphine, fentanyl, sufentanil and alfentanil significantly increased ICP and decreased CPP and MAP after bolus administration.38


Remifentanil is a mu-opioid agonist with analgesic effects and a rapid onset and a short duration of action. It can cause decreases in both cerebral metabolic rate and ICP, with minimal changes in CPP and cerebral blood flow.39 It can facilitate frequent awakening to evaluate neurological and respiratory parameters.40


Use of benzodiazepines increases the incidence of delirium significantly. Midazolam is an appealing sedative option for its rapid onset and short duration of effect with bolus administration, making it ideal for procedural sedation. It is also a very important drug in refractory status epilepticus. Benzodiazepines increase the seizure threshold and are useful antivonvulsivants.41,42 Midazolam accumulates in adipose tissue because of its high lipid solubility, thereby prolonging time to awakening.43


The use of barbiturates has been limited owing to their undesirable side effect profile, for example, immunosupressant properties and negative ionotropic effects. Pentobarbital is an extended drug used for refractory status epilepticus or elevated ICP.44–46 They are considered second-line therapy for the control of ICP, after propofol. 


Dexmedetomidine is an alpha-2 agonist acting on the central nervous system, with a rapid onset and termination of activity. It offers mild to moderate sedation without significant respiratory depression, analgesic effects and less delirium than with benzodiazepines.47 Side effects include bradycardia and hypotension, with bradycardia being the most typical haemodynamic effect.48
Grof and Bledsoe demonstrated that neurocritically ill patients may require high doses of dexmedetomidine to achieve desired levels of sedation and to wean off adjunctive analgesic and/or sedative agents. Infusions may be started at doses from 0.4–1mg/kg/hour in neurocritical care patients to achieve target levels of sedation.44
In a study by Erdman et al, limiting dose titration to every 30 min and omitting a bolus dose resulted in no significant difference in the prevalence of hypotension or bradycardia between dexmedetomidine and propofol.49
It also appears to shorten time to extubation and enhance arousability and the patient’s ability to communicate with caregivers. Dexmedetomidine may reduce delirium after long-term sedation as compared with midazolam,50 and also reduce the overall neurocognitive adverse events of sedation, such as agitation, anxiety, and delirium, compared with propofol.51
However, safety and efficacy of this drug have not been evaluated in some ICU patient groups, such as patients with acute neurologic disorder (for example, stroke and head trauma).2

Newer alternatives

Inhalative sedation in the ICU is starting to spread all over Europe and has been recommended as an alternative in a German consensus guideline.52 The device used is called AnaConDa (Anaesthetic Conserving Device), which makes possible the administration of anaesthetic agents (isoflurane and sevoflurane)  in any ventilator commonly found in the ICU.53–55
Isoflurane, sevoflurane and desflurane have shown some benefits compared with intravenous sedation. They have a low metabolism and, due to their low solubility, are eliminated quickly and offes shorter and more predictable wake-up times than intravenous agents.54-58 They can also prevent the development of bronchospam,53,55 have cardioprotective effects54 and are haemodynamically more stable than intravenous drugs.55
Some volatile anesthetics abolish cerebral autoregulation at high doses; it has been reported that at 1.0 MAC sevoflurane, the autoregulation of cerebral blood flow remained intact, but that this was impaired at 2.0 MAC. They also have a direct neuroprotective effect in periods of in vitro59 and in vivo60 ischaemia or administered prior to it (anaesthetic conditioning).61 Preconditioning has been described in in vitro62 and in vivo63 models of cerebral ischaemia. Sevorane also has a role in post-conditioning; its application could be of interest once cerebral ischaemia has occurred.
Lee et al64 found that isoflurane post-conditioning reduced brain injury due to ischaemia in rats. However, at 0.5 MAC, adequate neuroprotection is not obtained, which means that the effect of sevoflurane post-conditioning in focal ischaemic lessions is dose-dependent.65
In a prospective study of sevoflurane sedation in patients with acute stroke or subarachnoid haemorrhage, sufficient sedation levels without clinically relevant ICP increases were achieved in 68% of patients.66 However, serious adverse events observed in the remaining 32% raise considerable safety concerns. Mean arterial pressure (MAP) had to be stabilised actively to maintain CPP. Based on these observations, it was concluded that that the alleged neuroprotective potential of sevoflurane do not outweighs the risk of adverse events and sevoflurane sedation should probably not be used in this specific patient population.66
Volatile sedation has historically been considered unsafe in neurocritical care units around the world. However, as previously mentioned and suggested in several studies, the potential neuroprotective benefit of inhalative sedation has to be more studied. 
As well as inhalative sedatives, the use of ketamine has also been debated because of the concern raised by early studies that it was associated with increased ICP.67 In studies examining the cerebral haemodynamic effects of ketamine after acute brain injury, ICP was reduced and CPP remained stable or increased, without significant changes in cerebral haemodynamics.68 A systematic review concluded that ketamine was not associated with an increased risk of ICP elevation, as said previously.69
Ketamine is a short-acting NMDA receptor antagonist with a rapid onset of action. It does not alter systemic haemodynamics or respiratory drive, therefore it can be used in non-intubated patients. At doses of 1–5mg/kg/hour, it can be used as an adjunct to other sedatives to improve their effects and thus limit drug requirements. 


Sedation and analgesia is frequently used in the management of critically ill patients and is related to a longer hospital stay and more difficult weaning from mechanical ventilation. However, in neurointensive care units, it is also a therapeutic strategy. Therefore studies have been developed to achieve efficient sedation and avoid the adverse effects as much as possible. 
Midazolam and propofol are the most frequently used first-line sedatives; however, use of benzodiazepines is less common because 
of their deleterious effects, such as prolonging mechanical ventilation time and increasing awakening times. 
New trends, such as inhalative sedation or ketamine, are beginning to garner more attention, but more studies are required to fully confirm their use.


1 Oddo M et al. Optimizing sedation in patients with acute brain injury. Critical Care 2016;20:128. 
2 Badenes R, De Fez M. Sedation in neurocritical units. In Khan Z (ed): Challenging Topics in Neuroanesthesia and Neurocritical Care. Springer International Publishing Switzerland 2017; 259–74. 
3 Brummel NE, Girard. TD. Preventing delirium in the intensive care unit. Crit Care Clin 2013:29:51–65.
4 Otterspoor LC, Kalkman CJ, Cremer OL. Update on the propofol infusión syndrome in ICU management on patients with head injury. Curr Opin Anaesthesiol 2008;21:544–51.
5 Smith HA et al. Delirium: an emerging frontier in the management of critically ill children. Crit Care Clin 2009;25:593–614. 
6 Kress JP et al. Daily interruption of sedative infusions in critically ill patients undergoing mechanical ventilation. N Engl J Med 2000;342:1471–7.
7 Hogarth DK, Hall J. Management of sedation in mechanically ventilated patients. Curr Opin Crit Care 2004;10:40–6.
8 DeJonge B et al. Using and understanding sedation scoring systems: a systematic review. Intensive Care Med 2000;26:
9 Schnakers C et al. The Nociception Coma Scale: a new tool to assess nociception in disorders of consciousness. Pain 2010;148:215–9.   
10 Riess ML et al. Sedation assessment in critically ill patients with bispectral index. Eur J Anesthesiol 2002;19:18–22. 
11 Deogaonkar A et al. Bispectral index monitoring correlates with sedation scales in brain-injured patients. Crit Care Med 2004;32:2403–6.
12 Ogilvie MP, Pereire BM, Ryan ML, et al. Bispectral index to monitor propofol sedation in trauma patients. J Trauma 2011;71:1415–21. 
13 Skoglund K et al. The neurological wake up test increases stress hormone levels in patients with severe traumatic brain injury. Crit Care Med 2012;40:216–22.
14 Burry L et al. Daily sedation interruption versus no daily sedation interruption for critically ill adults patients requiring invasive mechanical ventilation. Cochrane Database Syst Rev 2014;7:CD009176.
15 Heibok R et al. Effects of the neurological wake-up test on clinical examination, intrancranial pressure, brain metabolism and brain tissue oxygenation in severely brain-injured patients. Crit Care 2012;16:R226. 
16 Penson RT et al. Randomized placebo-controlled trial to the activity of the morphine glucuronides. Clin Pharmacol Ther 2000;68:667–76.
17 Ickx B et al. Propofol infusion for induction and maintenance of anaesthesia in patients with end-stage real disease. Br J Anaesth 1998;81:854–60. 
18 Hiraoka H et al. Changes in drug plasma concentrations of an extensively bound and highly extracted drug, propofol, in response to altered plasma binding. Clin Pharmacol Ther 2004;75:324–30.
19 Servin F et al. Pharmacokinetics of propofol administered by continuous infusion in patients with cirrosis. Preliminary results. Anaesthesia 1988;43:23–4.
20 Pentikainen PJ et al. Pharmacokinetics of midazolam following intravenous and oral administration in patients with chronic liver disease and in healthy subjects. J Clin Pharmacol 1989;29:272–7.
21 Chauvin M et al. Sufentanil pharmacokinetics in patients with cirrhosis. Anesth Analg 1989;68:1–4.
22 Mazoit JX et al. Pharmacokinetics of unchanged morphine in normal and cirrhotic subjects. Anesth Analg 1987;66:296–8. 
23 Skoglund K, Enblad P, Marklund N. Effects of the neurological wake up test in intracranial pressure and cerebral perfusion pressure in brain injured patients. Neurocrit Care 2009;11:135–42. 
24 Citerio G, Cormio M. Sedation in neurointensive care: advances in understanding and practice. Curr Opin Crit Care 2003;9:120–6.
25 Barr J et al. Propofol dosing regimens for ICU sedation based upon an integrated pharmacikenetic-pharmacodynamic model. Anesthesiology 2001;95:324–33.
26 Oshima T, Karasawa F, Satoh T. Effects of propofol on cerebral blood flow and the metabolic rate of oxygen in humans. Acta Anesthesiol Scand 2002;46:831–5.
27 Engelhard K, Werner C. Inhalational or intravenous anesthetics for craniotomies? Pro inhalational. Curr Opin Anaesthesiol 2006;19:504–8.
28 Trescot AM et al. Opioid pharmacology. Pain Physicia 2008;11: S133–53.
29 Weinbroum AA et al. Midazolam versus propofol for long-term sedation in the ICU: a randomized prospective comparison. Intensive Care Med 1997;23:1258–63. 
30 Power KN et al. Propofol treatment in adult refractory status epilecpticus. Mortality risk and outcome. Epilepsy Res 2011;94:53–60. 
31 Meierkord H et al. European Federation of Neurological Societies. EFNS guideline on the management of status epilepticus in adults. Eur J Neurol 2010;17:348–55. 
32 Marik PE, Varon J. The management of status epilepticus. Chest 2004;126:582–91. 
33 Ebert TJ et al. Sympathetic responses to induction of anesthesia in humans with propofol or etomidate. Anesthesiology1992;76:725–33.
34 Gonzalez-Correa JA et al. Effects of propofol on the leukocyte nitric oxide pathway: in vitro and ex vivo studies in surgical patients. Naunyn Schmiedeberg’s Arch Pharmacol 2008;376:331–9.
35 Dorsout MF et al. Role of propofol and its solvent, intralipid, in nitric oxide induced peripherial vasodilation in dogs. Br J Anesth 2002;89:492–8.
36 Samain E et al. The effect of propofol on angiotensin II-induced Ca 2(+) mobilization in aortic smooth muscle cells from normotensive and hypertensive rats. Anesth Analg 2009;90:546–52. 
37 Hutchens MP, Memtsoudis S, Sadovnikoff N. Propofol for sedation in neurointensive care. Neurocrit Care 2006;4:54–62.
38 Roberts DJ et al. Sedation for critically ill adults with severe traumatic brain injury: a systematic review of randomized controlled trials. Cric Care Med 2011;39:2743–51.
39 Peeters MY et al. Diseases severity is a major determinant for the pharmacodynamics of propofol in critically ill patients. Clin Pharmacol Ther 2008;83:443–51. 
40 Karabinis A et al. Safety and efficacy of analgesia-based sedation with remifentanil versus standard hypotic-based regimens in intesive care patients with brain injuries: a randomised, controlled trial. Crit Care 2004;8:R268–80.
41 Hanley DF, Pozo M. Treatment of status epilepticus with midazolam in the critical care setting. Int J Clin Pract 2000;54:30–5.
42 Walder B, Tramer MR, Seeck M. Seizure-like phenomena and propofol : a systematic review. Neurology 2002;58:1327–32. 
43 Bauer TM et al. Prolonged sedation due to accumulation of conjugated metabolites of midazolam. Lancet 1995;346:145–7. 
44 Grof TM, Bledsoe KA. Evaluating the use of dexmedetomidine in neurocritical care patients. Neurocrit Care 2010;12:356–61. 
45 Chen HI et al. Barbiturate infusion for intractable intracranial hypertension and its effect on brain oxygenation. Neurosurgery 2008;63:880–6. 
46 Marshall GT et al. Pentobarbital come for refractory intra-cranial hypertension after severe traumatic brain injury: mortality predictions and one-year outcomes in 55 patients. J Trauma 2010;69:275–83. 
47 Yahwak JA et al. Determination of a lorazepam dose threshold for using the osmol gap to monitor the propylene glycol toxicity. Pharmacotherapy 2008;28:984–91. 
48 Jakob SM et al. Dexmedetomidine vs. midazolam or propofol for sedation during prolonged mechanical ventilation: two randomized controlled trials. JAMA 2012;307:1151–60. 
49 Erdman M et al. A comparison of severe hemodynamic disturbances between dexmedetomidine and propofol for sedation in neurocritical care patients. Crit Care Med 2014;42(7):1696–702. 
50 Riker RR et al. SEDCOM (Safety and Efficacy of Dexmedetomidine Compared with Midazolam) Study Group: Dexmedetomidine vs. midazolam for sedation of critically ill patients: a randomized trial. JAMA 2009;301:489–99. 
51 Jakob SM et al. Dexmedetomidine for long-term sedation investigators: Dexmedetomidine vs. midazolam or propofol for sedation during prolonged mechanical ventilation: two randomized controlled trials. JAMA 2012;307:1151–60. 
52 Martin J et al. Evidence and consensus-based German guidelines for the management of analgesia, sedation and delirium in intensive care-short versión. Ger Med Sci 2010:8;Doc02. 
53 Millane TA, Bennet E, Grounds RM. Isoflurane and propofol for long-term sedation in the intensive care unit a crossover study. Anaesthesia 1992;47:768–74. 
54 Spencer EM, Willats SM. Isoflurane for prolonged sedation in the intensive care unit: efficacy and safety. Intensive Care Med 1992;18:415–21.
55 Meiser A et al. Desflurane compared with propofol for postoperative sedation in the intensive care unit. Br J Anaesth 2003;90:273–80.
56 Halpenny D. Sevoflurane sedation. Can J Anaesth 2000;47:193–4.
57 Ibrahim AE et al. Speed of recovery and side-effect profile of sevoflurane sedation compared with midazolam. Anesthesiology 2001;94:87–94. 
58 Kong KL, Willats SM, Prys-Roberts C. Isoflurane compared with midazolam for sedation in the intensive care unti. BMJ 1989:298:1277–80.
59 Kalenka A, Gross B, Isoflurane MH. Anesthesia elicits protein pattern changes in rat hippocampus. J Neurosurg Anesthesiol 2010;22:144–54. 
60 Tanzi RE, Bertram L. Twenty years of the Alzheimer´s disease amyloid hypothesis: a genetic perspective. Cell 2005;120:545–55. 
61 Selkoe DJ. Alzheimer`s disease: genes, proteins and therapy. Physiol Rev 2001;81:741–66. 
62 Eckenhoff RG et al. Inhaled anesthetic enhancement of amyloid-beta oligomerization and cytotoxicity. Anesthesiology 2004;101:703–9.
63 Xie Z et al. The common inhalation anesthetic isoflurane induces apoptosis and increases amyloid beta protein levels. Anesthesiology 2006;104:988–94. 
64 Lee JJ, Li L, Jung HH, Zuo Z. Postconditioning with isoflurane reduced ischemia-induced brain injury in rats. Anesthesiology 2008;108:1055–62. 
65 Xie Z et al. The inhalation anesthetic isoflurane induces a vicious cycle of apoptosis and amyloid beta-protein accumulation. J Neurosci 2007;27:1247–54. 
66 Purruker JC, Renzland J, Uhlmann L. Volatile sedation with sevoflurane in intensive care patients with acute stroke or subarachnoid haemorrhage using AnaConDa®: an observational study. Br J Anaesth 2015;114:934–43.
67 White PF, Way WL, Trevor AJ. Ketamine – its pharmacology and therapeutic uses. Anesthesiology 1982;56:119–36. 
68 Zeiler FA et al. The ketamine effect on intracranieal pressure in nontraumatic neurological illness. J Crit Care 2014;29:1096–106. 
69 Himmelseher S, Durieux ME. Revising a dogma: ketamine for patients with neurological injury? Anesth Analg 2005;101:524–34.