Acta Anaesthesiol Scand 2009; 53: 935–940 Printed in Singapore. All rights reserved
r 2009 The Authors Journal compilation r 2009 The Acta Anaesthesiologica Scandinavica Foundation ACTA ANAESTHESIOLOGICA SCANDINAVICA
doi: 10.1111/j.1399-6576.2009.02020.x
Prognostic factors and outcome after drowning in an adult population ´ . BALLESTEROS1, M. GUTIE´RREZ-CUADRA2, P. MUN˜OZ3 and E. MIN˜AMBRES1 M. A
Departments of 1Critical Care Medicine, Servicio de Medicina Intensiva and 2Internal Medicine, Hospital Universitario Marque´s de Valdecilla, Santander, Spain and 3Servicio Ca´ntabro de Salud, Santander, Spain
Background: Drowning remains an actual problem. Although medical assistance has improved, it still has high rates of morbidity and mortality. We set out to explore the clinical characteristics and outcome of drowning patients admitted to the intensive care unit (ICU) of tertiarycare university hospital. Methods: We designed a retrospective observational study to analyse all drowning patients admitted to our ICU after successful cardiopulmonary resuscitation. The study was conducted during 1 January 1992–31 December 2005. There was no exclusion. We used a univariate analysis to evaluate the effect on patient and management characteristics on survival. Results: There were 43 patients (five children and 38 adults), with male predominance. Fifteen patients, all adults (34.9%), died. Submersion time, age, Glasgow
Coma Score (GCS), pupillary reactivity and acute physiology and chronic health evaluation (APACHE II) at ICU admission were related to mortality. Non-survivors presented a higher glycaemia level at ICU admission than survivors (P 5 0.005). Conclusions: The outcome is closely related to the patient’s clinical status on arrival to the hospital. We have found that submersion time, age, GCS, pupillary reactivity and APACHE II at ICU admission were related to mortality. Further research in prospective studies is needed.
D
Despite the alarmingly high number of drownings and near-drownings that occur worldwide, the number of articles on the subject is small and mainly focused on the paediatric population. Most studies relate to epidemiology, prevention and forensic aspects, but new research data are limited. In fact, there are few population-based surveillance studies on drowning incidents or prospective clinical studies of prognostic factors and outcomes of drowning, especially in the adult population.6 The advances in critical care have provided new, more aggressive management of these patients. However, attempts at management of intracranial pressure or specific neuro-resuscitative strategies have been unsuccessful in improving outcome.7,8 Comatose patients who have been resuscitated after ‘reasonable sub-mersion time’ regardless of the rhythm should not be rewarmed 434 1C. If the patient is not already hypothermic, hypothermia treatment could be considered based on extrapolation from what is known from out-of-hospital cardiac arrest with the following global cerebral
ROWNING and near-drowning are major global public health problems. Nearly 500,000 people worldwide die annually from drowning.1 Death by submersion represents the second most common cause of accidental death in children, only exceeded by motor vehicle accidents.2,3 One of the most significant characteristics of this clinical entity is the absence of international criteria to describe this concept. For this reason, not all drowning cases are included in these statistics because of a lack of an appropriate and internationally accepted definition that permits all relevant cases to be counted.4 This variability in the definitions and outcomes makes it very difficult to assess and analyse studies, either individually or as a whole, and draw any conclusions that will influence future practices. Recently, a new internationally accepted definition of drowning was developed.1 The American Heart Association has provided standardized criteria to improve vigilant measures, to facilitate prevention programmes and to enhance the collection of epidemiological data.5
Accepted for publication 7 April 2009 r 2009 The Authors Journal compilation r 2009 The Acta Anaesthesiologica Scandinavica Foundation
935
M. A. Ballesteros et al.
ischaemia–reperfusion injury.9–12 Nevertheless, future search is needed in this respect. Special efforts have been made to establish the prognosis of near-drowning patients at arrival to emergency services, but still most attempts have been made in the paediatric population. Nevertheless, there is no one clinical score that is applied in the accident location or during resuscitation that predicts with 100% accuracy, as to which patient will survive the drowning event. The aim of this study was to evaluate the profile of our near-drowning population and the influence of several clinical variables on outcome in all neardrowning victims admitted to our intensive care unit (ICU) during the last 14 years.
Methods The notes of all successfully resuscitated submersion victims who sustained either a cardiac arrest (confirmed by emergency medical systems), or required bystander cardiopulmonary resuscitation (CPR) at the scene, were analysed retrospectively. Patients were admitted to either the paediatric intensive care unit or the adult ICU at Marque´s de Valdecilla University Hospital (a third-level hospital with 1000 beds; 50 ICU beds). This hospital services a regional population of 500,000 inhabitants. It is located in Cantabria, a coastal region in the north of Spain, which is a common tourist destination and location for water activities in the summer months. The study period was from January 1992 to December 2005. In all cases, first assistance was made on site by the emergency services. Once stabilized, they were sent to our ICU. All cases received standard resuscitation controlling haemodynamic parameters and respiratory support was provided if needed.13 The pre-hospital care was carried out by basic life-supported units and advanced life-supported units depending on the cases. In Cantabria, there are four physician staff pre-hospital emergency care units available 24 h a day since 1992. Clinical data were recorded at ICU admission and collected in a retrospective way by means of analysis and revision of clinical charts. Neurological status at ICU admission was evaluated by the Glasgow Coma Score (GCS)14 and pupillary response (presence or absence of pupillary reactivity). Outcome was evaluated at ICU discharge by mortality. Neurological status from the survivors was evaluated 3 months after drowning by the
936
Glasgow Outcome Scale (GOS).15 Water temperature was assessed by going back to the local registry of the National Weather Institute Agency, which reports the water temperature daily. The Local Ethics Committee discussed the matter and did not consider patients’ consents necessary. Statistical analysis was performed using the statistic software SPSS 11.5. The U-Mann–Whitney test and the w2 test were performed according to the variable characteristics. The level of statistical significance was fixed to Po0.05.
Results Forty-three patients were finally analysed, with males predominating [32 patients (74.4%)]. There were five (11.5%) children (age o14), 29 (67.4%) adults and nine (20.7%) elderly individuals (age 65). Most of the submersion incidents [29 cases (67.4%)] occurred during summer months (June– August). The event was witnessed in 27 cases (62.8%). Most of the submersion incidents were in salt water [38 cases (88.4%)]. The average time of arrival to the emergency room was 30 min (range: 10–83 min) and the average time of arrival to ICU was 90 min (range: 40–210 min). Twenty-six patients received bystander CPR at the scene and 21 patients were in cardiac arrest at the time of arrival of the first emergency medical service (EMS) unit. Out of 26 patients receiving bystander CPR, 22 demonstrated restoration of spontaneous circulation when the first EMS unit arrived. Four patients receiving CPR bystander CPR also received CPR by EMS staff. The demographic characteristics of the submersion victims are summarized in Table 1. The initial cardiac rhythm was asystole in 15 patients, pulseless electrical activity in four and ventricular fibrillation in one. Twenty-one patients were intubated at the scene. Three patients were Table 1 Characteristics of the population at ICU admission. Age Gender (male) GCS APACHE II Submersion time (minutes) Water temperature (1C) Central body temperature (at ICU)
42.9 22.9 32 (74.41%) 10.7 5.3 14.7 9.9 6.2 5.1 19.1 2.8 34.9 1.7
All results are expressed in mean standard deviation except gender (absolute number and percentage). APACHE II, acute physiology and chronic health evaluation; GCS, Glasgow Coma Score; ICU, intensive care unit.
Prognostic factors and outcome
transported with on-going CPR to the emergency room. Standard resuscitation and support measures were used. The most common re-warming method was external re-warming with thermal blankets [30 cases (69.8%)]. Mechanical ventilation was used in 21 patients (48.8%). Prophylactic antibiotics were used in 27 (62.8%) patients to avoid infectious complications. We did not find any differences between the use/non-use of antibiotics in terms of septic complications or outcome (data not shown). The average ICU stay was 4.5 days. Those patients who had, most likely, had only a respiratory arrest (only bystander CPR) all survived. Of those who received CPR by EMS staff (most likely had a cardiac arrest) 42% survived. The neurological status of survivors was evaluated by the GOS based on the data 3 months after the incident. Most survivors [23 cases (82.1%)] had a good recovery (GOS 5). Moderate disability (GOS 4) was observed in four (14.3%) of the survivors and a vegetative state (GOS 2) was observed in one of them (3.5%). In addition, neurologic outcome using GOS was evaluated separately according to whether patients had received CPR by the EMS staff or not. We observed that of 21 patients receiving CPR by EMS staff, five (23.8%) had a good recovery, two (9.5%) suffered a moderate disability and 15 (71.4%) died. Of 22 patients not receiving CPR by EMS staff, 18 (81.8%) had good recovery, two (9.1%) suffered a moderate disability and two (9.1%) were in a vegetative state. The mortality rate was 34.9% (15 patients died). The median interval between ICU admission and death was 4.5 days (range: 1–85 days) in nonsurvivors. All patients who died were adults or elderly people. None of the children analysed died. The main causes of mortality observed were new cardiac arrest [five cases (33.3%)] and post-anoxic encephalopathy [four cases (26.7%)]. Other causes
of death were multiorganic failure [one case (6.7%)], encephalic death [one case (6.7%)], pneumonia [one case (6.7%)], heart failure [one case (6.7%)] and unclear cause [two cases (13.3%)]. The average submersion time for those who died was 12 4.2 min compared with 4.5 3.8 min for those who survived (P 5 0.0001). Non-survivors had worse clinical and analytical parameters at ICU admission (Table 2). We observed that patients with a GCS score lower than or equal to 5 had a higher mortality rate than those patients with a GCS45 (92.3% vs. 6.9%; P 5 0.0001). According to univariate statistic analysis, age, submersion time, water temperature, glycaemia at ICU admission, GCS, acute physiology and chronic health evaluation (APACHE II) score and the absence of pupillary reactivity were related to ICU mortality (Table 3). For instance, an increase of 10 years in age (years) is associated with a 1.51-fold risk of a fatal outcome. We observed, after stabilization in ICU, 27 patients (65.2%) did not show electrocardiographic abnormalities. Table 2 Clinical and analytical parameters of the population at ICU admission.
Heart rate Systolic tension (mmHg) Diastolic tension (mmHg) Haematocrit (%) Haemoglobin (g/dl) Glycaemia (mg/dl) Creatinine (mg/dl) pH Presence alveolar infiltrates Witnessed event
Survivors
Non-survivors
99.7 19.2 120 23 66 13 42 4.7 13.7 2 132 36 0.9 0.3 7.15 0.2 16 (80%)
78.8 29 82 26 44 17 38.7 3.6 12.5 1.5 253 107 1.6 0.7 6.94 0.2 4(20%)
P o0.001 o0.0001 o0.001 0.012 0.038 0.005 o0.0001 0.034 0.148
21 (77.8%) 6(22.2%)
0.268
All results are expressed in mean standard deviation. ICU, intensive care unit.
Table 3 Univariant analysis of the main variables. N Constant Age (years) Submersion time (minutes) Water temperature (1C) Glycaemia (mg/dl) GCS (points) No pupillary reactivity APACHE II score (points)
43 30 22 34 42 41 34
b 0.041 0.350 0.014 3.376 0.296
SE
OR (95% CI)
0.320 0.017 0.138 0.253 0.006 0.113 0.948 0.091
1.04 1.42 0.61 1.01 0.641 29.25 1.34
(1.01–1.08) (1.08–1.86) (0.37–0.99) (1.00–1.03) (0.51–0.80) (4.56–187.70) (1.13–1.61)
P 0.018 0.011 0.047 0.013 o0.005 o0.005 0.001
OR is expressed with respect to each increase or decrease in the unit of measurement in which the variable is expressed. APACHE II, acute physiology and chronic health evaluation; GCS, Glasgow Coma Score; b, logistic regression coefficient; SE, standard error; OR, odds ratio.
937
M. A. Ballesteros et al.
Thirty patients (69.8%) evidenced thoracic radiological alterations, emphasizing 50% of patients with alveolar infiltrates. Moreover, this pulmonary alteration was not related to water type (fresh/salt) in a statistical manner (50% vs. 60%).
Discussion This study is the first study carried out in Southern Europe about drowning in the context of ICU. Near-drowning is a current problem and should be recognized as a major public health problem with significant opportunities for prevention. There are few population-based surveillance studies on drowning incidents or prospective clinical studies of prognostic factors and outcomes of drowning events, especially in an adult population.16 Although there have been several attempts, the lack of standardized definitions of drowning makes the analysis of these studies very difficult.1 In our series, the severity of this clinical entity is reflected. We found a mortality rate of 34.9%; this rate is higher than previously described in the literature. This high mortality can be explained by the fact that our study was mainly focused on an adult population instead of a paediatric population, and mortality was related to age (see Table 3). In fact, no child died in our series. Nevertheless, Suominen et al.17 observed a mortality rate similar to ours, in a study with 61 patients (35 adults). The submersion time should be the most important prognostic factor after a near-drowning incident. We observed that the submersion time was correlated with survival as in previous reports.17 However, this factor is sometimes difficult to ascertain. Frequently, the time the victim was underwater is unknown and it needs to be estimated.16 There is no clear cut-off value for the submersion time, although 410 min is considered a cut-off point for non-survival.18 In addition, it is difficult to separate the effect of submersion time from other confounding factors, such as water temperature and efficiency of initial resuscitation. The majority of our drowning population was in saltwater. Although evidence suggests that aspirated fresh water is more likely to produce pulmonary oedema than seawater, we found similar results between both types of water. There are even reports about the use of a surfactant after freshwater near-drowning in order to treat the adult respiratory distress syndrome.19 The water temperature is another aspect to consider. Cold water could play a protective role
938
against hypoxic brain damage, diminishing the metabolic rate and it explains why survival is much better following resuscitation after drowning rather than after other aetiologies.20,21 Specifically, hypothermia reduces the cerebral metabolism by decreasing oxygen consumption, glucose utilization and lactate concentration. In our series, no comatose drowning patient was treated with moderate hypothermia but hypothermia could be a neuroprotective tool in the management of drowning patients with brain damage. If the patient is not already hypothermic, hypothermia treatment could be considered based on extrapolation from what is known from an out-of-hospital cardiac arrest with the following global cerebral ischaemia–reperfusion injury. Although therapeutic hypothermia is in full development, and therapeutic hypothermia is implemented in the daily clinical practice of an increasing number of ICU,22 the effectiveness of induced hypothermia for drowning victims has not been established, and evaluation of this approach is warranted. Some authors have reported that induction of mild hypothermia should be considered in cardiac arrest survivors12,23–25. The Scandinavian Society of Anaesthesiology and Intensive Care Medicine also recommends therapeutic hypothermia after restored spontaneous circulation in cardiac arrest survivors (patients with initial ventricular fibrillation, initial pulseless electrical activity and asystole).26 Much of the clinical literature on drowning has focused on predictors of outcome.21,27,28 Nichter et al.27 carried out a retrospective review and they found that sex, age, time of submersion, central temperature, pH, lack of response to pain and pupillary non-reactivity were independent predictors of outcome. We observed than non-survivors had higher levels of glycaemia than survivors as in other paediatric drowning patients.29 Recent findings support that hyperglycaemia can adversely affect the outcome in critically ill patients. In fact, hyperglycaemia after cerebral ischaemia exacerbates brain injury and worsens the outcome of stroke patients.30 Some papers have shown that hyperglycaemia is a marker of a severe physiologic insult after injury and that strict glycaemic control may improve outcome.31 Although intensive glycaemic control has been advocated for improving the outcome in these patients, the results remain unclear. Hypoxia is the most important factor related to outcome and subsequent quality of life in drowning victims. The neurological status at ICU admission evaluated by GCS and pupillary response has
Prognostic factors and outcome
been related to a poor outcome.28,32 Lavelle and Shaw,28 in a 6-year retrospective review of 44 neardrowning cases, observed that non-reactive pupils and a GCS o5 on arrival to the ICU were the best independent predictors of a poor neurological outcome. Our results are equivalent to these studies because we observed that GCS and the absence of the pupillary response were related to mortality, and GCS 5 was associated with a poor neurological outcome. Both GCS and pupillary response reflect a hypoxic brain injury. In fact, hypoxia participates in the secondary brain injury and it is related to cell death and prognosis.33,34 Therefore, this emphasizes the importance of hypoxic brain injury for survival, which is related to submersion time and a rapid assistance by medical services. The importance of aggressive pre-hospital care must be emphasized, because no positive effects on neurological outcome in near-drowning victims have been provided with late therapies.35,36 The few number of patients included in our study did not allow us to carry out a multivariate statistical analysis.
Conclusions Even though our study is limited by its retrospective nature, the small number of patients included and the type of people included, mainly adults, we found that submersion time, age, GCS, pupillary reactivity and APACHE II at ICU admission were related to mortality. The first steps in treatment are a quick assistance and prompt resuscitation to preserve the organism of hypoxic damage and treat secondary complications. Our results suggest that drowning outcome is closely related to the patient’s clinical status on arrival to the hospital. Further research in prospective studies is needed.
Acknowledgements We would like to thank Dr Antonio Gonza´lez for his help with the preparation of this paper. Competing Interest: None declared.
References 1. van Beeck EF, Branche CM, Szpilman D, Modell JH, Bierens JJ. A new definition of drowning: towards documentation and prevention of a global public health problem. Bull World Health Organ 2005; 83: 853–6.
2. Sibert JR, Lyons RA, Smith BA, Cornall P, Sumner V, Craven MA, Kemp AM. Preventing deaths by drowning in children in the United Kingdom: have we made progress in 10 years? Population based incidence study. BMJ 2002; 324: 1070–1. 3. Centers for Disease Control and Prevention (CDC). Nonfatal and fatal drownings in recreational water settings – United States, 2001–2002. Morb Mortal Wkly Rep 2004; 53: 447–52. 4. Papa L, Hoelle R, Idris A. Systematic review of definitions for drowning incidents. Resuscitation 2005; 65: 255–64. 5. American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Drowning. Circulation 2005; 112: IV-133–5. 6. Idris AH, Berg RA, Bierens J, Bossaert L, Branche CM, Gabrielli A, Graves SA, Handley AJ, Hoelle R, Morley PT, Papa L, Pepe PE, Quan L, Szpilman D, Wigginton JG, Modell JH. Recommended guidelines for uniform reporting of data from drowning: the ‘‘Utstein style’’. Resuscitation 2003; 59: 45–57. 7. Ibsen LM, Koch T. Submersion and asphyxial injury. Crit Care Med 2002; 30 (Suppl.): S402–8. 8. Hasibeder WR. Drowning. Curr Opin Anaesthesiol 2003; 16: 139–46. 9. Bernard SA, Gray TW, Buist MD, Jones BM, Silvester W, Gutteridge G, Smith K. Treatment of comatose survivors of out-of-hospital cardiac arrest with induced hypothermia. N Engl J Med 2002; 346: 557–63. 10. The Hypothermia after Cardiac Arrest Study Group. Mild therapeutic hypothermia to improve the neurologic outcome after cardiac arrest. N Engl J Med 2002; 346: 557–63. 11. Min˜ambres E, Holanda MS, Domı´nguez MJ, Rodrı´guezBorrega´n JC. Therapeutic hypothermia in neurocritical patients. Med Intensiva 2008; 32: 227–35. 12. Hovdenes J, Laake JH, Aaberge L, Haugaa H, Bugge JF. Therapeutic hypothermia after out-of-hospital cardiac arrest: experiences with patients treated with percutaneous coronary intervention and cardiogenic shock. Acta Anaesthesiol Scand 2007; 51: 137–42. 13. American Heart Association; Emergency Cardiac Care Committee and Subcommittees. Guidelines for cardiopulmonary resuscitation and emergency care. JAMA 1992; 268: 2171–302. 14. Teasdale G, Jennett B. Assessment of coma and impaired consciousness. Lancet 1974; 2: 81–4. 15. Jennett B, Bond M. Assessment of outcome after severe brain damage. Lancet 1975; 1: 480–4. 16. Nolan JP, Morley PT, Vanden Hoek TL, Hickey RW, Kloeck WG, Billi J, Bo¨ttiger BW, Morley PT, Nolan JP, Okada K, Reyes C, Shuster M, Steen PA, Weil MH, Wenzel V, Hickey RW, Carli P, Vanden Hoek TL, Atkins D. Therapeutic hypothermia after cardiac arrest: an advisory statement by the Advanced Life Support Task Force of the International Liaison Committee on Resuscitation. Circulation 2003; 108: 118–21. 17. Suominen P, Baillie C, Korpela R, Rautanen S, Ranta S, Olkkola KT. Impact of age, submersion time and water temperature on outcome in near-drowning. Resuscitation 2002; 52: 247–54. 18. Bierens JJ, Knape JT, Gelissen HP. Drownings. Curr Opin Crit Care 2002; 8: 578–86. 19. Onarheim H, Vik V. Porcine surfactant (Curosurf) for acute respiratory failure after near-drowning in 12 year old. Acta Anaesthesiol Scand 2004; 48: 778–81. 20. Goleen F. Mechanisms of body cooling in submerged victims. Resuscitation 1997; 35: 107–9.
939
M. A. Ballesteros et al. 21. Quan L, Wentz KR, Gore EJ, Copass MK. Outcome and predictors of outcome in pediatric submersion victims receiving prehospital care in King County, Washington. Pediatrics 1990; 86: 586–93. 22. Oksanen T, Pettila¨ V, Hynynen M, Varpula T. Therapeutic hypothermia after cardiac arrest: implementation and outcome in Finnish intensive care units. Acta Anaesthesiol Scand 2007; 51: 866–71. 23. Busch M, Soreide E, Lossius HM, Lexow K, Dickstein K. Rapid implementation of therapeutic hypothermia in comatose out-of-hospital cardiac arrest survivors. Acta Anaesthesiol Scand 2006; 50: 1277–83. 24. Skulec R, Kovarnik T, Dostalova G, Kolar J, Linhart A. Induction of mild hypothermia in cardiac arrest survivors presenting with cardiogenic shock syndrome. Acta Anaesthesiol Scand 2008; 52: 188–94. 25. Nordmark J, Rubertsson S, Mo¨rtberg E. Intracerebral monitoring in comatose patients treated with hypothermia after a cardiac arrest. Acta Anaesthesiol Scand 2008; 53: 289–98. 26. Castre´n M., Silfvast T, Rubertsson S, Niskanen M, Valsson F, Wanscher M, Sunde K. Scandinavian Clinical practice guidelines for therapeutic hypothermia and post-resuscitation care after cardiac arrest. Acta Anaesthesiol Scand 2009; 53: 280–8. 27. Nichter MA, Everett PB. Childhood near-drowning: is cardiopulmonary resuscitation always indicated? Crit Care Med 1989; 17: 993–5. 28. Lavelle JM, Shaw KN. Near drowning: is emergency department cardiopulmonary resuscitation or intensive care unit cerebral resuscitation indicated? Crit Care Med 1993; 21: 368–73. 29. Ashwal S, Schneider S, Tomasi L, Thompson J. Prognostic implications of hyperglycemia and reduced cerebral blood flow in childhood near-drowning. Neurology 1990; 40: 820–3.
940
30. Gentile NT, Seftchick MW, Huynh T, Kruus LK, Gaughan J. Decreased mortality by normalizing blood glucose after acute ischemic stroke. Acad Emerg Med 2006; 13: 174–80. 31. Sperry JL, Frankel HL, Vanek SL, Nathens AB, Moore EE, Maier RV, Minei JP. Early hyperglycemia predicts multiple organ failure and mortality but not infection. J Trauma 2007; 63: 487–93. 32. Bratton SL, Jardine DS, Morray JP. Serial neurologic examinations after near drowning and outcome. Arch Pediatr Adolesc Med 1994; 148: 167–70. 33. Ballesteros MA, Lo´pez Hoyos M, Mun˜oz P, Marin MJ, Min˜ambres E. Apoptosis of neuronal cells induced by serum of patients with acute brain injury: a new in vitro prognostic model. Intensive Care Med 2007; 33: 58–65. 34. Min˜ambres E, Ballesteros MA, Mayorga M, Marin MJ, Mun˜oz P, Figols J, Lo´pez-Hoyos M. Cerebral apoptosis in severe traumatic brain injury patients: an in vitro, in vivo, and postmortem study. J Neurotrauma 2008; 25: 581–91. 35. Salomez F, Vincent JL. Drowning: a review of epidemiology, pathophysiology, treatment and prevention. Resuscitation 2004; 63: 261–8. 36. Va¨yrynen T, Kuisma M, Ma¨a¨tta¨ T, Boyd J. Medical futility in asystolic out-of-hospital cardiac arrest. Acta Anaesthesiol Scand 2008; 52: 81–7.
Address: ´ ngeles Ballesteros Marı´a A Department of Critical Care Medicine Servicio de Medicina Intensiva Hospital Universitario Marque´s de Valdecilla Av Valdecilla s/n. 39008 Santander Spain e-mail:
[email protected]