European Heart Journal Advance Access originally published online on February 21, 2007
European Heart Journal 2007 28(6):773-774; doi:10.1093/eurheartj/ehl537
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Predicting survival with good neurological recovery at hospital admission after successful resuscitation of out-of-hospital cardiac arrest: the OHCA score: reply
Intensive Care Unit
Delafontaine Hospital
2 rue du Dr Delafontaine
Saint Denis 93205
France
Intensive Care Unit
Jacques Cartier Institute
Avenue Noyer Lambert
Massy 91300
France
Tel: +33 142356107 Fax: +33 142356233 E-mail address: christophe.adrie{at}wanadoo.fr
Tel: +33 160136272 Fax: +33 160136278 E-mail address: m.monchi{at}free.fr
The main concern raised by Sunde et al. is a very important one and points to a need for bearing in mind the message of our manuscript.1 Under no circumstances should any scoring system (including the OHCA score) be used to help predict survival in the individual patient. As stated in our discussion (page 2844), ‘The probability predicted by our score is the probability in the average patient, not the individual patient’. Decisions to withdraw life support are never taken after 24 h only on the basis of the SAPS or APACHE score,2,3 which are widely used in ICU patients; similarly, such decisions should never be taken on the basis of the OHCA score at admission.1 The OHCA score merely estimates the probability of a good (or poor) outcome. It is a valuable tool for designing epidemiological studies, constituting a group of patients whose disease is within a predefined severity range, or adjusting data on the basis of severity. The OHCA should never be used to take a decision in an individual patient. Therefore, the criticism by Sunde et al. of our statistical analysis is inappropriate: a receiver-operating characteristic value of 0.88 indicates a very good predictive ability4–6 but, of course, does not indicate usefulness for predicting outcomes in individual patients. This issue is discussed in detail in a very recent review by Le Gall,7 one of the designers of the SAPS.2
We designed a classic outcome prediction model on the basis of a development cohort and then we evaluated it in a validation cohort according to a process that has been used for many other scores.2,4,5 Its originality lies in its continuous nature, which avoids the class jump phenomenon seen with some severity scores. We do not know whether this score can be extrapolated to other healthcare systems; this point needs to be evaluated. Adjustments according to the healthcare system and improvements in patient management over time may be required, as with the SAPS (for which the third version was introduced recently).
In their retrospective study comparing patients treated with hypothermia to historical controls, Oddo et al.8 found evidence that hypothermia might substantially improve outcomes of patients with cardiac arrest due to ventricular fibrillation. However, they stated that ‘The outcome after cardiac arrest due to nonventricular fibrillation rhythms was poor and did not differ significantly between the two groups’. Although we also use therapeutic hypothermia in the most severe cases, there is no proof to date that this treatment is effective and safe in the overall population of cardiac arrest patients. Neither are there any comparative trials establishing the usefulness of emergency percutaneous coronary repermeabilization or goal-directed therapy in successfully resuscitated cardiac arrest patients. These treatments are costly, and many ICUs in Europe and North America refrain from aggressive initial treatments. The development of a severity score for the post-resuscitation phase may help to conduct studies aimed at demonstrating that survival rates in successfully resuscitated cardiac arrest patients are often comparable with those in patients with other critical conditions.
We apologize for the error in Table 4. An erratum will be published to indicate that 127 (60%) should be read instead of 155 (74%) in all patients receiving therapeutic hypothermia.
In brief, we agree with Sunde et al. that no withdrawal decisions should be taken on the basis of any score (including the OHCA score) and that all patients should receive optimal initial post-resuscitation care until a reliable, individual neurological evaluation can be done in order to maximize the patient's chances of a good outcome. We must avoid confusion ‘between interpreting an estimated probability of mortality and predicting whether or not a given patient will live or die’.7
References
- Adrie C, Cariou A, Mourvillier B, Laurent I, Dabbane H, Hantala F, Rhaoui A, Thuong M, Monchi M. (2006) Predicting survival with good neurological recovery at hospital admission after successful resuscitation of out-of-hospital cardiac arrest: the OHCA score. Eur Heart J 27:2840–2845.
[Abstract/Free Full Text] - Le Gall JR, Lemeshow S, Saulnier F. (1993) A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 270:2957–2963.[Abstract]
- Knaus WA, Draper EA, Wagner DP, Zimmerman JE. (1985) APACHE II: a severity of disease classification system. Crit Care Med 13:818–829.[ISI][Medline]
- Lemeshow S, Teres D, Pastides H, Avrunin JS, Steingrub JS. (1985) A method for predicting survival and mortality of ICU patients using objectively derived weights. Crit Care Med 13:519–525.[ISI][Medline]
- Lemeshow S, Teres D, Avrunin JS, Gage RW. (1988) Refining intensive care unit outcome prediction by using changing probabilities of mortality. Crit Care Med 16:470–477.[ISI][Medline]
- Hosmer DW. (1989) Applied Logistic Regression(John Wiley and Sons, New York).
- Le Gall JR. (2005) The use of severity scores in the intensive care unit. Intensive Care Med 31:1618–1623.[CrossRef][ISI][Medline]
- Oddo M, Schaller MD, Feihl F, Ribordy V, Liaudet L. (2006) From evidence to clinical practice: effective implementation of therapeutic hypothermia to improve patient outcome after cardiac arrest. Crit Care Med 34:1865–1873.[CrossRef][ISI][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||