[Prediction of outcome of elderly patients with respiratory failure: comparison of current scoring systems]

To evaluate the predictive power of current scoring systems applied in elderly patients with respiratory failure.

Data of 159 elderly patients with respiratory failure, also meeting the criteria of multiple organ dysfunction syndrome in the elderly (MODSE), being admitted to PLA General Hospital from January 2009 to March 2010 were analyzed with four scoring systems, i.e. acute physiology and chronic health evaluation (APACHE II, APACHE III), simplified acute physiology scoreII (SAPSII), as well as multiple organ dysfunction score (Marshall score), for the prediction of their outcome on the first day after admission, and the predictive power of respective system was compared.

One hundred and fifty nine patients were involved, with a mean age of (75.26±7.87) years, and a 28-day mortality of 53.5%. Incidence of organ failure and prognosis scores in nonsurvivors were significantly higher than those in survivors (APACHE II score: 27.88±8.19 vs. 17.73±6.21; APACHE III score: 101.10±29.81 vs. 64.24±21.98; SAPSII score: 59.70±17.68 vs. 38.05±12.10; Marshall score: 8.59±3.15 vs. 4.90±2.66; organ failure: 4.60±1.14 vs. 2.88±0.88, all P<0.01). Among the four scoring systems, SAPSII presented excellence in sensitivity (91.76%), while APACHE III in specificity (91.89%). Area under curve of receiver operator characteristic of all scoring systems reached 0.820 (0.829 for APACHE II , 0.836 for APACHE III , 0.844 for SAPSII, 0.820 for Marshall score).

All the four current scoring systems , i.e. APACHE II, APACHE III, SAPSII and Marshall score, presented satisfactory predictive power in prognosis of elderly patients with respiratory failure.