Early prediction of the adult respiratory distress syndrome by a simple scoring method

Early identification of patients who will develop the adult respiratory distress syndrome (ARDS) is important in prognosis, management, and candidate selection for studies that evaluate both pathophysiology and early therapy. The purpose of this study was to improve the accuracy of early prediction of ARDS by using a simple scoring method that combines the weighted risks from individual risk factors. By a discriminant function analysis of 136 patients with one or more clinical conditions predisposing to risk of ARDS, simple linear equations were derived in which several different risk factors, weighted by coefficients, served as the variables. We developed several equations from which individual "ARDS scores" could be calculated, and we found retrospectively that half the 46 patients who had developed ARDS scored higher than a particular "index score," while about 90% of the patients who never developed ARDS scored less than that index despite the presence of significant risk factors. We then prospectively tested these equations in 87 other patients and found similar results. In summary, a simple clinical scoring system which can be used soon after admission can identify, with reasonable accuracy, individual patients who are almost assured of developing ARDS.