3.4. Fasting state
In a subgroup of 213 intubations where an aspirate could be obtained, the effect of fasting on the pH cut-off point was calculated. In fasting patients, 78.7% (n = 63) of pH measurements was ≤ 5.5 as opposed to 76.6% (n = 102) in patients without fasting. The possibility to obtain an aspirate was 18.2% higher (95.3% versus 77.1%) in patients who were not fasting (P < 0.01).
3.5. pH measurement versus radiological control
It was possible to obtain an aspirate from the nasogastric tube in 270 of 331 intubations. However, as previously indicated in the research method, a number of patients (n = 29) had pulled out their NG tube or were not deemed fit enough to be transferred to radiology. That left a total of 241 intubations in which the aspirate could be compared to the X-ray results. One hundred eighty of these pH measurements showed a pH ≤ 5.5. With this cut-off point, the probability that the NG tube had been correctly placed in the stomach was 98.9% (positive predictive value). Two measurements showed a false positive test (pH ≤ 5.5) with placement in the distal oesophagus (two different intubations on the same patient on antacids). The low pH measured in this patient was the result of the existence of a large hiatal hernia (see Table 2). With a pH ≥ 6, the probability that the tip of the NG tube was located outside the stomach was 19.7% (negative predictive value). The results of pH measurements where an aspirate could be obtained, showed a sensitivity (the probability that pH results from aspirates correctly identified that the NG tube had been placed in the stomach) of 78.4% with a specificity (the probability that pH results from aspirates correctly identified that the NG tube was not in the stomach) of 84.6%. When you take into account all attempts of aspiration (including those without obtaining an aspirate) the sensitivity was 66% and the specificity 93%