OBJECTIVE: In mechanical ventilation, the assessment of pulmonary mechanics, mainly of total compliance (Crs), total resistance (Rrs), and intrinsic positive end-expiratory pressure (PEEPint), is clinically important. By using airway pressure (Paw) and flow (V'aw), the least squares fit (LSF) method allows the continuous calculation of these parameters. The objective of this work was to study the influence of imprecise breath detection, phase shift between airway pressure and flow signals, and noise on the determination of Crs, Rrs, and PEEPint. METHODS: Paw and V'aw were mathematically simulated as well as recorded in mechanically ventilated patients. Crs, Rrs, and PEEPint were computed off-line using the LSF method. The boundaries of the breath data window and the phase relationship between Paw and V'aw signals were manipulated and noise was superimposed. RESULTS: Both simulated and patient data gave similar results. Crs and Rrs were not sensitive to imprecise breath detection. If the first portion of the breath was missed, the mean relative error on PEEPint was 20% or 53% when the exact beginning of inspiration was missed by 0.1 or 0.3 sec, respectively. Paw lag of 66 ms with respect to V'aw yielded a relative error of -15 +/- 4% (mean +/- SD) for Rrs, -5 +/- 2% for Crs, and +13 +/- 16% for PEEPint. Paw lead of 66 ms with respect to V'aw yielded a relative error of +5 +/- 4% for Rrs, +7 +/- 3% for Crs, and +14 +/- 18% for PEEPint. Noise had very little impact on the accuracy of Crs, Rrs, and PEEPint. CONCLUSIONS: We conclude that the LSF method allows the assessment of Crs, Rrs, and PEEPint even with high levels of noise in patients with normal lungs provided that Paw and V'aw signals are precisely synchronised and a reliable breath detection algorithm is used.
OBJECTIVE: In mechanical ventilation, the assessment of pulmonary mechanics, mainly of total compliance (Crs), total resistance (Rrs), and intrinsic positive end-expiratory pressure (PEEPint), is clinically important. By using airway pressure (Paw) and flow (V'aw), the least squares fit (LSF) method allows the continuous calculation of these parameters. The objective of this work was to study the influence of imprecise breath detection, phase shift between airway pressure and flow signals, and noise on the determination of Crs, Rrs, and PEEPint. METHODS: Paw and V'aw were mathematically simulated as well as recorded in mechanically ventilated patients. Crs, Rrs, and PEEPint were computed off-line using the LSF method. The boundaries of the breath data window and the phase relationship between Paw and V'aw signals were manipulated and noise was superimposed. RESULTS: Both simulated and patient data gave similar results. Crs and Rrs were not sensitive to imprecise breath detection. If the first portion of the breath was missed, the mean relative error on PEEPint was 20% or 53% when the exact beginning of inspiration was missed by 0.1 or 0.3 sec, respectively. Paw lag of 66 ms with respect to V'aw yielded a relative error of -15 +/- 4% (mean +/- SD) for Rrs, -5 +/- 2% for Crs, and +13 +/- 16% for PEEPint. Paw lead of 66 ms with respect to V'aw yielded a relative error of +5 +/- 4% for Rrs, +7 +/- 3% for Crs, and +14 +/- 18% for PEEPint. Noise had very little impact on the accuracy of Crs, Rrs, and PEEPint. CONCLUSIONS: We conclude that the LSF method allows the assessment of Crs, Rrs, and PEEPint even with high levels of noise in patients with normal lungs provided that Paw and V'aw signals are precisely synchronised and a reliable breath detection algorithm is used.