Approval for this study was obtained from the Ethics Committee of the First People's Hospital of Xuzhou. A total of 67 patients with AECOPD, comprising of 41 men and 26 women, aged 52–81 years, hospitalized between 2011 and 2013, were included. The patients were mechanically ventilated and had either an endotracheal or tracheostomy tube. Of the 67 cases, 38 patients had serum phosphate levels <0.87 mmol/l and were assigned to the hypophosphatemic group. The remaining 29 cases with a serum phosphate concentration ≥0.87 mmol/l constituted the normophosphatemic group. Differences in the forced expiratory volume in 1 sec (FEV1) and in the vital capacity between the hypophosphatemic and normophosphatemic groups were <70%. Specifically, FEV1 accounted for 47.0%, in the normophosphatemic group, and 48.1%, in the hypophosphatemic group, of the predictive value (F=1.414 under statistical processing, or less than tabular 3.12), with no significant difference between the two groups (P>0.05). In addition, the Acute Physiology and Chronic Health Evaluation II scores in the two groups were comparable (23±5.2 and 21±4.9, respectively; P>0.05), confirming similar severity of illness between the two groups.

Blood samples collected from all 67 cases were analyzed using a Hitachi 7600 automatic biochemical analyzer (Holliston, MA, USA). The obtained serum phosphate results were used in patient allocation, as stated above, for the normophosphatemic group (≥0.87 mmol/l) and the hypophosphatemic group (<0.87 mmol/l). Prior to weaning from ventilator support, synchronized intermittent mandatory ventilation was implemented, in which the ventilator breaths were synchronized with patient inspiratory effort. Patient respiratory parameters, including the tidal volume of autonomous respiration, the minute ventilation volume and the static lung compliance, were monitored and recorded.

If a patient, 24 h after weaning at PCO₂ <60 mmHg and PO₂ >60 mmHg, continued to maintain a respiratory rate of <30 times/min, normal cardiovascular function, along with no paradoxical respiration and no irritation, the weaning was considered successful; otherwise, as failure.

Statistical analysis was performed with SPSS 10.0 software (SPSS, Inc., Chicago, IL, USA). Comparison of categorical data (% failure-to-wean) between the hypophosphatemic and normophosphatemic groups was performed using the Pearson's Chi-square test. Measurable data were subjected to the F-test. P<0.05 was considered to indicate a statistically significant difference.

Duration of MV applied to the 67 cases of AECOPD under examination varied from 21 h to 3.2 months, with an average of 75 h. By the end of weaning from ventilator support, 7 cases succumbed.

Of the 67 AECOPD patients, 29 were cases of normal serum phosphate and 38 were cases of hypophosphatemia. Thus, the morbidity rate of hypophosphatemia was 56.72% in AECOPD patients treated with MV. In the normophosphatemic group, 26 cases were successfully weaned and 3 cases failed (10.34%). By contrast, in the hypophosphatemic group, 25 cases were successfully weaned and 13 cases failed (34.21%). The percentage of weaning failure was significantly higher in the hypophosphatemic vs. the normophosphatemic groups (P<0.05; Table I).

When synchronized intermittent mandatory ventilation was used, the tidal volume of spontaneous respiration was smaller in the normophosphatemic group than that in the hypophosphatemic group (368 and 328 ml, respectively, P<0.01). By contrast, the minute ventilation volume of the normophosphatemic group was similar to the hypophosphatemic group (4,850 and 4,570 ml, respectively, P>0.05). A significant difference was identified in static lung compliance between the two groups (39 l/cm H₂O and 26 l/cm H₂O, respectively, P<0.01). Results shown in Table II suggested that hypophosphatemia was associated with respiratory muscular weakness (a decrease in the volume of spontaneous respiration), reduced static lung compliance and impaired pulmonary function, leading to weaning failure in AECOPD patients.

Most AECOPD patients consumed less food when being treated with MV to assist breathing. Even for those who consumed an adequate amount of food, the phosphorus content of the foods may have been insufficient. Moreover, clinical therapy often focused on replenishing energy with vitamins K⁺, Na⁺, Cl⁻, as well as Ca⁺⁺ and Mg⁺⁺, while ignoring the supply of phosphorus.

Patients with AECOPD often received intravenous injection of large amounts of glucose and insulin over a long period of time. They may also experience severe infection. These conditions may lead to higher cellular phosphorus absorption, negligible intracellular phosphorus-to-serum transfer, and acid-base imbalance. Under MV, respiratory alkalosis may produce a further decrease in serum phosphorus.

The use of diuretics, theophyllines and hormonic drugs may increase phosphorus excretion. Hypophosphatemia may lead to decreases in 2,3-diphosphoglyceric compound and ATP in erythrocytes (showing a shift to the left in an oxygen dissociation curve). It may critically reduced oxygen transfer from erythrocytes to tissues, resulting in tissue hypoxia, and the eventual failure of weaning. The results of our study show that hypophosphatemia causes respiratory muscular weakness (a decrease in the volume of autonomous respiration), reduction of static lung compliance and a loss of respiratory function, leading to weaning failure.

MV was a highly effective treatment in AECOPD patients suffering from severe respiratory failure. However, continuous application of a ventilator may result in uncontrolled infection, airway injury, and increased cost of treatment (2–5). Concurrently, there was a significant association between severe hypophosphatemia and the mortality of patients under ventilator support (11,12). Additionally, hypophosphatemia affected the recovery of respiratory function, causing ventilator dependence, and, consequently, weaning failure. Clinics typically ignore this type of problem. Nevertheless, ventilator assistance providing support to underlying diseases is an independent factor in the deaths of patients infected with pseudomonas aeruginosa and having hematological tumor, low-phosphorus, low-protein and hyperlactatemia (20). Since the low-phosphorus factor prevails continuously in AECOPD patients using MV, the patients should be tested frequently and have phosphorus replenished according to test results, while constantly bearing in mind that excessive phosphorus intake is also harmful (9,10).

In summary, our findings suggest that hypophosphatemia significantly affected the success of weaning from MV in patients with AECOPD. Serum phosphorus levels therefore serve as a reference index to determine the success of weaning.