The Lack of Effect of Routine Magnesium Administration on Respiratory Function in Mechanically Ventilated Patients: Discussion
Molloy and associates had evaluated the effects of hypomagnesemia on respiratory muscle strength in 11 patients admitted to the hospital for treatment of alcohol withdrawal and in 6 patients with chronic obstructive pulmonary disease. They found that infusion of 6 g of magnesium increased PEmax by 24 percent and Pimax by 46 percent. Fiaccadori and associates studied the correlation between muscle magnesium levels and serum magnesium levels in patients with chronic obstructive pulmonary disease admitted to their ICU and found that serum magnesium was a poor indicator of tissue magnesium levels. They concluded that 47 percent of their patients with normal serum magnesium levels had evidence of tissue magnesium deficiency and therefore measuring serum magnesium was not useful to detect tissue magnesium deficiency. Since the routine measurement of serum magnesium is not a sensitive method of detecting tissue magnesium deficiency, we decided to test whether the routine infusion of magnesium to ventilated patients with respiratory failure might improve respiratory function. The time course of action of magnesium in the ICU setting has not been fully described. Aikawa and associates have previously studied the kinetics of magnesium distribution in normal subjects and in patients with diabetes or hepatic disease. They found in both normal subjects and patients that infused magnesium exchanged slowly between plasma and muscle requiring greater than 18 h to begin to equilibrate. We therefore decided that the most conservative approach to analyze our data would be to compare only the values obtained immediately following magnesium infusion with those preceding the infusion in order to assess whether the magnesium infusion had any effects. canadianneighborpharmacy.com
We expected that there would be two distinct populations within our study sample: those with magnesium deficiency, and those without magnesium deficiency. That the changes in Pimax and VC followed a normal distribution implies that there is not a subgroup within the larger group, ie, there is not a bimodal pattern in the changes consistent with a “responder” group. The changes in PEmax (Fig 3) are not normally distributed, but there is no visible bimodal pattern or any discernible “responder” subgroup. Based on previous data, we expected that approximately 50 percent of our subjects would demonstrate large (30 percent or greater) increases in Pimax and PEmax; the other 50 percent should demonstrate little change over 24 h. Instead, we found a more homogeneous population, implying that either there were few subjects with tissue magnesium deficiency or that even after correcting magnesium deficiency, there were other causes for continued respiratory muscle weakness. We used a crossover design when we organized this study, knowing that there might be a significant carryover effect due to the potentially slow onset of action of magnesium. We expected that there would be large changes that would clearly define the responder group, but instead we found, on the whole, a homogeneous population.