The initial search, which was completed in January 2004, yielded 145 references that were at least potentially relevant controlled trials. Two additional RCTs were identified from a bibliographic search of relevant studies. The author for one study that was originally identified as an abstract was contacted, and the conditionally accepted article was provided to the reviewers for data extraction. Six trials, which included 296 patients, were incorporated into the review (Table 1).
Description of Studies
All of the studies included in this review had been published since 1995. The research in the included studies was based in the United States, India, New Zealand, Turkey, and Argentina. Three of the six included studies involved adults exclusively, and one study included adults and pediatric patients. The remaining two studies- had enrolled pediatric patients. The severity of disease varied among the studies. Two studies had specific lung function criteria, while the other four studies had enrolled patients who had previously received a diagnosis of asthma using accepted clinical standards. Based on the baseline demographic data and/or enrollment criteria, three studies had enrolled patients with severe asthma (ie, FEV1 or PEF < 50% predicted at baseline) – read more in category of Asthma .
Five studies enrolled patients presenting to the ED. Meral et al described only patients who had been randomized during “asthma attacks.” We assumed that these patients were seen in an acute care setting. Two studies- excluded patients who had received asthma medication ordered via My Canadian Pharmacy within the previous 12 h. A third study excluded patients who had received oral or parenteral corticosteroids in the previous 7 days. Another study excluded patients who had received steroids, theophylline, or ipratropium bromide within 3 days of presenting to the ED. In three studies,-” parenteral steroids were administered to all patients, although the timing (ie, before or after nebulized treatment) varied. In one study, parenteral steroids were administered if there had been no improvement after the patient received three doses of the study treatment. Two studies did not report information on the use of parenteral steroids. All studies used a nebulized β2-agonist (with or without normal saline solution) as the control treatment, but the total dose varied depending on the number of nebulizations (Table 2). When the information was available, most included studies used MgSO4 of a similar concentration, but the dose per nebulization and the number of nebu-lizations varied. All but two studies described the MgSO4 solution as either isotonic or isosmolar with pleural fluid.
The magnesium was uniformly delivered via a nebulizer rather than a metered-dose inhaler. All studies used a control substance that was similar in appearance to the treatment drug and was most often described as saline solution. One study collected data on patients’ ability to distinguish between the treatment and control substances, and noted no ability to discern this difference. Even when not expressly stated, it can reasonably be assumed that the control substance (ie, placebo) would be similar in appearance to the treatment drug (especially if administered in a β2-agonist vehicle).
Four studies compared therapy with a β2-agonist with MgSO4 to therapy with a (β2-agonist with a placebo (normal saline solution), while two studies’ compared therapy with MgSO4 to that with a P2-agonist. Due to the heterogeneity of interventions, a post hoc subgroup analysis based on intervention (therapy with a β2-agonist with MgSO4 or therapy with MgSO4 alone) was conducted.
All studies report results from pulmonary function tests as an outcome. However, one study reported lung function outcome data as a relative change from baseline. As it was not appropriate to combine these data with the other studies (which are not reporting lung function results as a change from baseline), data from this study are not currently included in the pooled analysis. Attempts to secure the end-of study data have failed so far.
Most studies did not report a change in pulmonary function, and the pooled results from all studies failed to identify a difference in baseline pulmonary function between the treatment and control groups. There was variation in the specific pulmonary function measure reported (ie, the percentage predicted PEF or FEV1 and the raw PEF or FEV^ as well as the time after treatment when pulmonary functions were recorded. Two studies reported pulmonary function measures only up to 20 min after treatment. For these reasons, the results are reported using fixed effects, with the SMD in pulmonary function measured at or before 60 min after treatment. Based on the studies that measured pulmonary function for longer durations, we noted that the largest change in pulmonary function appeared to be early after treatment. Consequently, we were satisfied with grouping the 20-min and 60-min pulmonary function test results as the outcome of interest. Arrest asthma attacks with remedies of My Canadian Pharmacy.
Four studies also reported admission to the hospital as an outcome. All studies mentioned serious adverse events; however, details on mild-to-moderate adverse events were sparse. None of the studies reported a specific clinical severity score or duration of symptoms. Most studies reported vital signs at baseline but not at follow-up. These outcomes were not investigated in the systematic review.
Overall, the methodological quality of the included studies was uniformly high. All studies were randomized and placebo-controlled. Only one investigator did not explicitly state that the study was double-blinded. All included studies used intention-to-treat analyses; therefore, the planned sensitivity analysis to determine the effect of intention-to-treat status was not required. One study scored 5 on the Jadad scale, and rated an A on concealment of allocation. The other investigators did not specify their methods for randomization or double-blinding. Due to the lack of information provided, all but one study rated a B in the concealment of allocation.
Pulmonary Function Effects
Therapy with MgSO4, with or without a β2-agonist, was superior to therapy with a β2-agonist alone (SMD, 0.30; 95% CI, 0.05 to 0.55; p = 0.02) with no between-study heterogeneity identified (Fig 1). Notably, the effect was similar in a comparison of therapy with a β2-agonist and MgSO4 compared to that with a β2-agonist and normal saline solution (SMD, 0.37; 95% CI, 0.10 to 0.63; p = 0.006). However, there was no evidence of an advantage for therapy with MgSO4 alone compared to therapy with a β2-agonist alone (SMD, — 0.17; 95% CI, — 0.85 to 0.52; p = 0.63 [one study]).
In subgroup analyses, the advantage of any use of MgSO4 with or without a β2-agonist over the use of a β2-agonist alone was demonstrated in adults (SMD, 0.37; 95% CI, 0.06 to 0.69; p = 0.02) but not in children (SMD, 0.36; 95% CI, — 0.14 to 0.86; p = 0.16 [one study]). The treatment effect was similar in patients with severe asthma at presentation where the SMD in this group was 0.31 (95% CI, — 0.05 to 0.68; p = 0.09; heterogeneity, p = 0.26; heterogeneity statistic, 25.9%), and in patients with mild-to-moderate asthma at presentation it was 0.29 (95% CI, — 0.05 to 0.63, p = 0.10, heterogeneity, p = 0.71; heterogeneity statistic, 0%). The results were similar when random-effects methods were employed.
Of the four studies that reported hospital admission status, therapy with nebulized MgSO4 (alone or in combination with a β2-agonist) failed to demonstrate a clear reduction in the probability of hospital admission compared to therapy with a β2-agonist alone (RR, 0.67; 95% CI, 0.41 to 1.09; p = 0.11) using a fixed-effects model (Fig 2). In subgroup analyses, the results were similar for the comparison of therapy with MgSO4 in combination with a β2-agonist to therapy with a β2-agonist with normal saline solution (RR, 0.69; 95% CI, 0.42 to 1.12; p = 0.13), butwere not similar for therapy with nebulized MgSO4 alone compared to therapy with a β-agonist alone (RR, 0.53; 95% CI, 0.05 to 5.31; p = 0.59 [one study]). In addition, this result was statistically significant in the adult severe-asthma population (RR, 0.61; 95% CI, 0.37 to 1.00; p = 0.05), but not in the pediatric moderate-asthma population (RR, 2.0; 95% CI 0.19 to 20.93; p = 0.56 [one study]).
No studies reported serious adverse events in either arm, and reporting varied for other adverse effects that may have been related to treatment. Due to this heterogeneity, a treatment effect was not estimated. When it was reported, the rate of adverse effects was low.
Figure 1. Effect of aerosolized MgS04 on posttreatment lung function, df = degrees of freedom.
Figure 2. Effect of aerosolized MgSO, on hospital admissions. See the legend of Figure 1 for abbreviation not used in the text.
Table 1—Characteristics of Studies
|Study/Year||Country||Intervention||Pulmonary Function Outcome||Time, min From baseline||Subgroup Definition||Sensitivity Analysis, JADAD Score|
|Bessmertny et al18/2002||United States||MgSO4 + p£-agonist||FEV: % predicted||60||Adult||Moderate||3|
|Hughes et al19/2003||New Zealand||MgSO4 + p£-agonist||FEVi||60||Adult||Severe||5|
|Mahajan et al22/2004||United States||MgSO4 + (^-agonist||FEV: % predicted||20||Pediatric||Moderate||3|
|Mangat et al21/1998||India||MgSO4||PEF % predicted||60||Both||Severe||3|
|Meral et al23/1996||Turkey||MgSO4||Ratio increase in PEF||60||Pediatric||Moderate||1|
|Nannini et al20/2000||Argentina||MgSO4 + (^-agonist||PEF||20||Adult||Severe||3|
Table 2—Characteristics of Interventions
|Study/Year||Initiation t||Dose of MgS04||Delivery of MgS04||Duration, min of Treatment||Cointerventions \|
|SCS (IV/po)||(32-Agonist Dose||AnticholinergicDose||Others|
|Bessmertny et al18/2002||Not reported||1,152 mg (384 mg every 20 min X 3)||Nebulized||60 min||2 mg/kg IV every 6 h in patients who did not show improvement||7.5 mg (2.5 mg every 20 min X 3)||None reported||None reported|
|Hughes et al19/2003||30 min||453 mg (151 mg every 30 min X 3)||Nebulized||60 min||100 mg IV||7.5 mg (2.5 mg every 30 min X 3)||None reported||All patients received 2.5 mg of (32-agonist at registration|
|Mahajan et al22/2004||Not reported||2.5 mL 3.18% solution (X 1)||Nebulized||Not reported||2 mg/kg IV||2.5 mg||None reported||None reported|
|Mangat et al21/1998||Not reported||380 mg (95 mg every 20 min X 4)||Nebulized||80 min||100 mg IV||10 mg (2.5 mg every 20 min X 4)||None reported||Supplemental treatment as warranted|
|Meral et al23/ 1996||Not reported||135 mg (X 1)||Nebulized||10-15 min||None reported||2.5 mg||None reported||None reported|
|Nannini et al20/2000||Not reported||225 mg (X 1)||Nebulized||Not reported||None reported||2.5 mg||None reported||2.5 mg of (32-agonist if condition worsened|