Twenty-one RCTs were included in the review. Seven RCTs (n=1,411 participants) assessed the effects of laxatives, seven trials (n=2,639) assessed the effects of prucalopride, three trials (n=610) assessed the effects of lubiprostone and three trials (n=1,582) assessed the effects of linaclotide. Twelve of the 21 included studies were at low risk of bias.
Laxatives: Compared to placebo, laxatives were associated with a significantly lower rate of failure to respond to therapy (RR 0.52, 95% CI 0.46 to 0.60; seven studies) with an NNT of 3 (95% CI 2 to 4). Subgroup analyses according to type of laxative reported significant benefits for osmotic and stimulant laxatives when compared to placebo. There was no evidence of significant heterogeneity. Sensitivity analyses indicated that treatment effects remained similar. Laxatives significantly reduced individual chronic idiopathic constipation symptoms. Compared to placebo, mean number of stools per week was significantly higher with laxatives (WMD 2.55, 95% CI 1.53 to 3.57; six studies); subgroup analyses confirmed the benefits separately for both osmotic and stimulant laxatives. Significant heterogeneity (I2=100%) was identified in the analyses. Compared to placebo, laxatives were associated with significantly more frequent diarrhoea (RR 13.75, 95% CI 2.82 to 67.14; two studies) with an NNH of 3 (95% CI 2 to 6).
Prucalopride: Compared to placebo, prucalopride was associated with a significantly lower rate of failure to respond to therapy (RR 0.82, 95% CI 0.76 to 0.88; seven studies) with significant heterogeneity (I2=60%) between studies and an NNT of 6 (95% CI 5 to 9). When sensitivity analyses were undertaken, there was no longer evidence of significant heterogeneity. Prucalopride was associated with a significantly greater risk of adverse events such as headache, nausea and diarrhoea.
Lubiprostone: Compared to placebo, lubiprostone was associated with a significantly lower rate of failure to respond to therapy (RR 0.67, 95% CI 0.56 to 0.80; three studies) with an NNT of 4 (95% CI 3 to 7). Lubiprostone was associated with a significantly greater risk of adverse events such as nausea and diarrhoea. No significant heterogeneity was found in the analyses.
Linaclotide: Compared to placebo, linaclotide was associated with a significantly lower rate of failure to respond to therapy (RR 0.84, 95% CI 0.80 to 0.87; three studies) with an NNT of 6 (95% CI 5 to 8). A significantly higher proportion of patients on linaclotide reported a decrease in severity of abdominal discomfort and bloating. Linaclotide was associated with a significantly greater risk of diarrhoea (RR 3.08, 95% CI 1.27 to 7.48; three studies). No significant heterogeneity was found in the analyses.
Where funnel plots were able to be generated, there was generally no evidence of significant asymmetry that might indicate publication bias.