Introduction

Kidney stones are a common condition, with an estimated prevalence of about 8% in the general population ("Croppi E - 2012" [1]). Kidney stone disease is often burdened by a high frequency of recurrence, thus entailing high costs associated with its treatment. Although dietary and lifestyle interventions are thought to be effective in most cases, patients with active and recurrent disease, especially with known associated metabolic abnormalities, may benefit from pharmacological interventions such as thiazides, citrate and allopurinol. However, direct comparisons between these agents are rare, thus reducing the confidence in choosing the best pharmacological treatment for kidney stones.

Network meta-analysis (NMA) is a statistical technique that allows obtaining estimates of effect from indirect comparisons of therapeutic approaches by common comparators ("Mills EJ - 2013" [2]).

We conducted a NMA of randomized controlled trials (RCTs) included in a recent meta-analysis of therapeutic approaches in kidney stone disease ("Fink HA - 2013" [3]), in order to compute indirect comparisons of the effect of thiazide drugs, citrate and allopurinol to reduce recurrence in patients with calcium stones.

Methods

We analyzed data from RCTs that evaluated the effect of thiazides, citrate or allopurinol included in a recently published meta-analysis, and obtained estimates of effect for each treatment strategy compared with each other by means of NMA. Estimates were reported as risk ratios (RRs) and 95% confidence intervals (CIs). The outcome of interest was recurrence of kidney stones (either clinical or radiological).

Results

Our analysis included 11 RCTs enrolling 903 patients. Of these, 5 evaluated thiazides (421 patients), 4 citrate (263 patients) and 2 allopurinol (219 patients), all of them against diet/placebo (Figure 1). When compared with diet/placebo, the estimates for thiazides, citrate and allopurinol overlapped with those reported by Fink and colleagues, showing a significant benefit for all those approaches. Indirect estimates showed that, compared with citrate, both thiazides (RR 2.11, CI 1.13, 3.92) and allopurinol (RR 2.41, CI 1.25, 4.63) were less effective in reducing recurrences. When compared with each other, the effect of thiazides and allopurinol was similar (Figure 2). Heterogeneity was not statistically significant.

Conclusion

Indirect comparisons obtained by NMA suggest that citrate might be more effective than thiazide and allopurinol in reducing recurrences in patients with idiopathic calcium stones. However, direct comparisons obtained from properly conducted RCTs are needed to confirm our findings.