EURURO Vol. 71 No. 4

506 692

that meta-analyses based on randomized trials represent
one of the highest levels of evidence , but meta-analyses
depend on the quality of the included RCTs. The majority of
published RCTs used for meta-analyses are very heteroge-
neous, include small numbers of patients (ie, are under-
powered), and have known publication bias. The continued
use of poor-quality data in each ‘‘new’’ meta-analysis calls
the validity and reliability into question.
Nevertheless, the available evidence resulted in a grade
A recommendation for MET using a-blockers In the
2016 issue of the EAU guidelines, the recommendations
have been downgraded to grade C, recognizing the recent
publications of multicenter, randomized, double-blinded
and placebo-controlled studies. The EAU guidelines
continue to recommend offering a-blockers as MET as
one of the treatment options after informing the patient
about the lack of efficacy in recent studies
4. Discussion
The contradictory results of meta-analyses, on the one side,
and large high-quality trials, on the other side, show the
dilemma of defining and valuing best evidence. Well-
performed, large, multicenter, placebo-controlled RCTs are
assigned the highest level of evidence. In contrast, meta-
analyses of small RCTs help generate hypotheses for more
reliable RCTs rather than providing the best possible
evidence
The primary outcomes of the paper by Furyk et al were
stone expulsion on CT at 28 d and time to stone expulsion
Pickard et al defined the primary end point of their
study as the difference in the need for further urologic
interventions . They avoided proving stone passage
with CT imaging because it was not standard clinical
practice in the United Kingdom and would have involved
additional costs and radiation doses for study participants.
One can argue that this appears to be a less precise
surrogate marker to evaluate the true efficacy of MET for
stone expulsion; however, in both studies, the event rates
were almost identical, showing that stone passage occurred
in almost 80%.
The study by Pickard et al was not powered to assess
the efficacy of MET in stones >5 mm in the upper or middle
ureter. In addition, there were no significant differences in
pain scales or number of rescue pain medications. These
were secondary outcomes assessed with patient surveys,
which suffered from decreased follow-up rates compared
with the primary outcome (62% vs 97%).
The question is whether to base treatment decisions on
meta-analyses composed of single-center, small, mainly
low-quality trials favoring MET or on a few large high-
quality trials with findings of no significant effect.
Further work is required to investigate the phenomenon
of large, high-quality trials showing smaller effect size than
meta-analyses of several small, single-center, lower quality
studies. Results of meta-analyses should be subjected to
careful sensitivity analyses to test the robustness of the
findings and interpreted on principle with caution, even if
the pooled effect is statistically significant. The question
also arises as to whether all available evidence must be
included in meta-analyses.
During both the AUA and EAU annual meetings, a panel
discussion on the topic was held. At the EAU meeting, the
panel agreed that the largest high-quality study, by Pickard
et al , was not powered for stones >5 mm, and the EAU
debate resulted in the recommendation for further high-
quality multicenter and double-blinded RCTs with a well-
controlled end point of stone passage rate. In contrast, at the
AUA annual meeting, the conclusion was to continue
offering a-blockers in particular for informed patients with
distal ureteral stones of 5–10 mm, obviously in view of the
AUA guidelines work.
5. Conclusions
The contradictory results of meta-analyses of small RCTs
compared with large multicenter trials show the vulnera-
bility of meta-analyses, which always have to be taken into
account. This vulnerability emphasizes the responsibility of
careful planning for RCTs. To minimize publication bias,
trial registration should be a prerequisite of ethics
committees for the approval of studies.
The current evidence on whether or not to use MET
indicates that a potential benefit of MET (a-blockers) is
most likely for distal ureteral stones >5 mm. For smaller
ureteral stones (i.e., the majority of ureteral stones), there is
no proven benefit. It is important to inform patients about
the possible but as yet unproven benefit using a-blockers as
well as their off-label use and possible side effects.
Further well-designed, double-blinded, placebo-con-
trolled, multicentric RCTs with clearly agreed-upon end
points are suggested to address the clinical question.
Conflicts of interest: The authors have nothing to disclose.
References
[1]
[2]
[3]
.
[4]
[5]
E U R O P E A N U R O L O G Y 7 1 ( 2 0 1 7 ) 5 0 4 – 5 0 7 506