Articles Medical therapy to facilitate urinary stone passage: a meta-analysis John M Hollingsworth, Mary A M Rogers, Samuel R Kaufman, Timothy J Bradford, Sanjay Saint, John T Wei, Brent K HollenbeckSummary Background Medical therapies to ease urinary-stone passage have been reported, but are not generally used. If eﬀ ective, Lancet 2006; 368: 1171–79 such therapies would increase the options for treatment of urinary stones. To assess eﬃ cacy, we sought to identify See Comment page 1138 and summarise all randomised controlled trials in which calcium-channel blockers or α blockers were used to treat Department of Urology urinary stone disease.
(J M Hollingsworth MD, T J Bradford MD, J T Wei MD, B K Hollenbeck MD);Division of Methods We searched MEDLINE, Pre-MEDLINE, CINAHL, and EMBASE, as well as scientiﬁ c meeting abstracts, up General Medicine, Department to July, 2005. All randomised controlled trials in which calcium-channel blockers or α blockers were used to treat of Internal Medicine ureteral stones were eligible for inclusion in our analysis. Data from nine trials (number of patients=693) were pooled. (M A M Rogers PhD, The main outcome was the proportion of patients who passed stones. We calculated the summary estimate of eﬀ ect S R Kaufman MA, S Saint MD); Veterans Aﬀ airs/University of associated with medical therapy use using random-eﬀ ects and ﬁ xed-eﬀ ects models. Michigan Patient Safety Enhancement Program Findings Patients given calcium-channel blockers or α blockers had a 65% (absolute risk reduction=0·31 95% CI (M A M Rogers, S R Kaufman, 0·25–0·38) greater likelihood of stone passage than those not given such treatment (pooled risk ratio 1·65; 95% CI S Saint);Center for Practice Management and Outcomes 1·45–1·88). The pooled risk ratio for α blockers was 1·54 (1·29–1·85) and for calcium-channel blockers with steroids Research, Ann Arbor Veterans was 1·90 (1·51–2·40). The proportion of heterogeneity not explained by chance alone was 28%. The number needed Aﬀ airs Health Services Research to treat was 4. & Development Center of Excellence, Ann Arbor (S Saint), MI, USA Interpretation Although a high-quality randomised trial is necessary to conﬁ rm its eﬃ cacy, our ﬁ ndings suggest that Correspondence to: medical therapy is an option for facilitation of urinary-stone passage for patients amenable to conservative Brent K Hollenbeck management, potentially obviating the need for surgery.
1500 E Medical Center Dr TC 3875-0330, Ann Arbor,
these drug classes stems from our understanding of MI 48109-0330, USA
The lifetime risk of urinary stone disease (urolithiasis) is
ureteral smooth-muscle physiology and urinary
estimated to be between 5% and 12% in Europe and the obstruction.11–17 Despite growing evidence from clinical USA,1–4 aﬄ
icting 13% of men and 7% of women.5 Since trials in support of its eﬃ
50% of patients will have a recurrence of renal colic used. Two explanations for underuse are: ﬁ rst, that within 5 years of their ﬁ rst episode,6 urolithiasis is a minimally invasive surgical techniques, such as chronic disease with substantial economic consequences
shock-wave lithotripsy and ureteroscopy have evolved to
and great public health importance. In the USA alone, allow for resolution of stone burden,18,19 but carry nearly 2 million outpatient visits were needed for the measurable risks and are costly;18,20–27 and second, that disease in 2000, with expenditures for inpatient and reports of empirical data for medical therapies have outpatient claims totalling US$2·1 billion.7
appeared only in urological publications, and therefore,
Although patients with urolithiasis might be the availability of such therapies might not be well known
asymptomatic, many have pain and thus commonly to physicians from other disciplines. Since many present to emergency or outpatient departments. specialists—such as emergency-department physicians, Provided that these patients do not need renal pelvic internists, and family practitioners—serve as the initial decompression—ie, they do not have a solitary kidney or
conduit into the health-care system for patients with
obstructing pyelonephritis—and that pain relief can be urolithiasis, a knowledge gap might exist. Therefore, we obtained, a trial of conservative non-surgical therapy is obtained data from clinical trials to derive a quantitative warranted, since many of these stones pass spontaneously.
estimate of ureteral-stone expulsion associated with
Indeed, studies have shown spontaneous passage rates medical therapy. of 71%–98% for small (≤5 mm) distal ureteral stones,8–10 with urinary-stone size and location being the two most Methods important predictors of stone passage.9 In view of this Eligibility criteria relation, investigators have sought ways of assisting the We used guidelines from the Quality of Reporting of process with the use of drugs, thereby reducing the need
Meta-Analyses conference.28 Inclusion criteria were
established before the search. Randomised controlled
Use of calcium-channel blockers and adrenergic trials of urolithiasis in any language were eligible. Only
α-antagonists for expulsive medical therapy has been those studies in which a calcium-channel blocker or an proposed as a way to enhance stone passage. Interest in adrenergic α-antagonist was used as the main therapy
www.thelancet.comVol 368 September 30, 2006 Articles
antispasmolytics, anticholinergic therapy, or cortico-
19 additional articles identiﬁed through
steroids. In the instance of many publications from a
single institution or group, we contacted the primary
investigator to determine whether these reports were generated from the same study population. Results from
duplicate populations were excluded. We speciﬁ ed a
minimum follow-up period of one week to allow for
and from parallel search by reference librarian
English and non-English language publications were
searched for human studies relating to expulsive medical
therapy from Jan 1, 1981, to July 31, 2005. We did an
electronic search of MEDLINE, Pre-MEDLINE, CINAHL,
and EMBASE. We searched by exploding and combining
the following medical subject heading terms:
“calcium channel blocking agent” with “urolithiasis” and “alpha adrenergic receptor blocking agent” with
9 trials assessed ureteral-stone passage related to expulsive medical therapy
“urolithiasis”. A research librarian at the University of Michigan Medical School did an independent search to
Figure 1: Study selection process
conﬁ rm the exhaustiveness of our search.
In an eﬀ ort to identify further reports, we hand searched
for ureteral-stone disease were included; therefore, we abstracts from the annual meetings of the World Congress excluded trials in which medical therapy was examined of Endourology, the European Association of Urology, as an adjuvant to surgery. For the purpose of ascertaining
and the American Urological Association between 1999
trial eligibility, control groups were deﬁ ned a priori as and 2005. We corresponded with the ﬁ rst or senior those not having received any additional medical therapy
authors of the published trials and conference abstracts
to ease urinary-stone passage—eg, other vasodilators, that met our inclusion criteria to clarify questions about
Location % female* Mean stone size randomised completed (years, SD)* (mm, SD)*
For those who completed For those who completed
Data are number (SD) unless otherwise indicated. All studies took place in outpatient settings. n/a=data not available. *Values are for the starting population unless indicated otherwise.
Table 1: Study and patient summary characteristics
www.thelancet.comVol 368 September 30, 2006 Articles
their studies, and to ﬁ nd out about unpublished or Collaboration33—ie, method of randomisation, conceal-continuing studies. We also contacted major drug ment of allocation, blinding, loss to follow-up, and manufacturers to inquire about unpublished industry-
intention-to-treat analysis. We then did sensitivity analyses
based on the criteria for which there was variation from the Cochrane criteria. Data reviews and statistical analysis
Subgroup analyses by drug type were done on the basis
Two reviewers independently extracted data from every of a priori decisions and included an assessment of study using a standardised form. To reduce bias, one of tamsulosin alone, nifedipine alone, all adrenergic the reviewers was blinded to the source of the publication
α antagonists, and the combined use of nifedipine and
and to the authors’ names. Inconsistencies between corticosteroids. Studies that used chemically diﬀ erent reviewers’ data were resolved through discussion until a drugs with similar mechanisms of action (such as consensus was reached. We were able to contact tamsulosin and terazosin) or diﬀ erent formulations of investigators from ﬁ ve of the studies for which we needed
the same drug (such as nifedipine immediate-release
clariﬁ cation. Three of the four drug companies with versus sustained-release) were grouped together for which we corresponded replied to our inquiries.
We wanted our ﬁ ndings to be generalisable to all adults
Stata version 9.0 (Stata Corp, College Station, Texas)
with ureteral stones amenable to a trial of conservative was used for all calculations. management. The primary endpoint in the studies was the proportion of patients who passed stones (cumulative
Role of the funding source
incidence). Therefore, we summarised eﬀ ect size using a
The sponsors had no role in study design, data collection,
pooled risk ratio with 95% CIs, comparing the proportion
data analysis, data interpretation, or writing of the report.
of patients taking medical therapy who passed stones The corresponding author had full access to all the data in with the proportion of those not taking medical therapy the study and had ﬁ nal responsibility for the decision to (controls) who passed stones. We used both submit for publication. Mantel-Haenszel ﬁ xed-eﬀ ects and DerSimonian and Laird random-eﬀ ects models to produce across-study risk
ratios. Since both models yielded similar results, the 415 studies were identiﬁ ed in the electronic database Mantel-Haenszel ﬁ xed-eﬀ ects model only is reported search (ﬁ gure 1). The review of meeting abstracts yielded here. Except where indicated otherwise, we did our 19 additional studies. We excluded from detailed review analysis on an intention-to-treat basis with the assumption
any articles that were either non-research reports, such
that dropouts failed to pass their stones. To assess heterogeneity, we used Cochran’s Q-test of heterogeneity
(which follows a χ² distribution) and the I² statistic (which
measures the proportion of inconsistency in individual
studies that cannot be explained by chance).29 On the basis of the pooled risk ratio and the baseline risk, we calculated
the number needed to treat (NNT). The 95% CIs for numbers needed to treat were calculated with the Newcombe-Wilson hybrid score method.30
To assess the eﬀ ect of individual studies on the summary
estimate of eﬀ ect, we did an inﬂ uence analysis, in which the pooled estimates were recalculated omitting one study at a time. We assessed publication bias using Rosenthal’s
fail-safe number—ie, the number of non-signiﬁ cant,
unpublished studies that would be needed to reduce a
statistically signiﬁ cant observed result to non-signiﬁ cance
at α=0·05).31 We also calculated the more recently described Rosenberg fail-safe number,32 which is weighted
by study variance and, therefore, might be more appropriate to meta-analyses that combine weighted eﬀ ect sizes. Publication bias is generally regarded as a concern if the fail-safe number is less than 5n +10, when
n is the number of studies included in the meta-analysis.
Since there is no gold standard for assessment of the
validity of randomised controlled trials, we add-ressed quality on the basis of every study’s conformity
with the criteria suggested by the Cochrane Figure 2: Percentage of patients who passed stones stratiﬁ ed by study group and mean stone size
www.thelancet.comVol 368 September 30, 2006 Articles
both of these studies also received treatment with the
Treatment group Number needed to
antispasmolytic phloroglunicol or its trimethoxy-benzene
derivative. Similarly, Staerman and colleagues38 randomly
Control group event rate
assigned patients nifedipine versus phloroglucinol.
Červenàkov and colleagues randomly assigned patients
standard therapy (including the antispasmolytic, diazepam)
versus standard therapy and tamsulosin.35 Because of the
smooth-muscle relaxation activity of antispasmolytics, the
control groups of these trials were given potentially active therapies that might have promoted ureteral-stone passage.
*Based on relative risk=1·65 applied to the control group event rates.
For this reason, they were not eligible for inclusion in our
Table 2: Number needed to treat for various background proportions of spontaneous stone passage
main analysis. Indeed, corticosteroids have been shown to increase stone passage rates.39 Although these studies did
as editorials or commentaries, or studies on the wrong not meet the strict inclusion criteria, we did a separate topic—eg, trials that used diﬀ erent interventions, trials sensitivity analysis in which their data were used to with diﬀ erent outcomes measured, or observational examine the eﬀ ect of their inclusion on the overall risk studies.
There were ﬁ ve additional randomised studies that
In seven of the studies pooled, both treatment and
made a substantial contribution to the literature, control groups received scheduled or on-demand doses of including the ﬁ rst trial of expulsive medical therapy.34–38
non-steroidal anti-inﬂ ammatory drugs (NSAIDs).40–46 These
These studies were excluded from the ﬁ nal meta-analysis drugs are highly eﬀ ective in the symptomatic relief of acute because none of them had a true control group. Speciﬁ cally, renal colic.47,48 Additionally, NSAIDs might augment Borghi and colleagues34 randomly assigned patients nife-
urinary stone expulsion.49,50 Yet the only randomised,
dipine and corticosteroids versus corticosteroids alone. double-blind, placebo-controlled trial to investigate the Dellabella and co-workers36,37 did two studies examining eﬀ ect of NSAIDs on stone-passage rates showed no the eﬃ
cacy of medical therapy. In their ﬁ rst,36 patients diﬀ erence between NSAIDs and control.51 None of the
with urolithiasis were randomly assigned tamsulosin and
studies that we pooled were designed to examine the eﬀ ect
corticosteroids versus corticosteroids alone. 2 years later, of NSAIDs on our primary outcome; however, we did a in their second study,37 patients were randomly assigned post-hoc subgroup analysis for the studies that included tamsulosin or nifedipine and corticosteroids versus NSAIDs in both treatment and control groups, as well as
See Online for webtable 1
corticosteroids alone. Additionally, the control groups in for the two studies in which no NSAIDs were used.52,53
We then reviewed the remaining 38 articles in detail to
determine if they met inclusion criteria. The ﬁ nal study
population consisted of nine relevant trials that examined
the use of calcium-channel blockers or α blockers to
augment urinary-stone passage.40–46,52,53
693 patients were randomised into the nine trials
included in the meta-analysis (table 1). All patients were
treated on an outpatient basis. The mean age of
participants ranged from 34·4 years to 46·5 years, and
the percentage of women in the studies varied from 25%
to 60%. Mean stone size ranged from 3·9 mm to 7·8 mm.
In all but one study,40 treated patients had stones located in the distal third of the ureter. There were 12 dropouts
No control group Borghi34
across all nine trials; seven patients from the intervention
groups and ﬁ ve from the control groups.
The medical treatments and follow-up as well as the
primary and secondary outcomes and recorded side-
eﬀ ects are shown in webtable 1. Treatment duration ranged from 7 days to 6 weeks, or until stone passage if
before than 6 weeks. Follow-up varied from 15 days to
48 days. In some trials, several drugs were given to the treatment groups; for three studies, corticosteroids were
given to the treatment groups in addition to the calcium-
Figure 3: Forest plot of risk ratios of stone passage, stratiﬁ ed by presence of control group
channel blocker nifedipine.40,42,43 The treatment and
Sizes of data markers are proportional to the weight of each study in the meta-analysis. Horizontal bars=95% CI.
control groups received NSAIDs in seven trials.40–46
www.thelancet.comVol 368 September 30, 2006 Articles
The primary outcome of interest—the proportion of
patients who passed their stones—occurred more often
Percentage of patients who Risk ratio (95% passed their stones
in the treatment groups than in the control groups in all nine studies. Figure 2 shows the percentage of
patients who passed stones, stratiﬁ ed by study group
and mean stone size for each of the studies. In six trials,
information was available for the mean time to stone
passage; mean time to passage ranged from 6 days in
several treatment groups to 20 days in one control
group.40,42,43,45,46,52 In ﬁ ve of these six trials, the treatment
group had shorter mean times to stone expulsion than
The ﬁ xed-eﬀ ects Mantel-Haenszel pooled risk ratio
was 1·65 (95% CI, 1·45–1·88), p<0·0001, indicating a
65% higher risk of stone passage associated with
medical therapy. There was no signiﬁ cant heterogeneity
in the studies pooled (χ2 test,p=0·196). The I2 statistic
was 28%. The pooled risk diﬀ erence was 0·31
(0·25–0·38). The number needed to treat was 4. When
dropouts were excluded, the pooled risk ratio was 1·66
(1·46–1·88), p<0·0001. The baseline occurrence of stone
passage in the control group, across all nine studies,
was 0·47. Table 2 shows the eﬀ ect of varying the
background occurrence of spontaneous stone passage
on the number needed to treat, which ranged from three
patients (background occurrence=60%) to 16 patients
(background occurrence=10%). The Rosenthal fail-safe
number was calculated as 175 studies, and the Rosenberg
fail-safe N was 105 additional non-signiﬁ cant studies
necessary to reduce the pooled risk ratio to
Calcium-channel blockers+steroids vs steroids
Figure 3 shows the study-speciﬁ c risk ratios and the
pooled estimate for the nine studies included, and for
the ﬁ ve that did not have a true control. When these ﬁ ve
α blockers vs calcium-channel blockers
studies were analysed with the previous nine, the overall
risk ratio remained signiﬁ cant at 1·52 (1·39–1·65),
p<0·0001. There was no signiﬁ cant heterogeneity with
inclusion of these studies (χ2=13·30,p=0·425). The I2
statistic was 2·3%, which indicates that the variation in
eﬀ ect estimates is probably due to sampling error
within trials rather than heterogeneity across trials.
The four studies in which treatment groups were given
Data are % (number) unless otherwise indicated. *Control groups also received phloroglucinol (or its trimethoxy
tamsulosin (without calcium-channel blockers) were
benzene derivative). Both treatment and control groups received cotrimoxazole and diclofenac. †Treatment and
pooled. This yielded a risk ratio of 1·52 (1·23–1·86),
control groups received a cocktail of medications including diazepam, tramadol, aesculus, and diclofenac. ‡Control group also received phloroglucinol. Both treatment and control groups received ketoprofene.
p<0·0001. When all studies in which an α blocker was used (n=5) were summarised, the pooled risk ratio was
Table 3: Outcomes of the studies pooled and of the studies excluded for absence of a true control,
1·54 (1·29–1·85), p<0·0001. In the two studies in which
stratiﬁ ed by treatment regimen
the intervention was nifedipine (without α blockers), the risk ratio was 1·51 (1·18–1·94), p=0·001. In three trials, to pool the two studies in which neither treatment groups the intervention was nifedipine with corticosteroids. nor control group received NSAIDs.52,53 The resulting risk When pooled, these studies resulted in a risk ratio of ratio was 1·74 (1·29–2·33), p<0·0001. The pooled risk 1·90 (1·51–2·40), p<0·0001. None of the pooled risk ratio for the seven studies in which NSAIDs were used in ratios showed statistically signiﬁ cant heterogeneity both treatment and control groups was 1·63 (1·41–1·88), across studies.
Although the eﬀ ects of calcium-channel blockers and
In some studies, corticosteroids were used in the
α blockers cannot be statistically isolated from those of treatment group; in others, they were used in both the other medications used, we did a subgroup analysis treatment and control groups. Table 3 summarises the
www.thelancet.comVol 368 September 30, 2006 Articles
given tamsulosin reported lower analogue pain scores.44 Fewer days lost from work, fewer emergency department
visits, and fewer surgical procedures in the treatment group than in the control group were reported in one
trial.40 Side-eﬀ ects were not rigorously reported for all
studies. However, the occurrence of therapy-related transient hypotension and palpitations was low at
Discussion The pooled results of the randomised trials suggest that
pharmacotherapy helps with passage of distal ureteral
stones. Patients treated medically with calcium-channel blockers or α blockers had a 65% greater likelihood of
spontaneous stone passage than did patients not given these drugs. This beneﬁ cial eﬀ ect was consistent for
both types of medical therapy. With the low risk-proﬁ le of these drugs and their wide therapeutic window, our
Figure 4: Inﬂ uence analysis
results suggest that treating physicians should consider a new algorithm for the management of urolithiasis, in
results for the 14 randomised controlled trials listed in which treatment begins with a course of medical the forest plot (ﬁ gure 3), with consideration of steroid therapy, unless medically contraindicated. use. When the studies of α blockers versus control41,44,46,52,53
Our ﬁ ndings are consistent with what is understood
were compared with those of α blockers and of ureteral pathophysiology associated with urinary-stone corticosteroids versus control,43 the incremental beneﬁ t obstruction. In animal models, ureteral stones result in of steroid use was small. A similar ﬁ nding was noted for
increased amplitude of ureteral smooth-muscle
steroid use with calcium-channel blockers. There were contraction, decreased frequency of contractions, and three studies in which the eﬀ ects of α blockers could be decreased ureteral pressure.12 Evidence suggests that directly compared with that of calcium-channel relaxing the ureter in the region of the stone and blockers.37,43,52 Two of these studies43,52 reported no increasing hydrostatic pressure proximal to the stone signiﬁ cant diﬀ erence in stone expulsion for the two drug
help to facilitate ureteral stone passage.17 Such relaxation
types, but one study37 noted that α blockers were better can be accomplished by giving adrenergic α-antagonists than calcium-channel blockers (risk ratio 1·26; and calcium-channel blockers,13,15 the eﬀ ects of which 1·10–1·44).
are mediated through the active calcium-channel pumps
We did an inﬂ uence analysis, in which the pooled and adrenergic α-1 receptors present in ureteral smooth
estimates were recalculated omitting one study at a time.
Figure 4 shows that the summary estimate of eﬀ ect
remained signiﬁ cant throughout this analysis. Resim
expulsive medical therapy for urolithiasis. Several
and colleagues’44 study had the largest eﬀ ect on the overall
studies have reported that patients given such treatment
estimate; with its omission, the overall risk ratio was 1·73
have a signiﬁ cantly reduced time to stone passage,42,43,45,46,52
signiﬁ cantly fewer pain episodes,44,46 lower analogue pain
Additionally, the three studies that included a scores,44 and need signiﬁ
description of the randomisation procedure (webtable 2)
analgesics.42,43,46 With the caveat that side-eﬀ ects were
were analysed separately. The pooled risk ratio from this
poorly categorised in these trials, calcium-channel
analysis was 2·14 (1·60–2·86), p<0·0001. We did a blockers and α blockers seemed to be well tolerated, separate sensitivity analysis of the six studies for which since there were only four patients across all studies
See Online for webtable 2
there was no loss to follow-up (webtable 2). The pooled who discontinued therapy.42,43risk ratio from this analysis was 1·52 (1·30–1·77),
When medical therapy for urolithiasis is successful,
p<0·0001. Finally, only those studies that were published
surgical intervention is unnecessary. This advantage is
as full manuscripts were pooled (webtable 2). The pooled
important because the risks related to surgical
risk ratio from this analysis was 1·68 (1·42–1·97), intervention are not trivial.23–25,27 Studies have reported p<0·0001.
overall complication rates after ureteroscopy of 10–20%,
Three trials showed less need for analgesics in the with major complications—eg, ureteral perforation,
treatment group than in the control group, as expressed
avulsion, and stricture—occurring in 3–5% of
by the mean amount of diclofenac used.42,43,46 Two studies
procedures.21–23 Accumulation of perirenal ﬂ uid and
showed fewer episodes of acute pain in patients given subcapsular bleeds have been reported in 15–32% of expulsive therapy,44,46 and one study showed that patients
patients treated with shock-wave lithotripsy.20,27 This risk
www.thelancet.comVol 368 September 30, 2006 Articles
is even more problematic since the re-treatment rate for
occurrence of stone passage in the controls is lower
shock-wave lithotripsy ranges from 4–50%.18,54
(30%), the sample size required would be 532; if the
Furthermore, the potential cost savings of expulsive occurrence is higher (60%), the study would require
medical therapies in lieu of surgical interventions is 110 patients. large. In the USA alone, total annual expenditure for
Thus the published evidence provides support for the
individuals with inpatient and outpatient claims for a use of expulsive medical therapy in the treatment of primary diagnosis of urolithiasis increased by 50% urolithiasis. Although minimally invasive procedures between 1994 and 2000. Of the US$2·1 billion spent on have evolved that allow for resolution of stone burden this disease in 2000, $490 million were for emergency with less morbidity than traditional open surgery, these department services.7 These numbers do not take into procedures expose patients to anaesthetic and surgical account the indirect costs, such as lost wages from risks that might be unnecessary. Although a large missed work.
conﬁ rmatory trial is advisable, our ﬁ ndings suggest that
Surgical intervention for urolithiasis is costly, with medical therapy might provide a viable alternative to
reported estimates ranging from US$2645 for surgery in patients with urolithiasis who are amenable to ureteroscopy to $4225 for shock-wave lithotripsy, with conservative management. repeated therapy often needed.18,26 Expulsive medical Contributors therapy, which relies mainly on generic drugs, is J M Hollingsworth participated in the study conception and design, the inexpensive. Based on drug-cost data obtained from the acquisition of data, the analysis and interpretation of data, drafting of the University of Michigan pharmacy, costs would range manuscript, critical revision of the manuscript, the statistical analyses,
and has seen and approved the ﬁ nal version. M A M Rogers participated
from US$10·74, for a 28-day course of doxazosin, to in the analysis and interpretation of data, drafting of the manuscript, $104·41 for a 42-day course of tamsulosin, the only critical revision of the manuscript, statistical analyses, supervision of the non-generic medication.
study, and has seen and approved the ﬁ nal version. S R Kaufman participated in the analysis and interpretation of data, critical revision of
Our main meta-analysis is potentially limited by clinical
the manuscript, statistical analyses, administrative and technical
(not statistical) heterogeneity, in view of the variation in support, supervision of the study, and has seen and approved the ﬁ nal the drugs given to the diﬀ erent treatment groups. To version. T J Bradford participated in the acquisition of data, the analysis address this issue, we analysed drug classes separately. and interpretation of data, critical revision of the manuscript,
administrative support for the study, and has seen and approved the
Both classes of drugs seemed to be beneﬁ cial, though the
ﬁ nal version. S Saint participated in the study conception and design,
results suggested that there might be an additive eﬀ ect the interpretation of data, critical revision of the manuscript, when combined with corticosteroids. Further exploration
administrative and technical support, supervision of the study, and has
of combined treatments might be useful.
seen and approved the ﬁ nal version. J T Wei participated in the analysis and interpretation of data, critical revision of the manuscript,
Furthermore, our results might have been aﬀ ected by supervision of this study, and has seen and approved the ﬁ nal version.
publication bias, in which positive studies are more likely
B K Hollenbeck participated in the study conception and design, the
to be submitted and published than negative ones. acquisition of data, analysis and interpretation of data, drafting of the Although the fail-safe numbers did not show evidence of
manuscript, critical revision of the manuscript, supervision of the study, and has seen and approved the ﬁ nal version.
publication bias, this potential drawback can only be adequately assessed through registration of prospective Conﬂ ict of interest statement
J M Hollingsworth, M A M Rogers, S R Kaufman, T J Bradford, and
trials and not through retrospective review of published B K Hollenbeck declare that they have no conﬂ ict of interest. S Saint and
studies, as exists here.55 We did speciﬁ cally ask T J Wei have been paid consultants to Sanoﬁ Aventis within the past investigators and the drug companies about unpublished
5 years (none of Sanoﬁ Aventis’ products were used in the trials
studies in our correspondence. All ﬁ ve investigators and
one of the three drug companies that responded to our Acknowledgments inquiry denied knowledge of unpublished data.
We thank Loris Borghi, Thomas Cooper, Bora Küpeli, Sefa Resim,
Erdal Yilmaz, and Christopher Parow, director of Global Scientiﬁ c
limitation relates to the overall quality of the trials, eight
Support at Abbott Laboratories, for responding to our enquiries about
of which were not blinded and six of which did not their studies; Joyce Chen from Global Drug Information Services at describe the randomisation procedures in detail. Schering-Plough Corporation, who provided us with an additional Additionally, most studies were done in Mediterranean literature search at our unsolicited request; Rita Galin, Russian
interpreter from the University of Michigan Health System, for her
countries, and it is unknown whether the treatment help in translating an article for review; and Alison Grodzinski, response would vary in patients from diﬀ erent settings.
research librarian at the University of Michigan Medical School, who
A deﬁ nitive high-quality randomised controlled trial is did an independent literature search to conﬁ rm the exhaustiveness of
cacy of calcium-channel our own. J M Hollingsworth is supported by a National Institutes of
Health research-training grant (T32 DK007758). M A M Rogers is
blockers and α blockers in patients with urolithiasis. supported by a Blue Cross Blue Shield of Michigan Foundation award Since preliminary studies tend to overestimate treatment
(793.RFP) an Agency for Healthcare Research and Quality award (R01
eﬀ ects, we used the lower conﬁ dence limit of the pooled
HS015571). S R Kaufman is supported by a National Institutes of Health/National Institute of Diabetes and Digestive and Kidney
relative risk (1·45) and determined that a two-armed trial
Diseases award (R21 DK067451-01A1) and a Blue Cross Blue Shield of
would need to include 226 patients (113 in each arm), in Michigan Foundation Award (793.RFP). T J Bradford is supported by a view of the background occurrence of stone passage of DOD FY05 Idea Development with Nested Resident and Medical 0·47, with α=0·05 (two-sided) and power=0·90. If the Student Traineeship Award (PC051081). S Saint is supported by a
www.thelancet.comVol 368 September 30, 2006 Articles
National Institutes of Health/National Institute of Diabetes and
20 Rubin JI, Arger PH, Pollack HM, Banner MP, Coleman BG,
Digestive and Kidney Diseases award (R21 DK067451-01A1), an
Mintz MC. Kidney changes after extracorporeal shock wave
Advanced Career Development Award from the Department of
lithotripsy: CT evaluation. Radiology 1987; 162: 21–24.
Veterans Aﬀ airs Health Services Research and Development Oﬃ
21 Harmon WJ, Sershon PD, Blute ML, Patterson DE, Segura JW.
(RCD-00-006), a Merit Award from the VA Health Services Research
Ureteroscopy: current practice and long-term complications. J Urol
ce (SAF 04-031), and a grant from the Blue Cross
1997; 157: 28–32.
Blue Shield of Michigan Foundation (793.RFP).
22 Schuster TG, Hollenbeck BK, Faerber GJ, Wolf JSJ. Complications
J T Wei is supported by a National Institutes of Health/The National
of ureteroscopy: analysis of predictive factors. J Urol 2001; 166:
Institute of Child Health and Human Development award (U01
HD041249-01), a National Institutes of Health/National Cancer
23 Daniels GF, Garnett JE, Carter MF. Ureteroscopic results and
Institute award (R01 CA095662), a National Institutes of Health/
complications: experience with 130 cases. J Urol 1988; 139: 710–13.
National Cancer Institute award (U01 CA111275), a National Cancer
24 el-Faqih SR, Husain I, Ekman PE, Sharma ND, Chakrabarty A,
Institute SPORE award (P50 CA069568), a National Institutes of
Talic R. Primary choice of intervention for distal ureteric stone: ureteroscopy or ESWL? Br J Urol 1988; 62: 13–18.
Health/National Cancer Institute award (U01 CA113913, and a National Institutes of Health developmental award (P50 DK065313).
25 Kramolowsky EV. Ureteral perforation during ureterorenoscopy:
treatment and management. J Urol 1987; 138: 36–38.
B K Hollenbeck is supported by a National Institutes of Health/
26 Lotan Y, Gettman MT, Roehrborn CG, Cadeddu JA, Pearle MS.
National Institute of Diabetes and Digestive and Kidney Diseases
Management of ureteral calculi: a cost comparison and decision
developmental award (P50 DK065313-01), the John and Suzanne Munn
making analysis. J Urol 2002; 167: 1621–29.
Endowed Research Fund of the University of Michigan Comprehensive
27 Preminger GM. Sonographic piezoelectric lithotripsy: more bang
Cancer Center, and the Clinical Scholars Science Program at the
for your buck. In: Lingeman JE, Newman DM, eds. Shock wave
lithotripsy urinary and biliary lithotripsy. New York: Plenum; 1989:
Ljunghall S. Renal stone disease. Studies of epidemiology and
28 Moher D, Cook DJ, Eastwood S, Olkin I, Rennie D, Stroup DF, for
calcium metabolism. Scan J Urol Nephrol 1977; 41: 1–96.
the QUOROM Group. Improving the quality of reports of
Pak CY, Resnick MI, Preminger GM. Ethnic and geographic
meta-analyses of randomised controlled trials: the QUOROM
diversity of stone disease. Urology 1997; 50: 504–07.
statement. Lancet 1999; 354: 1896–900.
Tiselius HG. Epidemiology and medical management of stone
29 Higgins JP, Thompson SG. Quantifying heterogeneity in a
disease. BJU Int 2003; 91: 758–67.
meta-analysis. Stat Med 2002; 21: 1539–58.
Sierakowski R, Finlayson B, Landes RR, Finlayson CD,
30 Newcombe RG. Interval estimation for the diﬀ erence between
Sierakowski N. The frequency of urolithiasis in hospital discharge
independent proportions: comparison of eleven methods. Stat Med
diagnoses in the United States. Invest Urol 1978; 15: 438–41.
1998; 17: 873–90.
Stamatelou KK, Francis ME, Jones CA, Nyberg LM, Curhan GC.
31 Rosenthal R. The ‘ﬁ le drawer problem’ and tolerance for null
Time trends in reported prevalence of kidney stones in the United
results. Psychol Bull 1979; 86: 638–41.
States: 1976-1994. Kidney Int 2003; 63: 1817–23.
32 Rosenberg MS. The ﬁ le-drawer problem revisited: a general
Coe FL, Keck J, Norton ER. The natural history of calcium
weighted method for calculating fail-safe numbers in meta-analysis.
urolithiasis. JAMA 1977; 238: 1519–23. EvolutionInt J Org Evolution 2005; 59: 464–68.
Pearle MS, Calhoun EA, Curhan GC. Urologic Diseases of
33 Higgins JPT, Green S, eds. Cochrane handbook for systematic
America Project. Urologic diseases in America project:
reviews of interventions 4.2.5). In: The Cochrane Library, Issue 3,
urolithiasis. J Urol 2005; 173: 848–57.
2005. Chichester, UK: John Wiley & Sons, 2005.
Hubner WA, Irby P, Stoller ML. Natural history and current
34 Borghi L, Meschi T, Amato F, et al. Nifedipine and
concepts for the treatment of small ureteral calculi. Eur Urol 1993;
methylprednisolone in facilitating ureteral stone passage: a
randomized, double-blind, placebo-controlled study. J Urol 1994; 152: 1095–98.
Miller OF, Kane CJ. Time to stone passage for observed ureteral calculi: a guide for patient education. J Urol 1999; 162: 688–90.
35 Cervenakov I, Fillo J, Mardiak J, Kopecny M, Smirala J, Lepies P.
Speedy elimination of ureterolithiasis in lower part of ureters with
10 Ueno A, Kawamura T, Ogawa A, Takayasu H. Relation of
spontaneous passage of ureteral calculi to size.
Int Urol Nephrol 2002; 34: 25–29. Urology 1977; 10:
36 Dellabella M, Milanese G, Muzzonigro G. Eﬃ
the medical management of juxtavesical ureteral stones.
11 Sigala S, Dellabella M, Milanese G, et al. Evidence for the presence
2003; 170: 2202–05.
of alpha1 adrenoceptor subtypes in the human ureter. Neurourol Urodyn 2005; 24: 142–48.
37 Dellabella M, Milanese G, Muzzonigro G. Randomized trial of the
cacy of tamsulosin, nifedipine and phloroglucinol in medical
12 Laird JM, Roza C, Cervero F. Eﬀ ects of artiﬁ cial calculosis on rat
expulsive therapy for distal ureteral calculi.
ureter motility: peripheral contribution to the pain of ureteric
J Urol 2005; 174: 167–72.
colic. Am J Physiol 1997; 272: R1409–16.
38 Staerman F, Bryckaert PE, Colin J, Youinou Y, Brandt B,
Lardennois B. Nifedipine in the medical treatment of symptomatic
13 Maggi CA, Giuliani S. A pharmacological analysis of calcium
channels involved in phasic and tonic responses of the guinea-pig
Eur Urol 2005; 37: 28.
ureter to high potassium. J Auton Pharmacol 1995; 15: 55–64.
39 Salehi M, Fouladi MM, Shier H, Mokhtari G, Dejabad V. Does
methylprednisolone acetate increase the success rate of medical
14 Malin JM, Deane RF, Boyarsky S. Characterisation of adrenergic
therapy for patients with distal ureteral stones.
Br J Urol 1970; 42: 171–74.
15 Morita T, Wada I, Suzuki T, Tsuchida S. Characterization of
40 Cooper JT, Stack GM, Cooper TP. Intensive medical management
alpha-adrenoceptor subtypes involved in regulation of ureteral
Urology 2000; 56: 575–78. Tohoku J Exp Med 1987; 152: 111–18.
41 Kupeli B, Irkilata L, Gurocak S, et al. Does tamsulosin enhance
16 Salman S, Castilla C, Vela NR. Action of calcium antagonists on
lower ureteral stone clearance with or without shock wave
ureteral dynamics. Actas Urol Esp 1989; 13: 150–52.
lithotripsy? Urology 2004; 64: 1111–15.
17 Sivula A, Lehtonen T. Spontaneous passage of artiﬁ cial concretions
42 Porpiglia F, Destefanis P, Fiori C, Fontana D. Eﬀ ectiveness of
applied in the rabbit ureter. Scan J Urol Nephrol 1967; 1: 259–63.
nifedipine and deﬂ azacort in the management of distal ureter
18 Bierkens AF, Hendrikx AJ, De La Rosette JJ, et al. Treatment of
stones. Urology 2000; 56: 579–82.
mid- and lower ureteric calculi: extracorporeal shock-wave
43 Porpiglia F, Ghignone G, Fiori C, Fontana D, Scarpa RM.
lithotripsy vs laser ureteroscopy. A comparison of costs, morbidity
Nifedipine versus tamsulosin for the management of lower ureteral
and eﬀ ectiveness. Br J Urol 1998; 81: 31–35.
stones. J Urol 2004; 172: 568–71.
19 Hochreiter WW, Danuser H, Perrig M, Studer UE. Extracorporeal
44 Resim S, Ekerbicer H, Ciftci A. Eﬀ ect of tamsulosin on the number
shock wave lithotripsy for distal ureteral calculi: what a powerful
and intensity of ureteral colic in patients with lower ureteral
machine can achieve. J Urol 2003; 169: 878–80.
calculus. Int J Urol 2005; 12: 615–20.
www.thelancet.comVol 368 September 30, 2006 Articles
45 Skrekas T, Liapis D, Kalantzis A, Argyropoulos A, Doumas K,
51 Laerum E, Ommundsen OE, Gronseth JE, Christiansen A,
Lycourinas M. Increasing the success rate of medical therapy for
Fagertun HE. Oral diclofenac in the prophylactic treatment of
expulsion of distal ureteral stones using adjunctive treatment with
recurrent renal colic. A double-blind comparison with placebo.
calcium channel blocker. Eur Urol Suppl 2003; 2: 82. Eur Urol 1995; 28: 108–11.
46 Yilmaz E, Batislam E, Basar MM, Tuglu D, Ferhat M, Basar H. The
52 Taghavi R, Darabi MR, Tavakoli K, Keshvari M. Survey of the eﬀ ect
cacy of 3 diﬀ erent alpha1-adrenergic blockers
of tamsulosin and nifedipine on facilitating juxtavesical ureteral
for distal ureteral stones. J Urol 2005; 173: 2010–12.
stone passage. J Endourol 2005; 19 (suppl 1): A9.
47 Cordell WH, Larson TA, Lingeman JE, et al. Indomethacin
53 Tekin A, Alkan E, Beysel M, Yucebas E, Asian R, Sengor F. Alpha-1
suppositories versus intravenously titrated morphine for the
receptor blocking therapy for lower ureteral stones: a randomized
treatment of ureteral colic. Ann Emerg Med 1994; 23: 262–69.
prospective trial. J Urol 2004; 17 (suppl 4): 304.
48 Cordell WH, Wright SW, Wolfson AB, et al. Comparison of
54 Gerber R, Studer UE, Danuser H. Is newer always better? A
intravenous ketorolac, meperidine, and both (balanced analgesia)
comparative study of 3 lithotriptor generations. J Urol 2005; 173:
for renal colic. Ann Emerg Med 1996; 28: 151–58.
49 Al-Waili NS. Prostaglandin synthetase inhibition with
55 Thornton A, Lee P. Publication bias in meta-analysis: its causes and
indomethacin rectal suppositories in the treatment of acute and
consequences. J Clin Epidemiol 2000; 53: 207–16.
chronic urinary calculus obstruction. Clin Exp Pharmacol Physiol 1986; 13: 195–99.
50 Ahmad M, Chaughtai MN, Khan FA. Role of prostaglandin
synthesis inhibitors in the passage of ureteric calculus. JPMA J Pakistan Med Assoc 1991; 41: 268–70.
www.thelancet.comVol 368 September 30, 2006
CLINICIAN’S CORNER A 67-Year-Old Man Who e-Mails His Physician Warner V. Slack, MD, Discussant I probably e-mail my doctor once every 2 weeks. If I havea concern, it might be more often than that. I think that health DR SHIP: Mr S is a 67-year-old retired public service worker issues are important, and by e-mailing my doctor and get-who lives in the Boston area with his wife. He has M
Chapter 4 Conventional Medical Therapies “Today’s standard, AMA-approved medicine is rooted in treating symptoms, rather than causes. Its dependence on drugs and surgery is ruinously expensive to patients, insurance companies, “Why I Left Orthodox Medicine” Conventional medical treatments for FMS and CFS is a controversial topic. Consider the following statements