approaches 80%, and achievement of pregnancy ap-
proaches 38% when the obstructive interval is
<
11 years.
6
Unfortunately, studies that have carefully evaluated and
reported data on vasal
fl
uid quality have generally been
small, retrospective, and based on data from single
institutions.
A meta-analysis evaluating the outcomes of VV did
not assess intraoperative vasal
fl
uid characteristics.
7
Although 1 other group has reviewed this topic,
8
the
data were not synthesized using a meta-analytic frame-
work. Therefore, we performed a systematic review and
meta-analysis of the published literature to evaluate the
association between the presence of sperm or sperm parts
in the vasal
fl
uid and patency after VV.
METHODS
Study Design
This study was a systematic review and meta-analysis. An a
priori protocol was written and agreed, by the authors, to
include study design, search strategy, inclusion and exclusion
criteria, primary outcomes, statistical methods, and bias assess-
ment. We followed the Preferred Reporting Items for Systematic
Reviews and Meta-Analyses guidelines for performing and
reporting a meta-analysis.
Literature Search
English-language studies reporting on outcomes of microscopic
VV for VR between November 1977 (the
f
rst report of
microsurgical VV
9
) and March 2014 were sought by electronic
search of MEDLINE, scanning the reference lists of identi
f
ed
articles, and correspondence with study investigators. The
computer-based search included variations of the terms
“
vasec-
tomy reversal
”
and
“
vasovasostomy.
”
Study Selection
Studies were eligible for inclusion if they used a microscopic VV
approach and reported on outcomes for
±
10 patients. Studies of
patients undergoing VR for reasons other than a desire for
fertility such as a history of epididymitis, hernia repair, idio-
pathic obstruction, or trauma were excluded. If multiple publi-
cations reporting on the same patient population were
identi
f
ed, only the latest study was included.
Data Collection
The following information was independently extracted by 2
reviewers from each article using a standardized form: study
population (including population source, sampling method used,
sample size, and patient demographic characteristics); geographic
location; publication year; mean patient age and obstructive in-
terval at the time of VR; number of patients with sperm or sperm
parts in intraoperative vasal
fl
uid; de
f
nition of postoperative
patency; and number of patients achieving patency.
Data Synthesis
All analyses were performed using only within-study compari-
sons to limit possible biases. The mean age and obstructive in-
terval at VR reported by each study were combined and
summarized using an arithmetic mean weighted by study sample
size. An odds ratio (OR) and corresponding 95% con
f
dence
interval (CI) for the association between the presence of
intravasal sperm or sperm parts and postoperative patency were
calculated for each study. To include the results of case series
reporting incomplete data, 0.5 was added to each count in 2-by-
2 contingency tables that contained a value of 0 in any cell.
10
Meta-analysis was performed using a random-effects model.
The consistency of
f
ndings across studies was assessed using the
Cochran Q test
11
and the I
2
statistic.
12
Publication bias was
assessed by the funnel plot and Egger et al
13
regression asym-
metry analysis. Statistical signi
f
cance was de
f
ned as a 2-tailed
P
value
<
.05. Analyses were performed using R version 3.0.0 (R
Foundation for Statistical Computing, Vienna, Austria).
RESULTS
Studies Included for Systematic Review and
Meta-analysis
Four case series
6,14-16
and 2 retrospective cohort studies
3,17
of 1293 eligible patients were identi
f
ed (
Fig. 1
). The
studies were published between 1977 and 2014 and took
place in Asia, Australia, and the United States (
Table 1
).
Reported sample sizes ranged from 11 to 605 patients un-
dergoing microsurgical VV. The weighted mean patient
age was 37.8 years and obstructive interval was 7.1 years at
the time of VR. The studies used wet-mount light micro-
scopy to assess for the presence of sperm in the vasal
fl
uid.
The de
f
nition of patency used by the studies varied
markedly. The Vasovasostomy Study Group
3
and Sigman
15
de
f
ned patency as the presence of whole sperm or sperm
parts in the postoperative semen analysis; Sheynkin et al
14
de
f
ned it as the presence of whole sperm but not sperm
parts; and Bolduc et al,
17
Kolettis et al,
6
and Smith et al
16
de
f
ned it as the presence of motile sperm. Three of the 6
studies
3,16,17
de
f
ned whole sperm as being
“
mostly normal
and motile
”
or
“
mostly normal and nonmotile.
”
In these 3
studies, sperm parts were de
f
ned as
“
mostly heads without
tails
”
or
“
only heads without tails
”
on intraoperative ex-
amination. Sigman
15
de
f
ned sperm parts as
“
short tails
”
or
“
sperm heads.
”
Kolettis et al
6
and Sheynkin et al
16
did not
differentiate between whole sperm and sperm parts.
Meta-analysis
The unadjusted OR of postoperative patency was 4.1
times higher (95% CI, 2.3-7.3) given the presence of
intravasal sperm or sperm parts as opposed to their
absence at the time of VR (
Fig. 2
). The pooled OR
should be interpreted with caution as only the 2 retro-
spective cohort studies reported meaningful data on this
association. Although surrogate statistical techniques
were used to include case series in this meta-analysis,
10
their data did not contribute signi
f
cantly to the pooled
OR. There was no evidence of statistically signi
f
cant
heterogeneity among the 6 studies (Q
¼
3.4; df
¼
5;
P
¼
.6; I
2
¼
22%) nor was there obvious publication bias by
the funnel plot (
Fig. 3
) or Egger et al
13
regression analysis
(z
¼²
0.4;
P
¼
.7).
Sensitivity Analysis
A sensitivity analysis in which case series were excluded
yielded an unadjusted OR of 4.3 (95% CI, 2.2-8.3) and
810
UROLOGY 85 (4), 2015