Study reference

Study characteristics

Patient characteristics

Prognostic factor(s)

Follow-up

Outcome measures and effect size

Comments

Davis, 2017

Type of study: retrospective observational cohort study.

Setting: hospital outpatients, single centre.

Country: Ireland

Source of funding: not reported.

Inclusion criteria: sample of consecutive men undergoing vasectomy reversal.

Exclusion criteria: none reported.

N= 106

Mean age ± SD:

Age at vasectomy: 33 ± 6 (24-52)

Age at reversal: 42 ± 7 (range 30-57)

Age of female partner: 34 ± 4 (27-45)

Sex: 100%M

Other important characteristics (confounders and effect modifiers):

Mean PVOI ± SD: 8 years (SD unknown) (3 months – 22 years).

Bilateral VV: N=103

Unilateral VV: N=3.

Describe prognostic factor(s) and method of measurement:

Prognostic factors were male age at reversal, male smoking status, female partner age at reversal, PVOI.

Methods of measurement were not reported.

Endpoint of follow-up: not reported.

For how many participants were no complete outcome data available?

Not reported.

Reasons for incomplete outcome data described?

Not reported. Presumably only patients with compete data were selected.

Incidence of live births: 41.0%

Live births

Determined by direct patient inquiry.

Results from multivariate logistic regression analysis. The outcome refers to the odds for a live birth compared to no live birth.

Significant associations:

PVOI: OR: 0.977 (95%CI unknown) (β=-0.023); P=0.007.

Female age at reversal: OR: 0.27 (95%CI unknown) (β=-1.309); P=0.043.

Non-significant associations:

Male age at time of reversal: OR 0.944 (95%CI unknown) (β= -0.058); P=0.287.

Male smoking: OR 0.621 (95%CI unknown) (β= -0.477); P=0.436.

Vasectomy reversal:

VV was performed by a single urologist. Under general anaesthesia via a midline 2-3 cm incision, the vasectomy site was bypassed, by end to end anastomoses, completing a modified 1-layer technique using 7-0 moncryl sutures over a 2-0 nylon suture stent in all procedures. Visual aid was by surgical loupes, not operating microscope. Post-operative care included scrotal support dressing, analgesia and suggested rest. Stent sutures were removed at day 5.

NB:

- The following factors were presumably dropped for multivariate analysis due to non-significant associations in univariate analysis: male age at vasectomy indication for reversal, indication for cancelled reversal, concomitant TESE. In addition, patency was studied as a prognostic factor but the OR was extremely high and non-significant. This factor was probably studied by mistake.

- A proportion of couples with risk factors for failed vasectomy reversal (prolonged PVOI and/or increased female age) were offered intraoperative sperm retrieval by TESE with cryopreservation concomitant with reversal procedure. In the event of failed reversal such couples could proceed directly to ICSI. TESE technique: transverse 5mm incision through the tunica albuginea whereby testicular parenchyma was extruded to deliver a 5x5 mm fragment.

- Demographic data on male smoking status not reported.

- Odds ratios were calculated by hand.

- N (%): 24/116 (20.7%) people were excluded from study population. Of the 116 men scheduled for vasectomy reversal, there were 10 cancellations due to cost (N=3), change of partner (N=3) and change of mind (N=4).

- Reasons for missing pregnancy outcome data were: relocation (N=7), refused study inclusion (N=7).

Van Dongen (2011)

Type of study: retrospective cohort study.

Setting: hospital outpatients, single centre.

Country: The Netherlands.

Source of funding: not reported.

Inclusion criteria: all patients undergoing primary or repeat vasectomy reversal.

Exclusion criteria: not reported.

N= 162

Mean age ± SD:

Age at reversal: 42.6 ± 5.9 years.

Partner’s age: 32.4 ± 4.9 years.

Sex: 100%M

Other important characteristics (confounders and effect modifiers):

Male smokers: 29.6%

Partner smokers: 25.3%

Mean PVOI ± SD: 7.9 ± 4.4 years

Bilateral VV: 67.9%

Unilateral VV: 3.1%

Bilateral VE: 11.7%

Unilateral VE: 3.7%;

Combined unilateral VV and VE: 13.6%.

Describe prognostic factor(s) and method of measurement:

Prognostic factors were smoking status men, smoking status partner, male age, partner age, PVOI, surgical procedure, sperm concentration, total forward motility.

Standard data were obtained by chart review. Follow-up data, including pregnancies, use and timing of ARTs, smoking habits were collected by telephone interview using standardized questionnaire.

Endpoint of follow-up: follow-up was at least 1 year. Mean duration of follow-up was 3.4 years (SD unknown), median was 2.9 years (range unknown).

For how many participants were no complete outcome data available?

Not reported.

Reasons for incomplete outcome data described?

Not reported. Presumably only patients with compete data were selected.

Incidence of pregnancy: 58.6%

Pregnancy

Pregnancy status was determined by questionnaire via telephone.

Results from multivariate cox regression analysis. The outcome refers to the hazard ratio to obtain a pregnancy compared to no pregnancy during follow-up.

Significant associations:

Total forward motility <5% vs. >20%: HR 0.28 (95%CI 0.119-0.652); P=0.003.

Non-significant associations:

Smoker: HR 0.78 (95%CI 0.45-1.37); P=0.395.

Partner smoker: HR 1.34 (95%CI 0.76-2.36); P=0.318.

Age (years): HR 1.00 (95%CI 0.95-1.05); P=0.911.

Partner age (years): HR 1.00 (95%CI 0.96-1.04); P=0.992.

PVOI: HR 1.00 (95%CI 0.99-1.01); P=0.363;

Surgical procedure (VV vs. VE): HR 0.53 (95%CI 0.24-1.18); P=0.118.

Sperm concentration: HR 1.00 (95%CI 0.99-1.09); P= 0.594.

Total forward motility 5-20% vs. >20%: HR 0.50 (95%CI 0.246-1.028); P=0.060.

Vasectomy reversal:

surgery was performed under general anaesthesia. Before reconstruction, efflux of the proximal vas portion was evaluated for gross appearance and microscopic presence of sperm. If clear watery efflux or any efflux containing whole sperm was found, VV using a modified one-layer technique was performed using 9-0 ethylon sutures. If thick white ( ‘ toothpaste-like ’ ) or milky fluid containing no sperm or only sperm heads was found, VE using a two stitch intussception technique was performed with 10-0 ethylon sutures. Patients were released on the day of surgery.

NB:

- N (%): 24/186 (12.9%) men were excluded from study population. N=15 could not be contacted, N=9 did not attempt pregnancy.

Friedler (2002)

Type of study: retrospective observational cohort study.

Setting: hospital outpatients, single centre.

Country: Israel.

Source of funding: not reported.

Inclusion criteria: patients with obstructive (OA) and non-obstructive azoospermia (NOA) who underwent surgical sperm retrieval by PESA or TESE. Whenever mature sperm cells were obtained, ICSI was performed.

Exclusion criteria: not reported.

N= 175, of which 52 OA, 123 NOA.

Mean age ± SD:

Age at reversal:

NOA: 34.5 ± 5.9

OA: 35.6 ± 7.6

Age of female partner:

NOA: 30.9 ± 5.8

OA: 32.4 ± 6.8

Sex: 100%M

Other important characteristics (confounders and effect modifiers):

- NOA: 142 TESE cycles; 128 ICSI cycles.

- OA: 55 PESA cycles; 135 ICSI cycles.

Describe prognostic factor(s) and method of measurement:

Prognostic factors were age at reversal, serum FSH level, sperm origin (epididymal vs. testicular), sperm status (fresh vs. frozen), age of female partner, number of oocytes retrieved, number of oocytes injected.

Method of measurement was not reported. Presumably, data were extracted from patient files.

Endpoint of follow-up: not reported. Duration of follow-up not reported.

For how many participants were no complete outcome data available?

Not reported.

Reasons for incomplete outcome data described?

Not reported. Presumably only patients with compete data were selected.

Incidence of pregnancy:

Number of pregnancies/number of patients:

NOA: 30.9%

OA: 82.7%

Incidence of ongoing/delivered pregnancy rate/embryo transfer:

NOA: 26.5%

OA: 33.9%

Pregnancy

Pregnancy was determined based on sonographic demonstration of a gestational sac.

Results from multivariate logistic regression analysis. The outcome refers to the odds for a pregnancy compared to no pregnancy.

Significant associations:

Female partner’s age (90^th quantile: 38 years) <38 vs >38 years: OR 6.46 (95%CI unknown); P=0.0052.

Number of oocytes retrieved (10^th quantile: 4 oocytes) >4 vs. <4 oocytes: OR 95.3 (95%CI unknown); P=0.0285.

Number of oocytes injected (10^th quantile: 3 oocytes): OR 0.002 (95%CI unknown); P=0.0006.

Non-significant associations:

Sperm status: OR 1.26 (95%CI unknown); P=0.51.

Sperm origin: OR 1.11 (95%CI unknown); P=0.81.

Serum FSH level: OR 0.43 (95%CI unknown); P=0.35).

Male age: OR 1.67 (95%CI unknown); P=0.63).

NB:

- diagnosis of OA/NOA based on histological report.

- there were early abortions reported:

NOA: 7/38 (18.4%)

OA: 2/43 (4.6%)

- Analysis refer to the odds for clinical pregnancy, presumably this refers to all pregnancies without spontaneous abortions.

Important: results are for NOA and OA patients analysed together. Sensitivity analysis stratified for NOA/OA showed similar results (data not shown).

Gerrard (2007)

Type of study: retrospective observational cohort study.

Setting: hospital outpatients, single centre.

Country: USA.

Source of funding: not reported.

Inclusion criteria: all men who underwent microsurgical vasectomy reversals.

Exclusion criteria: patients with less than 12 months follow-up or those not actively attempting to conceive were excluded from the pregnancy rate analysis.

N= 294

Mean age ± SD:

Age at reversal: 39.7 ± 6.6.

Age of female partner: 31.3 ± 4.6

Sex: 100%M

Other important characteristics (confounders and effect modifiers):

Mean PVOI ± SD: 9.4 years ± 5.3.

Couples with female factors: 6.5%.

Couples with repeat procedures: 9.9%.

Sperm observed intra-operatively: 89.9%.

Describe prognostic factor(s) and method of measurement:

Prognostic factors were female partner age, male partner age, PVOI, female factors (not specified), repeat procedures (not specified), surgical technique (VV vs. VE), sperm observed intraoperatively, and follow-up time (when available, studied in separate analysis)

A semen analysis was obtained between 4 weeks – 3 months post-surgery, and generally every 3 months until pregnancy was established or the patient discontinued follow-up.

Follow-up data were obtained from review of the medical records, phone contact, and letters from patients.

Endpoint of follow-up:

All participants included in the pregnancy rate analysis (N=208) had a minimum of 12 months follow-up until pregnancy was reported. Exact duration of follow-up was known for N=152: mean duration of follow-up was 32.5 months ± 27.0.

For how many participants were no complete outcome data available?

N (%): 86 (29.3%) with no pregnancy data.

Reasons for incomplete outcome data described?

Patients with less than 12 months follow-up or those not actively attempting to conceive were excluded from the pregnancy rate analysis. (Percentages unknown)

Incidence of pregnancy: 54.3% (N=208)

Pregnancy

Pregnancy data was extracted by review of the medical records, phone contact, and letters from patients.

Results from multivariate logistic regression analysis. The outcome refers to the odds for a pregnancy compared to no pregnancy.

Significant associations (N=208):

PVOI: OR unknown (95%CI unknown); P=0.025

Presence of sperm or sperm parts intraoperatively: OR unknown (95%CI unknown);P= 0.033

Significant associations (for participants with >12 months follow-up N=152):

Presence of sperm or sperm parts intraoperatively: OR unknown (95%CI unknown);P= 0.048

Follow-up time: OR unknown (95%CI unknown); P=0.049.

Non-significant associations (regardless of duration of follow-up):

Female factors (OR unknown (95%CI unknown);P=0.092). Age of female partner, male age, repeat procedures, and surgical technique (ORs unknown (95%CIs unknown); P-values unknown).

Non-significant associations (for participants with >12 months follow-up N=152):

PVOI (OR unknown (95%CI unknown); P=0.083. Female factors, age of female partner, male age, repeat procedures, and surgical technique (ORs unknown (95%CIs unknown); P-values unknown).

Vasectomy reversal:

was performed by one of three urologic surgeons under general anaesthesia with either a modified one-layer technique or a formal two-layer technique.

NB:

- Patients followed for 1 year who did not establish a pregnancy were considered not to be pregnant.

- There were no effect sizes reported, only P-values.

- Pregnancy rate for N=152 (>12 months follow-up) was not reported.

Hinz (2008)

Type of study: retrospective observational cohort study.

Setting: hospital outpatients, single centre.

Country: Germany.

Source of funding: not reported.

Inclusion criteria: all men undergoing vasectomy reversal for obstructive azoospermia. NB Indications for surgery were prior vasectomy (n = 177), inflammation (n = 18), and prior unsuccessful vasectomy reversal (n = 13). In 4 cases, the reason for obstruction was not known or documented.

Exclusion criteria: patients whose partners achieved pregnancy by ART.

N=212

Mean age ± SD:

Age at reversal: 40.4 years (SD not reported).

Age of female partner: 31.3 years (SD not reported)

Sex: 100%M

Other important characteristics (confounders and effect modifiers):

Indications for surgery: prior vasectomy (N=177), inflammation (N=18), prior unsuccessful vasectomy reversal (N=13), unknown (N=4).

Silber score, stratified for age of female partner:

1-2: 47.9% (≤29 years), 40.0% (30-39), 35.7% (≥40 years).

Sperm granuloma, stratified for age of female partner:

No granuloma: 76.4% (≤29 years), 83.7% (30-39), 50.0% (≥40 years).

Describe prognostic factor(s) and method of measurement:

Prognostic factors were female partner age, Silber score grouping, sperm granuloma.

Method of measurement was not reported.

Endpoint of follow-up: not reported, duration of follow-up not reported.

For how many participants were no complete outcome data available?

Not reported.

Reasons for incomplete outcome data described?

Not reported. Presumably only patients with compete data were selected.

Incidence of pregnancy: 72.2%.

Pregnancy

Unknown how pregnancy was determined.

Results from multivariate logistic regression analysis. The outcome refers to the odds for a pregnancy compared to no pregnancy, corrected for male age.

Significant associations:

Female partner age ≥40 vs. ≤29 years: OR 0.24 (95%CI 0.07-0.88); P overall=0.010.

Silber score 5 vs. 1-2: OR 0.31 (95%CI 0.11-0.92); P overall =0.024.

Non-significant associations:

Female partner age 30-39 vs. ≤29 years: OR 1.62 (95%CI 0.82-3.20); P overall= 0.010.

Silber score 3-4 vs. 1-2: OR 1.33 (95%CI 0.64-2.76); P overall=0.024.

Sperm granuloma unilateral vs. no granuloma: OR 1.52 (95%CI 0.54-4.27); P overall=0.726.

Sperm granuloma bilateral vs. no granuloma: OR 1.02 (95%CI 0.27-3.85); P overall=0.726.

Vasectomy:

performed under general anaesthesia. In most cases, a multilayer technique was used to achieve VV anastomosis. Microscopic examination of the seminal fluid was performed intraoperatively, placing a drop of fluid on a slide and diluting with normal saline. A modified Silber score was assigned to each patient, ranging from grade 1-5:

Grade 1 – mostly morphologically normal, motile spermatozoa;

Grade 2 – mostly morphologically normal, nonmotile spermatozoa; Grade 3 – mostly spermatozoa heads with few intact spermatozoa;

Grade 4 – only spermatozoa heads without tails;

Grade 5 – no spermatozoa or spermatozoa fragments.

NB:

- The aim of the study was to examine the influence of female partner age on pregnancy. Analyses were corrected for male partner age, but data was not shown.

- Bilateral VV: 64.6%, Unilateral VV: 3.3%, Bilateral VE: 15.5%, Unilateral VE: 5.2%.

- Study population overlap with Magheli (2010), but this study has a shorter follow-up (though duration of follow-up was not reported for either study).

- N (%): 851/1063 (80.1%) were excluded from study population. Patients without follow-up fertility data were excluded from the study, reasons not reported.

Kim (2005)

Type of study: retrospective observational cohort study

Setting: hospital outpatients, single centre.

Country: South Korea.

Source of funding: not reported.

Inclusion criteria: men undergoing vasectomy reversal in whom one or more prior attempts at vasectomy reversal had failed.

Exclusion criteria: patients for whom postoperative semen analysis had never been done and without pregnancy report. Patients with postoperative semen analysis but without pregnancy follow-up data.

N= 44

Mean age ± SD:

Age at reversal: median 37 (32-56) years

Age of female partner: median 35 (28-49) years

Sex: 100%M

Other important characteristics (confounders and effect modifiers):

Median PVOI: 8.5 (3-23.7) years

Median interval since first vasectomy reversal attempt: 2.7 (0.7-9) years.

Age at vasectomy: median 29. (23-39.3) years.

Partner type: same (63.6%)

Sperm present at operation: 52.3%

Bilateral VV: 90.9% (9.1% others).

Anastomotic levels: straight vas (70.5%), convoluted vas (29.5%)

Patency rate: 86.4%

Describe prognostic factor(s) and method of measurement:

Prognostic factors were patient age, partner age, partner type, PVOI, interval since first vasectomy reversal, age at vasectomy, sperm at operation, reconstruction type, anastomotic levels, and patency.

Method of measurement was not reported. Presumably, data were established from patients records and via patient interviews: ‘Pregnancy and childbirth was established during follow-up visits and by telephone or mail.’

Endpoint of follow-up: mean duration of follow-up 4.1 years (0.7-8.2)

For how many participants were no complete outcome data available?

Not reported.

Reasons for incomplete outcome data described?

Not reported. Presumably only patients with compete data were selected.

Incidence of pregnancy: 56.8% (of which 2 pregnancies spontaneously aborted)

Pregnancies were achieved mean 6.5 months (2-48) post-operative.

Pregnancy:

Pregnancy and childbirth was established during follow-up visits and by telephone or mail.

Results from multivariate logistic regression analysis. The outcome refers to the odds for a pregnancy compared to no pregnancy.

Significant associations:

Partner age <35 vs. ≥35 years OR 4.127 (95%CI 1.058 – 16.097); P=0.041.

Non-significant associations:

Patient age: <40 years vs. ≥40 years OR 2.741 (95%CI 0.617-12.171); P=0.185.

Interval since first vasectomy reversal: <10 vs. ≥10 years OR 0.679 (95%CI 0.169-2.735); P=0.586.

Partner type, PVOI, age at vasectomy, sperm at operation, reconstruction type (bilateral VV vs. others), anastomotic levels (straight vs. convoluted vas), and patency were not studied in multivariate analysis as there were no significant associations in univariate analysis.

Vasectomy reversal:

Intravasal fluid was evaluated for sperm using a touch preparation examined by the surgeon and/or a pathologist. All patients underwent bilateral microsurgical VV with 2-layer end-to-end anastomosis if surgically possible, and this was done regardless of the presence of sperm in the intraoperative vas fluid. Four patients underwent unilateral VV only or unilateral VV with contralateral VE because the vassal length was severely compromised as a result of a previous VV.

NB:

- pregnancy included 2 spontaneous abortions.

- patients who achieved pregnancy with ART were excluded (this was not formulated as an explicit exclusion criterium).

- N (%): 38/82= 46.3% were excluded from study population. Defined in exclusion criteria, but number of exclusions per criterium not reported.

Magheli (2010)

Type of study: retrospective observational cohort study.

Setting: hospital outpatients, single centre.

Country: Germany

Source of funding: not reported.

Inclusion criteria: all men undergoing vasectomy reversal for obstructive azoospermia.

Exclusion criteria: patients whose partners achieved pregnancy by ART.

N=334

Mean age ± SD:

Data shown stratified for obstructive interval

Age at reversal:

<5 years: 39.1 ± 6.9

5-10 years: 40.5 ± 5.6

10-15 years: 42.8 ± 5.8

>15 years: 43.9 ± 8.4

Age of female partner:

<5 years: 31.6 ± 5.5

5-10 years: 31.3 ± 4.9

10-15 years: 32.0 ± 4.7

>15 years: 32.0 ± 5.2

Sex: 100%M

Other important characteristics (confounders and effect modifiers):

PVOI:

<5 years: 29.3%

5-10 years: 33.5%

10-15 years: 23.4%

>15 years: 13.8%

Sperm granuloma:

None: 76.3%

Unilateral: 15.9%

Bilateral: 7.8%

Describe prognostic factor(s) and method of measurement:

Prognostic factors were age, age of the female partner, granuloma (uni/bilateral vs. no granuloma), obstructive interval (5-10, 11-15, >15 years).

It was not specified what data were extracted by phone interview.

Endpoint of follow-up: not reported, duration of follow-up not reported.

For how many participants were no complete outcome data available?

Not reported.

Reasons for incomplete outcome data described?

Not reported. Presumably only patients with compete data were selected.

Incidence of pregnancy: 63.2%

When stratified: 63.3% bilateral VV; 58.3% bilateral VE.

Pregnancy:

Pregnancy was verified by biochemical testing and verification using ultrasound.

Results from multivariable cox regression analysis refer to the hazard ratio to obtain a pregnancy compared to no pregnancy during follow-up.

Significant associations:

Age of the female partner: HR 0.94 (95%CI 0.90-0.99); P=0.015.

Non-significant associations:

Age of the men: HR 1.00 (95%CI 0.97-1.04); P=0.907.

Granuloma unilateral vs. no granuloma: HR 1.29 (95%CI 0.67-2.47); P=0.203.

Granuloma bilateral vs. no granuloma: HR 0.52 (95%CI 0.22-1.21); P=0.203.

PVOI:

5-10 vs. < 5 years: HR 1.26 (95%CI 0.69-2.28); P overall=0.213.

11-15 vs. <5 years: HR 0.69 (95%CI 0.37-1.31); P overall=0.213

>15 vs. <5 years: HR 0.71 (95%CI 0.34-1.51); P overall=0.213.

Vasectomy reversal:

performed under general anaesthesia. In most cases, a multilayer technique was used to achieve VV anastomosis. Microscopic examination of the seminal fluid was performed intraoperatively, placing a drop of fluid on a slide and diluting with normal saline.

NB:

- patient population concerns the same population as Hinz (2008), but with longer follow-up. Duration of follow-up not reported. Silber score was not reported in this study, while Hinz did study this as a covariate.

- exact method of analysis not reported (presumably cox regression as HR were reported).

- N (%): 729/1063= 68.6% were excluded from study population. Patients without follow-up fertility data were excluded from the study, reasons not reported.

Peng (2017)

Type of study: retrospective observational cohort study.

Setting: hospital outpatients, single centre.

Country: China

Source of funding: ‘no external funding was received’.

Inclusion criteria: consecutive male patients with obstructive azoospermia undergoing microsurgical VE.

Exclusion criteria: none reported.

N=198

Mean age ± SD:

31.0 ± 5.8 years (range 20-51)

Age of female partner:

28.4 ± 4.4 years (range 18-42)

Sex: 100%M

Other important characteristics (confounders and effect modifiers):

Median infertile interval:

3 years (range 0.5-18)

Aetiology:

Epididymitis (30%)

Not reported (70%)

Describe prognostic factor(s) and method of measurement:

Prognostic factors were male age (<35 vs. ≥35 years), female age (<35 vs. ≥35 years), infertility interval (<2 vs. ≥2 years), FSH, testosterone, left and right testis volume, side of anastomosis, epididymal sperm motility, sperm concentration, sperm forward motility.

Method of measurement was not reported.

Endpoint of follow-up: not reported, median duration of follow-up was 25.3 months (range 12-48).

For how many participants were no complete outcome data available?

0%; all participants had complete data.

Reasons for incomplete outcome data described?

N/A

Incidence of pregnancy: 40%

Incidence of live birth: 34.8%

Pregnancy:

Natural pregnancy was verified by ultrasound.

Results from multivariable cox regression analysis refer to the hazard ratio to obtain a pregnancy compared to no pregnancy during follow-up.

Significant associations:

Male age (<35 years vs. ≥35 years): HR 0.407 (95%CI 0.203-0.816); P=0.011.

Sperm concentration (<20 vs. ≥20 million/ml): HR 4.988 (95%CI 2.777-8.957); P<0.001

Sperm forward motility (<25% vs. ≥25%): HR 1.751 (1.042-2.945); P=0.035.

Non-significant associations:

Data not shown, but P>0.05 for female age (<35 vs. ≥35 years), infertility interval (<2 vs. ≥2 years), FSH, testosterone, left and right testis volume, side of anastomosis, epididymal sperm motility.

NB:

- Study only included patients with natural pregnancy.

- The causes of unilateral reconstruction were contralateral intratesticular obstruction (n = 26), distal vas defense obstruction (n = 14), cryptorchidism (n = 5) and contralateral absent vas defense (n = 6). However, because most patients had no such causes, we explored both sides. If no sperm was found in the epididymis, or distal obstruction in the vas deferens was confirmed, we would abandon reconstruction in this side.

Smrkolj (2010)

Type of study: retrospective cohort study.

Setting: single centre hospital outpatients.

Country: Slovenia.

Source of funding: no financial or commercial conflict of interest reported.

Inclusion criteria: men with obstructive azoospermia with normal serum FSH, testicular volume >30ml and spermatozoa present in the testicular biopsy according to the WHO criteria. All men either underwent vaso-epididymostomy (N=72) or ICISI-TESE (N=34).

Exclusion criteria: none reported.

N total at baseline: 106

ICSI-TESE: 72

VE: 34

Important prognostic factors²:

Age ± SD:

ICSI: 37.7 ± 7.9

VE: 34.5 ± 5.8

Age of female partner ± SD:

ICSI: 32.5 ± 4.7

VE 30.3 ± 5.4

Sex:

ICSI: 100% M

VE 100% M

Congenital absence of vas deferens (%):

ICSI: 19.4%

VE: 0%

Groups comparable at baseline? Yes

Describe prognostic factor(s) and method of measurement:

Prognostic factors were male age, female age, duration of infertility, tobacco intake, testicular histology, inflammation markers in the ejaculate before anastomosis (reduced semen volume and leukocytospermia) and sperm count and motility during epididymal exploration on ductal system patency.

Methods of measurement were not reported.

Endpoint of follow-up: not reported. Follow-up occurred at 3 months post-operative and every 3 months thereafter. Mean duration of follow-up was not reported.

For how many participants were no complete outcome data available?

Not reported.

Reasons for incomplete outcome data described?

Not reported.

Incidence of pregnancy:

Unknown definition

ICSI-TESE: 25%

VE: 38.2%

P>0.05

% pregnancies/cycle:

ICIS-TESE: 25%

VE: 38.2%

P>0.05

%pregnancies/embryo transfer:

ICSI-TESE: 27.3%

VE: N.A.

Pregnancy:

Not defined in text. It presumably refers to all pregnancies, including miscarriages.

Results from multivariate logistic regression analysis. The outcome refers to the odds for a pregnancy compared to no pregnancy.

Significant factors:

Female age (<29 vs. ≥29 years): beta: 2.946; OR 19.034 (95%CI 1.733-209.061); P=0.016.

Testicular histology (normal spermatogenesis vs. mixed lesions): beta: 2.984; OR 19.772 (95%CI 1.192 – 328.085); P0.037.

Non-significant factors:

Male age, duration of infertility, tobacco intake, inflammation markers in the ejaculate before anastomosis (reduced semen volume and leukocytospermia) and sperm count and motility were not significantly associated. Data not shown.

Incidence of live birth:

Defined as % of pregnancies

ICSI-TESE: 77.8%

VE: 84.6%
P>0.05

NB. This is different from the of incidence of live birth / entire population.

Notes:

 - This study also reports on 404 men with non-obstructive azoospermia, which is not reported here.

- Duration of infertility was not shown.

Study reference

(first author, year of publication)

Was there a representative and well-defined sample of patients at a similar point in the course of the disease?

(yes/no/unclear)

Was follow-up sufficiently long and complete?

(yes/no/unclear)

Was the outcome of interest defined and adequately measured?

(yes/no/unclear)

Was the prognostic factor of interest defined and adequately measured?

(yes/no/unclear)

Was loss to follow-up / incomplete outcome data described and acceptable?

(yes/no/unclear)

Was there statistical adjustment for all important prognostic factors?

 (yes/no/unclear)

Davis, 2017

Yes, all men had obstructive azoospermia and were scheduled for vasovasostomy.

Unclear, duration of follow-up was not reported.

Yes, pregnancy was determined by patient inquiry.

Unclear, patient characteristics were presumably extracted from medical records, but this was not explicitly reported.

Unclear, presumably only patients with complete data were selected.

NB: 20.7% were not included in the study due to various reasons (patients did not undergo vasovasostomy due to cost (N=3), change of partner (N=3) and change of mind (N=4). In addition, outcome data were missing in 14/106 patients due to relocation (N=7), refused study inclusion (N=7).)

Unclear, some reported patient characteristics were not included in multivariate analysis. Presumably, these factors were not significantly associated with the outcome and therefore left out, but this is not reported by the authors.

van Dongen, 2012

Yes, all men had obstructive azoospermia and were scheduled for vasovasostomy and/or vasoepididymostomy.

Yes, follow-up was at least 1 year. Mean duration of follow-up was 3.4 years (SD unknown), median was 2.9 years (range unknown).

Yes, pregnancy was determined via questionnaire by telephone.

Yes, prognostic factors were defined in text and method of measurement reported.

Unclear, presumably only patients with complete data were selected.

NB: 12.9% of patients was excluded because of lack of outcome data (did not attempt pregnancy N=9, could not be contacted N=15).

Yes, all variables of interest were analysed in a multivariate cox regression model.

Friedler, 2002

No, sample included both men with obstructive and non-obstructive azoospermia with different histological origins of azoospermia.

Unclear, duration of follow-up was not reported.

Unclear, definition of pregnancy was stratified for all pregnancies and ongoing /delivered pregnancies (excluding abortions). Unclear whether the analyses were performed with what outcome. Method of determination was clear: pregnancy was determined by sonographic demonstration of a gestational sac.

Unclear, method of measurement was not reported. Presumably, data were extracted from patient files. Definitions of prognostic factors were clear.

Unclear, presumably only patients with complete data were selected.

NB: not reported how many people were excluded based on inclusion criteria.

Yes, all variables of interest were analysed in a multivariate logistic regression model.

Gerrard, 2012

Yes, all men had obstructive azoospermia and were scheduled for vasovasostomy.

Yes, minimum of 12 months until pregnancy was reported. Mean duration of follow-up was 32.5 months ± 27.0

Yes, pregnancy was determined via ‘[…] review of the medical records, phone contact, and letters from patients.’

Unclear, prognostic factor ‘female factors’ and ‘repeat procedures’ were not clearly defined. Data were extracted by ‘[…] review of the medical records, phone contact, and letters from patients.’

No, 29.3% did not have outcome data. Reasons not reported.

Unclear, multivariate analysis were not shown in Tables, only in text without effect sizes. In addition, not all factors from Table 1 that were presumably included in multivariate analysis were mentioned as non-significant factors.

Hinz, 2008

No, ethiology of obstruction differed. Majority of participants were men with obstructive azoospermia due to previous vasectomy, minority due to infectious or unknown cause.

Unclear, duration of follow-up was not reported.

Unclear, method of measurement and definition was not reported.

Unclear, method of measurement is not reported. Definitions were clear.

Unclear, presumably only patients with complete data were selected.

NB: 80.1% of patients did not have pregnancy follow-up data and were excluded from the analysis. In addition, patients who achieved pregnancy by ART were excluded, but data was not shown.

Unclear, analysis were only corrected for male partner age (data not shown).

Kim, 2005

No, it concerned men undergoing vasectomy reversal in whom one or more prior attempts at vasectomy reversal had failed. It was not defined how often prior attempts had failed.

Yes, follow-up occurred for a mean duration of 4.1 years. Pregnancy was achieved within a mean 6.5 months (2-48), which suggests that follow-up was long enough to include most pregnancies.

Unclear, it was reported that pregnancy included 2 spontaneous abortions but it was not explicitly stated whether these were included in the analysis or whether sensitivity analysis were performed to test for differences. Method of measurement was sufficient (patient inquiry by phone or mail or during follow-up visits).

Unclear, all factors were clearly defined but method of measurement was not explicitly stated. Presumably, data were established from patients records and via patient interviews: ‘Pregnancy and childbirth was established during follow-up visits and by telephone or mail.’

Unclear, presumably only patients with complete data were selected.

NB: 46.3% of patients were excluded from analysis because of missing pregnancy data. In addition, patients who achieved pregnancy by ART were excluded, but data was not shown.

Yes, all variables in multivariate analysis were selected based on univariate significance. All analyses were shown.

Magheli, 2010

Yes, all men had obstructive azoospermia and were scheduled for vasovasostomy and/or vasoepididymostomy

Unclear, duration of follow-up was not reported.

Yes, pregnancy was verified by biochemical testing and verification using ultrasound.

Unclear, method of measurement is not explicitly reported. Presumably data were extracted from patients files and during phone interviews. Definitions were clear.

Unclear, presumably only patients with complete data were selected.

NB: 68.6% of patients did not have pregnancy follow-up data and were excluded from the analysis. In addition, patients who achieved pregnancy by ART were excluded, but data was not shown.

Yes, all variables were included in analysis. However, in a previous publication (Hinz, 2008) Silber score was included as a covariate, which was not included here without reason.

Peng (2017)

Likely, men with obstructive azoospermia due to (suspected) miscellaneous reasons of epididymal obstruction. It was reported that in 30% obstruction was due to epididymitis; other reasons unknown.

Yes, median duration of follow-up was 25.3 months (range 12-48).

Yes, pregnancy was confirmed by ultrasound.

Unclear, method of measurement was not reported.

Yes, all patients had complete outcome data. Patients without pregnancy data were excluded from the study.

Yes, multivariate analysis corrected for most important prognostic factors.

Smrkolj (2010)

Likely, men with obstructive azoospermia with normal spermatogenesis presenting with infertility. Only from a few it is reported that origin of infertility is congenital absence of vas deferens or inflammation.

Unclear, mean duration of follow-up was not reported.

Unclear, it was not specifically reported whether pregnancy concerned all ongoing pregnancies. Presumably this included % miscarriages as well.

Unclear, method of measurement is not explicitly reported. In addition, it was unclear whether tobacco intake referred to male patient or female partner.

Unclear, no data reported.

Yes, multivariate analysis corrected for most important prognostic factors, such as age, smoking, obstructive interval etc.