Sterilization is intended to be permanent. Men and women are counseled not to undergo sterilization if there is any possibility that they may want children in the future because it is not always possible to reverse sterilization and reversal itself is costly. It is nevertheless well known that women and men do sometimes change their minds later. Some regret having been sterilized, and a smaller subset actually attempt to have their procedures reversed. 1
In Canada, sterilization is a popular contraceptive option. In the Canadian Fertility Survey of 1984, it was the most popular method of contraception, with 35% of women relying on female sterilization and 13% relying on male sterilization. 2 In the Canadian Contraceptive Study of 1998, these percentages had changed to 9% for female sterilization and 16% for male sterilization. 3 Until 1986, with the exception of Newfoundland, the rate of tubal ligation had always been higher than that of vasectomy in Canadian provinces. 4 By that year, Quebec, which had always showed the highest Canadian rate of vasectomy, experienced another rise in the vasectomy rate concurrently with a substantial drop in the tubal ligation rate. 4 In 1977, 48 per 1000 women and 13 per 1000 men in the 30–34 age group obtained a sterilization. 5 The comparable figures for 1999 were 11 per 1000 (tubal ligation) and 19 per 1000 (vasectomy), with the vasectomy rate peaking in 1985 at 25 per 1000. 5,6 Among Quebec women of the 1950 to 1951 birth cohort, 37% had a tubal ligation before age 50, at a mean age of 32 years. For women born 10 years later, it is estimated that 23% will have a sterilization before 50 years, at a mean age of 33 years. As for men, 24% of those born in 1945 or 1946 had a vasectomy before reaching age 55, at a mean age of 37 years, whereas in younger cohorts born in 1955 or 1956, or 1960 or 1961, it is estimated that 61% will have a vasectomy before age 55, at a mean age of 36 years. 5
In all Canadian provinces, men and women can obtain almost all their reproductive health care at no cost through a governmental health insurance plan. Every Canadian province pays physicians for each medical act performed. Using these payment data, we examined what happens after sterilization in the province of Quebec. Specifically, we examined the likelihood of sterilization reversal and of subsequent sterilization after sterilization reversal among men and women; among women, we also examined the likelihood of pregnancy after sterilization (contraceptive failure) and of pregnancy after sterilization reversal.
MATERIALS AND METHODS
All data required for this study were obtained from la Reégie de l'Assurance-maladie du Québec. In the province of Quebec, la Régie de l'Assurance-maladie du Québec is the organization to whom physicians report every medical procedure they perform to receive their fee. Virtually all physicians (99.5%) participate in this state insurance program (the main exceptions being plastic surgeons and ophthalmologists, whose services are not fully covered), and by law private insurance cannot be offered for services covered by la Régie de l'Assurance-maladie du Québec. La Régie de l'Assurance-maladie du Québec has maintained a computerized data bank of all medical procedures obtained by the population of Quebec since the establishment of the Canadian governmental health insurance plan in 1969. The data for this study were obtained for each person undergoing vasectomy or female sterilization from January 1, 1980 to December 31, 1999 and linked through a unique identifying number for each person. Information was collected for the following procedures: female sterilization, performed interval or postpartum, with or without induced abortion, with or without a curettage, through vaginal or abdominal route; vasectomy; vaginal delivery, cesarean delivery, induced abortion by surgical or induction route, treatment of ectopic pregnancy, curettage for spontaneous abortion; and tubal reanastomosis with or without microscopy, vasovasostomy with or without microscopy, and epididymovasostomy with or without microscopy. Approval for this study was obtained from the Institutional Review Panel at Princeton University.
In the province of Quebec, female sterilizations as well as treatment of ectopic pregnancies are mostly performed by gynecologists and, in remote areas, by general surgeons. Vasectomies are mostly performed by general practitioners and also by urologists. Reanastomosis of deferens vas and tubal reanastomosis are respectively performed by urologists and gynecologists. Deliveries, induced abortions, and treatment of spontaneous abortions can be performed by both gynecologists and general practitioners. Vasectomy, induced abortion, and treatment of spontaneous abortion are usually performed in outpatient clinics of a hospital or in local clinics. All other procedures are performed in hospitals (operating room, delivery room).
Kaplan–Meier survival curves were constructed for the following six analyses with the software Stata 7 (Stata Corp., College Station, TX): female sterilization to pregnancy (contraceptive failure), female sterilization to reversal, reversal of female sterilization to pregnancy, reversal of female sterilization to subsequent sterilization, vasectomy to reversal, and reversal of vasectomy to subsequent sterilization.
We restricted the analysis to women who had sterilizations at ages 15–49 and to men who had vasectomies at ages 18–49. To insure that the observations were independent, the unit of analysis was individuals, not procedures; for example, if a man experienced a vasectomy, a reversal, and a subsequent vasectomy, only the first vasectomy appeared in the analysis of time from vasectomy to reversal (we actually did all analyses both including and excluding these subsequent procedures, and the results were virtually identical). In each of these analyses, we constructed separate survivor curves for three age groups, corresponding approximately to terciles of age at sterilization or reversal (depending on the analysis) for males and females. We tested for differences among survivor curves by age tercile using the log-rank test. Exposure in the analysis was censored when individuals reached age 60; this censoring affected only those in the oldest age group. For the analyses of time from sterilization to pregnancy and from sterilization reversal to pregnancy, we omitted sterilizations and reversals, respectively, that occurred in 1999 because pregnancies may not have been resolved, and consequently not in the payment database, by December 31 of that year. We assumed that pregnancies occurred 9 months before a delivery, 3 months before an induced abortion, and 2 months before a spontaneous abortion or treatment of an ectopic pregnancy. We also examined whether there was any change between the 1980s and the 1990s, comparing survivor curves for each age group in the 2 decades using the log-rank test.
As seen in Table 1, during the 20-year period 1980–1999 there were 321,929 female sterilizations, 2496 pregnancies (contraceptive failures) after 311,960 female sterilizations during 1980–1998, 4678 reversals after female sterilization, 2536 pregnancies after 4369 reversals of female sterilization during 1980–1998, 692 subsequent female sterilizations after reversals, 310,827 vasectomies, 4528 reversals after vasectomy, and 778 subsequent vasectomies after 6694 reversals. For an explanation of the different numbers in these analyses, see Table 1's footnotes. All results that follow are based on Kaplan–Meier survivor curves.
Among women, 0.9% experienced a pregnancy after sterilization, 1.8% obtained a reversal after sterilization, 61% achieved a pregnancy after sterilization reversal and 48% achieved a delivery, and 23% obtained a subsequent sterilization after reversal. Among men, 2.4% obtained a reversal after vasectomy and 18% obtained a subsequent vasectomy after reversal.
Results for each of the six analyses are shown in Figures 1–6, separately for each approximate age tercile. In each of these figures, results by age tercile are monotonic, with the curve for the oldest age group at the bottom, the curve for the youngest age group at the top, and the curve for the middle age group in between those two. In each of these analyses except one, the curves highly statistically differ one from another (P < .01); in Figure 6, the curves for the youngest two age groups are statistically indistinguishable (P = .20).
In the youngest age group of women, 1.5% experienced a pregnancy after sterilization, 4.2% obtained a reversal after sterilization, 73% achieved a pregnancy after sterilization reversal, and 32% obtained a subsequent sterilization after reversal. In the youngest age group of men, 3.9% obtained a reversal after vasectomy and 21% obtained a subsequent vasectomy after reversal. In contrast, in the oldest age group of women, 0.4% experienced a pregnancy after sterilization, 0.2% obtained a reversal after sterilization, 46% achieved a pregnancy after sterilization reversal, and 13% obtained a subsequent sterilization after reversal. In the oldest age group of men, 1.0% obtained a reversal after vasectomy and 13% obtained a subsequent vasectomy after reversal.
Not surprisingly, the cumulative risk of contraceptive failure (Figure 1) and reversal (Figure 2) stabilized earlier for the oldest age group than for the two youngest age groups. In contrast, the cumulative risk of reversal for men (Figure 5) continued to rise for all three age groups.
In Figure 3, we display the time from reversal of female sterilization to pregnancy, regardless of the outcome of the pregnancy. The probability of pregnancy was much higher in the two youngest age groups (73% and 64%) than in the oldest age group (46%); most of the pregnancies after reversal occurred within 2 years of reversal. We reanalyzed the data by limiting pregnancies only to those resulting in a delivery (results not shown). The proportion of women achieving pregnancy fell from 73% to 59% in the youngest age group, from 64% to 51% in the middle age group, and from 46% to 34% in the oldest age group.
We found only one change between the 1980s and the 1990s that was both statistically significant and large enough to be clinically important. The proportion of sterilizations to women and men in the youngest age group fell from 37% to 30% (P < .01, Fisher exact test) and from 45% to 33% (P < 01), respectively. But among both women and men, the cumulative probability of reversal in the 1990s was significantly higher than that in the 1980s.
In Quebec, 0.84% of women experienced a pregnancy (contraceptive failure) within 10 years after sterilization, compared with 1.85% of women in the United States, based on the United States Collaborative Review of Sterilization. 7 Two reasons may explain this difference. First, in the American cohort 15% of sterilizations were performed using the Hulka clip, with a relatively high cumulative probability of failure of 3.7% at 10 years. 7 This technique has only rarely been used in Quebec during the past 2 decades. Instead, the Filshie clip–which appears to be more effective 8 – has been the predominant method of female sterilization (personal communication, L. Verschueren, Director, Laborie Inc., 2002). Second, most participants in the United States Collaborative Review of Sterilization study were enrolled in teaching institutions, and it is possible that the rate of contraceptive failure after tubal ligation in these settings is higher than in the general population. 7 In contrast, all sterilizations in the province of Quebec were included in our analysis, regardless of whether those sterilizations were performed in a teaching hospital. However, just as in the United States, the risk of pregnancy is not confined to the first year or two after sterilization; the cumulative probability of pregnancy increases with time since sterilization, from 0.3% within 1 year, to 0.7% within 5 years, to 0.9% within 15 years.
In Quebec, 1.0% of women and 1.0% of men obtained a reversal within 5 years after sterilization. The corresponding figures for the United States (United States Collaborative Review of Sterilization) are 0.2% of women and 0.4% of men. 1 Differences in age at sterilization cannot account for these differences because age at sterilization is actually slightly higher in Quebec than in the United States. Cost, however, may be a factor because reversals can be obtained at no cost to the individual in Quebec, whereas in the United States the cost of a reversal may not be covered by insurance, or may be only partly reimbursed. Fourteen years after sterilization, the rate of reversal in women of Quebec is 80% higher (1.8%) than at 5 years, whereas the corresponding rate in the United States is more than five times as high (1.1%). 9 Easier access to reversal of sterilization for women in Quebec therefore reflects both a more rapid access and better access in general.
Not surprisingly, given that fecundity of women drops rapidly before menopause, we found that women in the oldest age group at the time of sterilization obtained reversals relatively soon after sterilization or not at all. However, men, regardless of age at vasectomy, continued to seek reversals for many years after the vasectomy was performed. It is well known that return of fecundity after a vasectomy reversal declines with time since vasectomy. In a series of 1247 patients, the reported rates of return of sperm to semen and pregnancy, respectively, were 97% and 76% if the obstructive interval was less than 3 years, 88% and 53% if 3–8 years, 79% and 44% if 9–14 years, and 71% and 30% if 15 years or longer. 10 Success of reversal, particularly when done 15 years or more after vasectomy, also depends on spousal age. 11
Many more women may have requested information about sterilization reversal or experienced some form of regret. In the United States, the cumulative probability of women expressing regret after their husband's vasectomy 5 years poststerilization is 6%, similar to the 7% 5-year cumulative probability of regret among women after tubal sterilization. 1 Fourteen years poststerilization, the cumulative probability of expressing regret rises to 13%. 12 The most consistently documented risk factor for regret after sterilization is young age. 12–14 Our study confirms this finding, showing that for both men and women younger age is associated with a higher rate of reversal. In the youngest age groups of men and women in this study, approximately 4% of men and women obtained a reversal after sterilization. Tubal ligation obtained before age 30 in Quebec is generally the result of an early marriage and a rapid constitution of the family, which are characteristic of women with a low education level and less professional expectations. 15
A review of female sterilization reversal studies found that live-birth rates varied according to occlusion techniques, averaging 41% after reversal of electrocoagulation, 50% after reversal of the Pomeroy procedure, 75% after reversal of rings, and 84% after reversal of clips. 13 In other series, 16–20 the pregnancy rate after reversal of tubal ligation varied widely, from a low of 55% at 5 years 16 to a high of 84% at 12 months. 17 Our results fall within this range. Our results also confirm others' findings that the pregnancy rate after reversal is higher in younger women. 16,17,20
Several caveats are in order when interpreting our results. First, we have no information on the type of sterilization procedure, so we cannot compute separate life tables for each type. Second, our estimates are unambiguously lower bounds because we have no way of censoring those who migrate from Quebec or die; instead they remain in the life table but cannot experience a subsequent event (because they are in reality either no longer in Quebec or dead). Our expectation is that these two biases would have opposite effects on the differences by age groups. Younger persons are more likely to migrate than are older persons, so true age differences are greater than we estimate. In contrast, older persons are more likely to die than younger persons, so true age differences are smaller than we estimate. Third, we have no information on sterilizations performed before 1980. Therefore, we cannot be sure that the first sterilization in a person's record is in fact the first sterilization, and consequently we cannot analyze the time from first sterilization to reversal, from second sterilization to reversal, and so forth. We would expect that each subsequent reversal would be less likely than the prior one (and indeed that is the case in our data for those men who have multiple reversals), so that our results underestimate the likelihood of reversal of a first sterilization. Finally, our estimates of pregnancy after sterilization and sterilization reversal among women are lower bounds for the following reasons:
* Spontaneous abortions may not require a curettage.
* Induced abortion may be obtained out of the province of Quebec or under another code such as diagnostic curettage, intrauterine device excision, or polyp excision.
* Medically induced first-trimester abortions with methotrexate or mifepristone and misoprostol as well as ectopic pregnancies treated medically with methotrexate have no precise coding in the data bank of la Régie de l'Assurance-maladie du Québec; therefore they are not counted. However, these procedures are known to be far less common than surgical approaches and have been available only since the beginning of the 1990s for medical treatment of ectopic pregnancies and since 1998 for medical abortions. Medically induced first-trimester abortions can be obtained in only three regions of Quebec. In the region where most of these procedures are offered, they account for only about 4% of all induced abortions (about 125 medical abortions for 3000 surgical abortions annually).
* Very rarely, deliveries may occur at home. In these cases, women do not see physicians at the time of delivery and consequently these deliveries are not declared to la Reégie de l'Assurance-maladie du Québec.
In conclusion, male and female sterilizations are popular contraceptive options in Quebec. Both men and women experience regret after sterilization, in particular when it is done at a younger age. Because reversals are costly to the health care system and not always successful, the high incidence of reversal experienced in younger cohorts of men and women in Quebec suggests that better counseling about the likelihood of sterilization regret, the difficulty of sterilization reversal, and alternative long-term contraceptive methods is needed.
1. Jamieson DJ, Kaufman SC, Costello C, Hillis SD, Marchbanks PA, Peterson HB. A comparison of women's regret after vasectomy versus tubal sterilization. Obstet Gynecol 2002;99:1073–9.
2. Balakrishnan TR, Krotki K, Lapierre-Adamcyk E. Contraceptive use in Canada, 1984. Fam Plann Perspect 1985; 17:209–15.
3. Fisher WA, Boroditsky R, Bridges ML. The 1998 Canadian Contraceptive Study. Can J Hum Sex 1999;8:161–216.
4. Alderman PM, Gee EM. Sterilization: Canadian choices. CMAJ 1989;140:645–9.
5. Duchesne L. Les stérilisations. Donnees Sociodemographiques Bref 2001;5(3):4–5.
6. Rochon M. Les ligatures de trompes et les vasectomies au Québec: Évolution récente. Cah Quebecois Demogr 1991; 20:157–66.
7. Peterson HB, Xia Z, Hughes JM, Wilcox LS, Tylor LR, Trussell J. The risk of pregnancy after tubal sterilization: Findings from the U.S. Collaborative Review of Sterilization. Am J Obstet Gynecol 1996;174:1161–70.
8. Dominik R, Gates D, Sokal D, Cordero M, Lasso de la Vega J, Ruiz AR, et al. Two randomized controlled trials comparing the Hulka and Filshie clips for tubal sterilization. Contraception 2000;62:169–75.
9. Schmidt JE, Hillis SD, Marchbanks PA, Jeng G, Peterson HB. Requesting information about and obtaining reversal after tubal sterilization: Findings from the U.S. Collaborative Review of Sterilization. Fertil Steril 2000;74:892–8.
10. Belker AM, Thomas AJ Jr, Fuchs EF, Konnak JW, Sharlip ID. Results of 1,469 microsurgical vasectomy reversals by the Vasovasostomy Study Group. J Urol 1991;145:505–11.
11. Fuchs EF, Burt RA. Vasectomy reversal performed 15 years or more after vasectomy: Correlation of pregnancy outcome with partner age and with pregnancy results of in vitro fertilization with intracytoplasmic sperm injection. Fertil Steril 2002;77:516–9.
12. Hillis SD, Marchbanks PA, Tylor LR, Peterson HB. Poststerilization regret: Findings from the United States Collaborative Review of Sterilization. Obstet Gynecol 1999; 93:889–95.
13. Church CA, Geller GS. Voluntary female sterilization: Number one and growing. Popul Rep C 1990;10:1–23.
14. Ehn BE, Liljestrand J. A long-term follow-up of 108 vasectomized men. Good counseling routines are important. Scand J Urol Nephrol 1995;29:477–81.
15. Marcil-Gratton N. Le recours précoce à la ligature des trompes au Québec: Des suites indésirables? Sociol Societes 1987;XIX:83–95.
16. Kim SH, Shin CJ, Kim JG, Moon SY, Lee JY, Chang YS. Microsurgical reversal of tubal sterilization: A report on 1,118 cases. Fertil Steril 1997;68:865–70.
17. Kim JD, Kim KS, Doo JK, Rhyeu CH. A report on 387 cases of microsurgical tubal reversals. Fertil Steril 1997;68:875–80.
18. Yadav R, Reddi R, Bupathy A. Fertility outcome after reversal of sterilization. J Obstet Gynaecol Res 1998;24:393–400.
19. Gillett WR, Martin WL, Romans SE. Reversal of female sterilisation: Outcome of 210 referrals. N Z Med J 1993; 106:173–5.
20. Yoon TK, Sung HR, Kang HG, Cha SH, Lee CN, Cha KY. Laparoscopic tubal anastomosis: Fertility outcome in 202 cases. Fertil Steril 1999;72:1121–6.