Long-term risk of ectopic pvaries bymethod of tubal sta whole-population study

.D. Jam

oo meand , Cl

e of 28 years than in those sterilizedopic partial salpingectomy (adjusted

ter steriliza-follow-up.n hamperednd reliancerge studiestive risk of

9,ofpedic).tou-P,

ORIGINAL ARTICLE: CONTRACEPTIONEP in women previously sterilized (10). The Collaborative ReviewSterilization (CREST) Working Groufollowed 10,685 women in the UnitStates who underwent laparoscoptubal sterilization from 19781986 (9The cohort was surveyed annuallyrecord new cases of EP (9). A 10-year cmulative probability of 7.3/1,000 E

Received October 16, 2013; revised November 19, 2013; accepted November 27, 2013; publishedonline January 2, 2014.

E.M. has nothing to disclose. A.K. has nothing to disclose. R.H. reports personal fees from MSD andMerck-Serono. K.J.-A. has nothing to disclose. D.B.P. has nothing to disclose.

Supported by the National Health andMedical Research Council Capacity Building grant (APP573122),Canberra, Australia.

Reprint requests: Eva Malacova, Ph.D., School of Population Health, The University of WesternAustralia, 35 Stirling Highway, Crawley, Perth Western Australia 6009, Australia (E-mail: eva.malacova@uwa.edu.au).S (EP) as a complication of tubalsterilization, and even bilateral andrecurrent cases have been found (16).Of concern is the maternal risk fromEP, which accounts for 5% ofmaternal deaths in developed

and the seriousness of EP (7), there ispressing need to provide accuratelong-term risk estimates of EP inwomen undergoing these proceduresand characterize women at increasedrisk.

often reportedmanyyearsaftion, necessitating long-termResearch in this area has beeby insufcient sample sizes aon self-report. Only two lahave examined the cumulahazard ratio 14.57, 95% condence interval 3.5060.60) and electrodestruction (adjusted hazard ratio 5.65, 95% condenceinterval 2.3813.40), compared with those who had laparoscopic unspecied destruction of fallopian tubes.Conclusion(s): Women undergoing tubal sterilization at a young age are at particular risk forsubsequent EP. The risk among younger women doubled between 5 and 15 years after steriliza-tion. Laparoscopic electrodestruction and partial salpingectomy carried the highest risk of EP.(Fertil Steril 2014;101:72834. 2014 by American Society for Reproductive Medicine.)Key Words: Cohort studies, tubal sterilization, ectopic pregnancy, life tables, age

Discuss: You can discuss this article with its authors and with other ASRM members at http://fertstertforum.com/malacovae-ectopic-pregnancy-tubal-sterilization/

Use your smartphoneto scan this QR codeand connect to thediscussion forum forthis article now.*

* Download a free QR code scanner by searching for QRscanner in your smartphones app store or app marketplace.

everal case studies and serieshave reported ectopic pregnancy

countries (7). Given the widespreaduse of this contraceptive method (8)

Studies of the EP risk after tubalsterilization are challenging. Cases are(n 44,829).Intervention(s): Data on tubal sterilization were extracted from hospital records.Main Outcome Measure(s): Long-term risk of EP.Result(s): There were 89 EPs recorded during the observation period in women previouslycumulative probability of EP for all methods of tubal sterilization were 2.4/1,000 and 2.910-year cumulative probability of EP was 3.5 times higher in women sterilized before the agafter the age of 33 years. An increased risk of EP existed in women who received laparoscSetting: Hospitals in Western Australia.Patient(s): All women aged 1844 years undergoing tubal sterilization between 1990 and 2010 at Western Australian hospitals

sterilized. The 10-year and 15-year/1,000 procedures, respectively. TheDesign: Population-based retrospective cohort study.

Objective: To evaluate the risk of ectopic pregnancy (EP) associated with different methods of tubal sterilization.Eva Malacova, Ph.D.,a Anna Kemp, Phand David Brian Preen, Ph.D.a

a Centre for Health Services Research, SchUniversity of Western Australia, Crawley;AustraliaFertility and Sterility Vol. 101, No. 3, March 2014 0Copyright 2014 American Society for Reproductivehttp://dx.doi.org/10.1016/j.fertnstert.2013.11.127

728,a Roger Hart, M.D., C.R.E.I.,b,c Khadra

l of Population Health and b School of Woc Fertility Specialists of Western Australia015-0282/$36.00Medicine, Published by Elsevier Inc.regnancyerilization:

a-Alol, M.P.H.,a

ns and Infants' Health, Thearemont, Western Australia,and an elevated relative risk for bipolarand unipolar coagulation, interval

VOL. 101 NO. 3 / MARCH 2014

MATERIALS AND METHODS

unspecied destruction of fallopian tubes, and occlusion us-The study included all women aged 1844 years who wereadmitted to hospital in Western Australia (WA) for rst-time tubal sterilization from January 1, 1990, through June30, 2010. Women were included if they had a primary or sec-ondary hospital procedure for tubal sterilization with any ofthe following International Classication Diseases (ICD)codes: ICD-9 (56.6256.65, 59.80) (13), ICD-9-CM (66.266.29, 66.366.39, 66.6, 66.6366.64) (14), or ICD-10-AM(35638-08, 35638-10, 35688-0035688-04, 35717-0235717-03) (15). We restricted the analysis to women whohad undergone tubal sterilization of both fallopian tubes bythe same method, and who had not concurrently undergoneremoval of tubal pregnancy. Women with a history of EPwere excluded.We used a 10-year look-back period to capturenon-rst-time tubal sterilization procedures.

We used linked whole-population administrative healthdata provided by the WA Department of Health. Migrationto and fromWA has been shown to be lower than in any otherAustralian jurisdiction, largely due to its geographic separa-tion from other major Australian cities (16). Data were ex-tracted from all statutorily collected midwives notications,hospital separation records, and death registrations from19802010. Midwives notications contained informationfor all registered births in WA, which was used to identifypregnancies preceding tubal sterilization. Hospital morbiditydata included diagnostic and procedural records for all hospi-tal separations (public and private) in WA. This informationpartial salpingectomy, spring-clip and silicone rubber bandapplication, compared with postpartum partial salpingectomywas reported. However, results were prone to loss-to-follow-up. In addition, surgical sterilization procedures have changedmarkedly since the 1980s (due to a Filshie clip (Femcare-Ni-komed) introduction and lower invasiveness of procedures),thereby impacting the relevance of nding to current practice.Another study of 6,639 women who underwent sterilizationby Falope rings (Gyrus) by laparoscopy or Pomeroy techniqueby laparotomy within a single center in India (from 19861997) reported 0.6 EP per 1,000 procedures, all of which usedFalope rings and occurred within the rst 12 months after ster-ilization (10). This study did not distinguish between intervaland postpartum sterilizations, although the latter accountedfor about one-third of all procedures used in that center.

Postpartum sterilizations, frequently used in the UnitedStates, are rare in Europe or Australasia (11). In contrast,Filshie clip (also known as titanium clip) has become the ster-ilization of choice in many developed countries during thepast three decades but is yet to be evaluated by anypopulation-based studies (12). Furthermore, no study hasevaluated EP risk more than 10 years after sterilization. Weused whole-population administrative health data to estimatethe cumulative probability and risk of EP in a cohort ofwomen undergoing tubal sterilization, with up to 20 yearsof follow-up, and considering the impact of sterilizationmethod and age.was used to select the cohort, EP, and pelvic inammatorydisorders (PID), a strong predictor of EP (17), as previously

VOL. 101 NO. 3 / MARCH 2014ing titanium clip), ve minilaparotomy approaches (partialsalpingectomy, other unspecied destruction or occlusion offallopian tubes, electrodestruction, salpingectomy, and anopen abdominal approach using titanium clip), and a hyster-oscopic approach with the Essure device (transcervical steril-ization; Conceptus).

The primary outcome was hospitalization for rst-timeEP after incident tubal sterilization. An EP was dened by ahospital admission with a primary or secondary diagnosis:ICD-9 (57.43) (13), ICD-9-CM (633.0633.91) (14), or ICD-10-AM (O00.0O00.9) (15, 20). Women were considered tobe at risk for EP until they had a repeat sterilizationprocedure, a surgical reversal (tubal anastomosis),hysterectomy, undergone IVF, died, or reached age 45 years,whichever came rst. Follow-up for all women still at riskwas stopped (censored) on June 30, 2010.

Descriptive statistics were performed to summarize char-acteristics of the study cohort at the baseline. Kaplan-Meiercurves and cumulative life-table probabilities of EP werecalculated to assess time to EP overall and by age group(1827, 2833, 3444 years), as previously categorized bythe CREST study (21). We used Cox regression to calculatehazard ratios (HR) and 95% condence intervals (CI) to inves-tigate the association between type of surgical sterilizationand EP. Additional prognostic factors for EP, including age,Indigenous status, a history of PID, and parity, were also eval-uated. Data were analyzed using SAS version 9.3 (22).

The study was approved by the Human Research EthicsCommittees of the Department of Health WA and the Univer-sity of Western Australia.

RESULTSBetween 1990 and 2010, 45,332 women aged 1844 yearswere identied as having undergone incident tubal steriliza-tion in a WA hospital. Of these patients, we excluded 348women who had concurrent records for sterilization and EPon the same admission, 80 with a history of EP before theirrst sterilization procedure, and 75 with

rom

LaMLaLaLaMLaMLaMHTABLE 1

Baseline characteristics of women who underwent tubal sterilization f

Characteristic

Method of tubal sterilizationUnspecied destruction or occlusion of fallopian tubesOther unspecied destruction or occlusion of fallopian tubesOcclusion of fallopian tubes with Filshie clipOpen abdominal approach with Filshie clipPartial salpingectomyPartial salpingectomyElectrodestruction of fallopian tubesElectrodestruction of fallopian tubesSalpingectomySalpingectomyEssure procedure

Age at sterilization, y182728333444

Indigenous statusNon-IndigenousIndigenous

Parity01

ORIGINAL ARTICLE: CONTRACEPTION(SD 5 years) and 92% were non-Indigenous. Most women(88%) undergoing tubal sterilization had previously beenpregnant and 63% had experienced three or more pregnanciesbefore sterilization. Nearly 5% had a previous hospital recordfor PID, and 6% had previously miscarried.

A total of 174 women (0.4%) experienced an incident EPafter sterilization. Of those, 85 women were censored (78 un-derwent reversal of tubal sterilization, 5 repeat tubal steriliza-tion, 1 hysterectomy, and 1 IVF treatment before EP) and weretherefore not counted in the nal numbers of EP. For the re-maining 89 patients, time from tubal sterilization to EPranged from 47 days to 18 years (mean, 3.17 years), thusrendering luteal phase pregnancy unlikely.

Figure 1A shows the unadjusted Kaplan-Meier curve fortime to EP from incident tubal sterilization for all methodscombined. At 5 years after surgery, the cumulative probabilityof EP was 1.7/1,000 procedures, increasing to 2.4/1,000 pro-cedures by 10 years and 3.0/1,000 procedures by 15 years.

The cumulative probability of EP differed according toage at sterilization. Kaplan-Meier curves for time to EP forwomen aged 1827, 2833, and 3444 years are shown inFigure 1B. All three curves increased over time, with someoverlap between the youngest and the 28- to 33-year agegroups in the rst 4 years. In the youngest age group, the

2R3

History of pelvic inammatory disordersNoYes

Index of socioeconomic disadvantagea

LowestLowHighHighestMissing

Note: Data in parentheses are percent.a Highest index of disadvantage represents the lowest level of socioeconomic status.

Malacova. Ectopic pregnancy following tubal sterilization. Fertil Steril 2014.

7301990 through 2010.

Women with tubal sterilization (n[ 44,829)

Techniqueparoscopy 22,295 (49.7)inilaparotomy 4,904 (10.9)paroscopy 11,858 (26.5)parotomy 2,859 (6.4)paroscopy 108 (0.2)inilaparotomy 1,166 (2.6)paroscopy 789 (1.8)inilaparotomy 19 (0.04)paroscopy 195 (0.4)inilaparotomy 358 (0.8)ysteroscopy 278 (0.6)

5,726 (12.8)14,968 (33.4)24,135 (53.8)

41,473 (92.5)3,356 (7.5)

5,269 (11.8)1,608 (3.6)5-year cumulative probability of EP was 2.5/1,000 proce-dures, increasing to 4.5/1,000 procedures by 10 years and5.9/1,000 procedures by 15 years. In contrast, for womenaged 3444 years, the 5-year cumulative probability was1.1/1,000 procedures, increasing marginally to 1.3/1,000 pro-cedures by 10 years.

Cumulative probability varied according to tubal sterili-zation method (Table 2). Laparoscopic partial salpingectomyhad the highest 10-year probability of EP (21.5/1,000 proce-dures) followed by laparoscopic electrodestruction (8.4/1,000). Electrodestruction by minilaparotomy together withtranscervical sterilization and salpingectomy (laparoscopyand minilaparotomy) had the lowest probability (0/1,000).Of the seven sterilization methods where EP was recorded,partial salpingectomy (laparoscopy and minilaparotomy)and laparoscopic electrodestruction had a larger risk in therst 5 years of follow-up, whereas the risk for the otherfour methods increased consistently over time.

The associations between method of tubal sterilization,other prognostic factors, and the risk of EP are shown inTable 3. Women who laparoscopically underwent partialsalpingectomy (HR 14.40; 95% CI 3.4759.79), electrodes-truction (HR 5.38; 95% CI 2.2712.77), and, to a lesserextent, other unspecied destruction or occlusion via

9,794 (21.8)28,158 (62.8)

42,656 (95.2)2,173 (4.8)

7,794 (17.3)7,783 (17.3)7,771 (17.3)7,768 (17.3)

13,713 (30.6)

VOL. 101 NO. 3 / MARCH 2014

FIGURE 1

Kaplan-Meier survival curve showing the cumulative probability of ectopic pregnancy (EP) in women who underwent tubal sterilization. (A) Overall.(B) According to age group (1827 years, green square; 2833 years, blue circle; 3444 years, red diamond).Malacova. Ectopic pregnancy following tubal sterilization. Fertil Steril 2014.

Fertility and Sterilityminilaparotomy (HR 1.71; 95% CI 0.963.06) had anincreased risk of EP, compared with laparoscopic unspecieddestruction or occlusion. No EPs were observed for transcer-vical sterilization, salpingectomy (laparoscopy and minila-parotomy), and electrodestruction by minilaparotomyduring follow-up. No association existed between the otherthree methods of sterilization (partial salpingectomy by mini-laparotomy, open abdominal approach, and laparoscopic oc-clusion with titanium clip) and EP.TABLE 2

Long-term cumulative probability of EP, according to tubal sterilization p

Method of tubal sterilization TechniqueNo. ofwomen

No. ofEPs

Unspecied destruction orocclusion of fallopian tubes

Laparoscopy 22,295 40

Other unspecied destruction orocclusion of fallopian tubes

Minilaparotomy 4,904 16

Occlusion of fallopian tubes withFilshie clip

Laparoscopy 11,858 19

Open abdominal approach withFilshie clip

Laparotomy 2,859

TABLE 3

Hazard ratio for EP, according to tubal sterilization method and selected characteristics.

VariableNo. of women(n[ 44,829)

Hazard ratio (95% CI)unadjusted P valuea

Hazard ratio(95% CI) adjustedb P valueaMethod of tubal sterilization Technique

Unspecied destruction or occlusion of fallopian tubes Laparoscopy 22,295 Ref < .001 Ref < .001Other unspecied destruction or occlusion of fallopian tubes Minilaparotomy 4,904 1.71 (0.963.06) 1.61 (0.902.88)Occlusion of fallopian tubes with Filshie clip Laparoscopy 11,858 1.10 (0.631.91) 1.19 (0.682.06)Open abdominal approach with Filshie clip Laparotomy 2,859 0.97 (0.342.71) 0.92 (0.332.59)Partial salpingectomy Laparoscopy 108 14.40 (3.4759.79) 14.57 (3.5060.60)Partial salpingectomy Minilaparotomy 1,166 1.18 (0.294.90) 1.11 (0.274.62)Electrodestruction of fallopian tubes Laparoscopy 789 5.38 (2.2712.77) 5.65 (2.3813.40)Electrodestruction of fallopian tubes Minilaparotomy 19 0 0Salpingectomy Laparoscopy 195 0 0Salpingectomy Minilaparotomy 358 0 0Essure procedure Hysteroscopy 278 0 0Age at sterilization, y1827 5,726 3.23 (1.835.70) < .001 3.20 (1.815.66) < .0012833 14,968 2.02 (1.213.35) 2.04 (1.233.40)3444 24,135 Ref Ref

History of pelvic inammatory disordersNo 42,656 Ref .179 Yes 2,173 1.77 (0.823.84)

Indigenous statusNon-Indigenous 41,473 Ref .835 Indigenous 3,356 0.92 (0.422.00)

Parity0 5,269 Ref .718 1 1,608 1.66 (0.564.96) 2 9,794 0.95 (0.432.11) R3 28,158 0.98 (0.491.99)

Index of socioeconomic disadvantagec

Lowest 7,794 Ref .463 Low 7,783 0.80 (0.371.77) High 7,771 0.89 (0.411.93) Highest 7,768 1.18 (0.572.44) Missing 13,713 1.38 (0.742.56)

Note: CI condence interval; EP ectopic pregnancy; Ref reference group.a P values are based on the c2 test for trends.b Hazard ratios were derived from multivariable Cox regression and adjusted for method of sterilization and age, signicant at the 5% level.c Highest index of disadvantage represents the lowest level of socioeconomic status.

Malacova. Ectopic pregnancy following tubal sterilization. Fertil Steril 2014.

732

VOL.101

NO.3

/MARC

H2014

ORIG

INALARTIC

LE:CONTRA

CEPTIO

N

rience, as was the case with the high failure rates of siliconerubber band and spring clip found by previous researchers

(23). We did not have information about the operating sur-geons to investigate this issue, nor about the clinical decisionmaking.

Although case studies and series in this area exist, it isdifcult to compare the current ndings due to a paucity ofpopulation-based studies evaluating long-term outcomes inthis area. To our knowledge, only one large study (CREST)had the opportunity to include a large cohort with long-term (10-year) follow-up (9), and no study has included theentire population of women undergoing tubal sterilization.The only other large studies published were based on singletreatment centers in India and Norway, and had limitedfollow-up (10, 24). None of these three studies examinedpartial salpingectomy (adjusted HR 14.57; 95% CI 3.5060.60) or electrodestruction (adjusted HR 5.65; 95% CI2.3813.40) was associated with an increased risk of EP,compared with laparoscopic unspecied destruction or occlu-sion. Other methods of sterilization (minilaparotomy methodsof partial salpingectomy and other unspecied destruction orocclusion, open abdominal approach, and laparoscopic occlu-sion with titanium clip) showed nonsignicant positivetrends. Women aged 1827 years at sterilization had anincreased risk (adjusted HR 3.20; 95% CI 1.815.66) ofEP, compared with the oldest group. Similar results werealso obtained when socioeconomic status was added to themodel (data not shown).

DISCUSSIONOur results suggest that the risk of EP after tubal sterilizationis greatest in women sterilized when aged less than 28 years.The risk of EP doubled in this age group from 515 years aftersterilization. Overall, laparoscopic procedures resulted in ahigher rate of EP than minilaparotomy. Laparoscopic electro-destruction had a 7-fold and laparoscopic partial salpingec-tomy had a 16-fold increase in the risk of EP, comparedwith laparoscopic unspecied destruction or occlusion of fal-lopian tubes. In contrast, other unspecied destruction orocclusion by minilaparotomy, laparoscopic occlusion with ti-tanium clip, partial salpingectomy by minilaparotomy, andabdominal approach with titanium clip had a negligent andnonsignicant increase in risk. No risk existed for electrodes-truction by minilaparotomy, salpingectomy (laparoscopy andminilaparotomy), and transcervical sterilization.

Some limitations do exist that should be considered. Weadjusted for hospital admission for history of PID, represent-ing severe cases of PID, but could not identify instancestreated in outpatient settings or those left untreated. It is likelythat only the more serious cases were identied, which wouldhave greater implications for the outcomes evaluated. Weevaluated hospital admissions for EP and could not identifywomen treated as outpatients or those who lost their preg-nancy spontaneously. Although the probability of EP variedby sterilization method used, we cannot discount the possibil-ity that some of the variation could be due to surgeons expe-the titanium clip method. Although this method has beenwidely used in Australia and the United Kingdom for the

VOL. 101 NO. 3 / MARCH 2014past three decades (12), the US Food and DrugAdministration did not approve its clinical use until 1996and thus, would not have been available at the time of theCREST study, which examined data from 19781986.Consequently, the generalizability of results from the CRESTstudy to current clinical practice is uncertain. Directcomparisons with the previous cohorts are furthercomplicated by their use of methods different to those usedin Australia. More than half of the US cohort used siliconerubber band (Falope rings), spring clip (also known as Hulkaclip), and postpartum partial salpingectomy (9), whereas theNorwegian cohort used unipolar diathermi, spring clip,silicon rings, and endotherm coagulation (24) and theIndian cohort compared Falope rings with Pomeroystechnique (10). In addition, we were unable to distinguishbetween bipolar and unipolar electrodestruction, thusfurther complicating comparisons.

Interestingly, our more recent and whole-population datasuggest the 10-year cumulative probability of EP to be 67.1%lower than that previously reported by CREST (2.4/1,000 pro-cedures vs. 7.3/1,000 procedures) (9). This difference may beattributed to a much larger proportion of women aged 2534 years at sterilization and a sizeable proportion of non-Hispanic black women, who were found to be at increasedrisk of EP, in the US cohort. Our results thus suggest lowerrisk for women undergoing these procedures at present thanpreviously reported. This may be due to increased safety orbetter surgical standards in more recent times. Alternatively,womenmay be receiving better advice about the risks or somehigh-risk women may no longer be referred for theseprocedures.

It is possible that the signicantly increased risk of EP in2% of women who underwent partial salpingectomy,compared with those who underwent laparoscopic occlusionwith titanium clip, was due to chance. A degree of difcultyof surgery may lead an operator to perform partial salpingec-tomy, and therefore, those women who underwent partial sal-pingectomy might have had other medical conditions thatcould have decreased the effectiveness of sterilization, whichwere unable to be controlled in the present study.

We found, in agreement with the CREST study (9), thatwomen who underwent tubal sterilization before the age of27 years had the highest risk of EP and that the risk decreasedwith increasing age at sterilization. We expect that the strongeffect of age at sterilization on EP is a reection of higherfertility among younger women. Most women in our studyexperienced EP around the age of 3435 years, thus conrm-ing previous ndings indicating that women older than theage of 35 years are more likely to experience EP than youngerwomen (20, 25). Changes in the functionality and structure ofthe fallopian tubes have been proposed as a likely explanationfor this elevated rate of EP among older women (25),especially due to sexually transmitted infections and/or PID(17, 26). Our observation of an almost twofold increase inthe risk of EP after tubal sterilization for women with ahistory of PID supports this explanation (9, 27).

The main strength of our study is that we were able to

Fertility and Sterilitycapture all women in the State of WA (population morethan 2.3 million) (16) who underwent tubal sterilization

733

during a 20-year period. Using data routinely collected understatutory requirement ensured complete capture of tubal ster-ilizations and inpatient EPs during the entire observationperiod. This also reduces loss to follow-up, experienced byprevious studies (9, 10, 23). In addition, we implemented a10-year look-back period to increase the likelihood that alltubal sterilizations in this study represented patients rststerilization procedures. However, it is possible that therewere also women who had their rst procedure before 1980.

In summary, although tubal sterilization is intended to bepermanent and is generally effective, EP cannot be ruled outeven 15 years after sterilization. Young women may be atparticularly high risk of EP after tubal sterilization and their

4202.

11. Chan LM, Westhoff CL. Tubal sterilization trends in the United States. FertilSteril 2010;94:16.

12. Peneld AJ. The Filshie clip for female sterilization: a review of world expe-rience. Am J Obstet Gynecol 2000;182:4859.

13. World Health Organisation. The international statistical classication of dis-eases and related health problems, 9th revision (ICD-9). Geneva: WorldHealth Organisation; 1979.

14. National Coding Centre. The international statistical classication ofdiseases and related health problems, 9th revision, clinical modica-tion (ICD-9-CM). Sydney: National Centre for Classication in Health;1996.

15. National Centre for Classication in Health. The international statistical clas-sication of diseases and related health problems, 10th revision, Australianmodication (ICD-10-AM). Sydney: National Centre for Classication inHealth; 2006.

ORIGINAL ARTICLE: CONTRACEPTION7. Khan KS, Wojdyla D, Say L, Gulmezoglu AM, van Look PFA. World HealthOrganisation analysis of causes of maternal death: a systematic review. Lan-cet 2006;367:106674.

8. Jayaraman S, Mann M. Male and female sterilization. Obstet Gynaecol RepMed 2012;22:8591.

9. Peterson HB, Xia Z, Hughes JM, Wilcox LS, Tylor LR, Trussell J. The risk ofectopic pregnancy after tubal sterilization. N Engl J Med 1997;336:7627.

10. Roy KK, Banerjee N, Takkar D. Pregnancy following tubal sterilization: an 11-year survey. Int J Gynaecol Obstet 2000;68:534.risk may double between 5 and 15 years after sterilization.Electrodestruction and partial salpingectomy by laparoscopycarry the highest risk of EP, but account for a small proportionof overall procedures.

Acknowledgments: The authors thank the staff at the WAData Linkage Branch and data custodians at the WA Depart-ment of Health.

REFERENCES1. Shah JP, Parulekar SV, Hinduja IN. Ectopic pregnancy after tubal sterilization.

J Postgrad Med 1991;37:1720.2. Huang J, Chen S, Lin F, Change K, Ko T. Ectopic pregnancy after tubal ster-

ilization. Am J Emerg Med 1999;17:312.3. Kumar S, Oxorn H. Ectopic pregnancy following tubal sterilization. CMA

1978;119:1567.4. Fischer S, Keirse MJ. When salpingectomy is not salpingectomy-ipsilateral

recurrence of tubal pregnancy. Obstet Gynecol Int 2009; http://dx.doi.org/10.1155/2009/524864.

5. Levy SJ, Rosenzweig BA, Blumenthal L. Bilateral tubal pregnancies after tubalsterilization. Obstet Gynacol 1988;72:494.

6. Chou S, HsuM, Chow P, Chiang H, Su H, Hsu C. Recurrent ipsilateral ectopicpregnancy after partial salpingectomy. Taiwan J Obstet Gynecol 2009;48:73416. Clark A, Preen DB, Ng JQ, Semmens JB, Holman CDAJ. Is Western Australiarepresentative of other Australian States and Territories in terms of key so-cio-demographic and health economic indicators? Aust Health Rev 2010;34:2105.

17. Kamwendo F, Forslin L, Bodin L, Danielsson D. Epidemiology of ectopic preg-nancy during a 28 year period and the role of pelvic inammatory disease.STI 2000;76:2832.

18. Stewart LM, Holman CDAJ, Aboagye-Sarfo P, Finn JC, Preen DB, Hart R.In vitro fertilization, endometriosis, nulliparity and ovarian cancer risk. Gyne-col Oncol 2013;128:2604.

19. Australian Bureau of Statistics. An introduction to Socio-Economic Indexesfor Areas (SEIFA). Canberra: Australian Bureau of Statistics; 2006.

20. Creanga AA, Shapiro-Mendoza CK, Bish CL, Zane S, Berg CJ,Callaghan WM. Trends in ectopic pregnancy mortality in the United States.Obstet Gynecol 2011;117:83744.

21. Peterson HB, Xia Z, Hughes JM, Wilcox LS, Tylor LR, Trussell J. The risk ofpregnancy after tubal sterilization: ndings from the U.S. Collaborative Re-view of Sterilization. Am J Obstet Gynecol 1996;174:11618.

22. SAS Institute Inc. SAS/ETS Users Guide Version 9. Cary, NY: SAS InstituteIncorporated; 2003.

23. Peterson HB, Xia Z, Wilcox LS, Tylor LR, Trussell J, US Collaborative Reviewof Sterilization Working Group. Pregnancy after tubal sterilization withsilicone rubber band and spring clip application. Obstet Gynecol 2001;97:20510.

24. Qvigstad E, Jerve F. Ectopic pregnancy following tubal sterilization. Int JGynaecol Obstet 1982;20:27981.

25. Hoover KW, Tao G, Kent CK. Trends in the diagnosis and treatmentof ectopic pregnancy in the United States. Obstet Gynecol 2010;115:495502.

26. Simms I, Rogers PA, Nicoll A. The inuence of demographic change and cu-mulative risk of pelvic inammatory disease on the incidence of ectopicpregnancy. Epidemiology and Infection 1997;119:4952.

27. Peterson HB, Xia Z, Wilcox LS, Tylor LR, Trussell J. Pregnancy after tubalsterilization with bipolar electrocoagulation. Obstet Gynecol 1999;94:1637.VOL. 101 NO. 3 / MARCH 2014

Long-term risk of ectopic pregnancy varies by method of tubal sterilization: a whole-population studyMaterials and methodsResultsDiscussionAcknowledgmentsReferences