The relative risks of sterilization alone and in combination with abortion

ABORTION The relative risks of sterilization alone and in combination with abortion KARIN G. B. EDSTROM 1 In recent years many authors have claimed that the combination of tubal sterilization and induced abortion carries too high a morbidity and that the two operations should preferrably be performed separately. As this standpoint has serious practical consequences for many women undergoing abortion, a review of the literature was undertaken to see whether there are any data supporting it. No direct comparison of the morbidity accompanying the combined procedure and the total morbidity from the two procedures performed with an interval could be found. Indirect evidence suggests that the effect of an interval between the operations-if any-is negligible compared with the influence upon morbidity offactors such as methods ofsterilization and abortion, health status, and age. To gain direct information on the problem, a multicentre study has been designed by the Task Force on Sequelae and Complications of Induced Abortion. The outlines of this study are briefly described. In recent years, medical journals have made men- tion several times of the increased risk connected with combining an abortion operation with tubal sterilization. For instance, Huntingford (1) analysed the mortality connected with induced abortion in England during 1969. He found that 12 out of 17 deaths were related to simultaneous sterilization, the mortality rate for abortion combined with steriliza- tion being 110 per 100 000 operations and that for abortion only 10 per 100 000. The death rate was more closely correlated to simultaneous sterilization than to age. From such statistics the conclusion has sometimes been drawn that the combination of induced abor- tion and tubal sterilization should be avoided (2, 3). Before accepting this conclusion one would like to know: (a) whether mortality and morbidity from the combined procedure of abortion and sterilization is really higher than that of abortion and of sterilization performed separately, with an interval, in the same woman (" interval sterilization "); (b) to what extent hidden factors, such as health status, age, and method of operation affect mortality and morbidity for the combined versus separate operations; and ' Research Physician, WHO Research and Training Cen- tre, Karolinska Hospital, Stockholm, Sweden (present address: Medical Officer, Maternal and Child Health, World Health Organization, 1211 Geneva 27, Switzerland). (c) to what extent the imposing of an interval might influence mortality and morbidity. The purpose of this review is to try to find evidence already existing in the literature that can shed light on these questions, to outline what ques- tions remain to be answered, and to suggest how this could be done. MORTALITY Mortality from interval sterilization is difficult to assess with any accuracy from the literature, since only series with large numbers of cases can be used for this purpose and data from many sources have to be added together. These data usually cover post- partum, post-abortum, and interval sterilization without differentiating between them. In India steril- ization is practised on a large scale, both post partum and with an interval, and extensive data have been published, but the mortality figures are often somewhat vague. For instance, Anklesaria (4) re- ported 5 900 cases of post-abortion abdominal steril- ization with at least 4 deaths, giving a mortality of 68 per 100 000 or more. Presser (5), in a review on voluntary sterilization, added up 10 studies contain- ing more than 500 cases from 6 countries (1964-67) and she found 5 deaths among 20 000 reported cases, giving a mortality rate of 25 per 100 000 operations. Among these was a large but unknown proportion 3333 - 141 - BULL. WORLD HEALTH ORGAN., Vol. 52,1975 KARIN G. B. EDSTROM of post-partum sterilizations. The rate might be too low, as it is founded on literature reports only and not on reporting systems, as are many studies of abortion complications. It corresponds fairly well, however, to the rate estimated by Tietze (6) of 33 deaths per 100 000 operations. This mortality figure, however, should not be directly compared with the calculated mortality rate of 3 per 100 000 for legally induced abortion used instead of contraception, or 3 per 100 000 for oral contraception (7), since the two last-mentioned methods do not ensure permanent contraception. The risk of sterilization needs to be taken only once, and the rates have to be calculated for a certain period of time per individual in order to be compa- rable. The median age for female sterilization in countries such as India, Pakistan, Japan, and USA is around 32 years; for Puerto Rico a substantially lower median age of 26 years has been reported (5). This leaves a fertile period of at least 10 years for each sterilized woman during which she is no longer exposed to the risks of other contraceptive methods. Using Tietze's model to calculate the annual mor- tality for different contraceptive methods, the death rate for women using repeat hospital-induced abor- tion instead of contraception during these 10 years would be at least 30 per 100 000. This rate could possibly be reduced a little, owing to the fact that the women in the age groups concerned have a fecundity below average. On the other hand, the annual death rate of 3 per 100 000 is extracted from eastern European data, which reflect a very high level of training of doctors and the fact that almost all abortions are performed in healthy patients in the first trimester. If the total mortality for New York State during the first two years of legal abortion were used instead (8) the total death rate for 10 years would be 50 per 100 000. In New York, 20% of all abortions were performed after the 12th week. Steril- ization thus does not seem to carry a higher mor- tality than other contraceptive methods. MORBIDITY A general morbidity for sterilizing operations is more difficult to calculate, as the criteria used by the investigators differ so widely that the data from dif- ferent studies cannot be added up. Most studies in the literature give data from mixed material where the complication rates for post-partum, post-abor- tion, and interval sterilization are not separated. There is little material in which the complication rates are separate, and the groups are seldom matched for age and parity or for operating proce- dures. Studies on long-term complications of sterilizing operations are usually concerned mainly with psy- chiatric problems, feelings of regret, and sexual adjustment, and are not within the scope of this article. However, an excellent 'brief review by Thompson & Illsley (9) on social and psychological aspects of sterilization was recently published. Gynaecological complaints after sterilization-main- ly menstrual disturbances, dyspareunia, and dys- menorrhoea-have been studied and discussed, but very seldom in a comparative way with a proper prospective design and with preoperative and post- operative data for the same individual. Lu & Chun (10) in a noncontrolled follow-up study of 1 055 cases found menstrual changes (usually slight) after operation in 51.8% of the patients, but in only 4 women to an extent requiring hysterectomy. Secondary dysmenorrhoea was found in 16.8% and slight abdominal pain in 20.5 %. These postoperative symptoms were probably very slight, since 98.7% of the women were content to have had the operation performed. Whitehouse (11) in another long-term follow-up found postoperative menstrual distur- bances in 44.2% (leading to hysterectomy in 6.2%), and Phatak (12) in 21 %. In all these studies, how- ever, a large fraction of the patients were in the premenopausal period, when changes in the men- strual pattern are to be expected. Methods of sterilization The method of sterilization used (abdominal, vaginal, or endoscopic) is of importance for the risks. Mortality in relation to the sterilization method cannot at present be estimated with any degree of accuracy except for the mortality con- nected with vasectomy, which is apparently nil. For female sterilization there is no reporting system taking into consideration the type of operation performed, and in most studies of more than a few hundred patients there is not a complete account of the approach or technique used. Thus, the discussion must be limited to the morbidity reported. Table 1 gives the complication rates found in 9 reports (10, 13-20) of abdominal tubal ligation, mainly for post-partum patients. The rates vary from 1.5% to 25.8 %, the former figure reflecting not severe complications but only febrile periods. The three last-mentioned investigations include a more thorough follow-up than those indicated earlier. 142 Table 1. Morbidity from female sterilization by the abdominal route Author(s) Reference No. Year of Country Type of Timing b No. of cases Tion ratc publication operatjon caes tiotaltcopli Kroener (13) 1969 USA fimbriectomy post partum 200 1.5 Hofmeister (14) 1970 USA n.s. a mainly post partum 465 2.6 White (15) 1966 USA n.s. interval 361 3.0 Lu & Chun (10) 1967 Hong Kong Pomeroy post partum 1 049 5.1 Hayes & Wolfe (16) 1970 USA n.s. mainly post partum 790 6.1 Akhter (17) 1973 Canada mixed interval 122 7.3 Tietze & Lewit (18) 1972 USA n.s. post abortum (suction) 380 20.3 Mabray et al. (19) 1970 USA n.s. post partum 734 20.7 Black & Sclare (20) 1968 Scotland Pomeroy n.s. 209 25.8 a n.s. = not stated. b Timing of operation in relation to preceding pregnancy. c Percentage of operated patients. Table 2. Morbidity from female sterilization by vaginal route (for explanations, see footnotes for Table 1) Author(s) Reference Year of Country Type of Timing No.ofcases Total com- No. publication Cuty operation plication rate John & Dunster (21) 1972 United Kingdom n.s. interval 128 2.3 Yupze et al. (22) 1972 Canada fimbriectomy mixed 1 890 2.1 Smith & Symmonds (23) 1971 USA fimbriectomy mixed 161 3.1 Tietze & Lewit (18) 1972 USA n.s. post abortum (suction) 131 13.7 Sogolow (24) 1971 USA Pomeroy post abortum (suction) 127 14.0 Laufe & Summerson (25) 1972 USA Pomeroy interval 263 8.2 Laufe & Kreutner (26) 1971 USA hysterectomy post abortum 77 26.0 Roach et al. (27) 1972 USA hysterectomy n.s. 100 34.0 v. Nagell & Roddick (28) 1971 USA hysterectomy mixed 100 22.0 Table 2 lists 9 studies (18, 21-28) on tubal steril- ization by the vaginal route, with equally varying rates and also a more mixed composition as regards timing of the operation. According to these studies, vaginal hysterectomy does not seem to be a very recommendable method if complication rates only are considered. Endoscopic sterilizations are listed in Table 3 (18, 29-36) and they undoubtedly give the lowest complication rates. The very low rates given here could, of course, possibly also be dependent on the fact that this procedure is used mainly for interval sterilization. The most recent developments in sterilization techniques, namely hysteroscopy with electrocauteri- zation of the tubal corners and the intrauterine instillation of chemicals, have not yet been evaluated from this point of view. These techniques, however, may be quite irrelevant to the present discussion, as they are technically unsuitable for application in direct connexion with abortion. A comparison between vaginal operations and abdominal operations is given in Table 4 (15, 17, 18, 37, 38). Some of these studies include very few cases indeed, but the cautious conclusion might be drawn 143ABORTION 144 KARIN G. B. EDSTROM Table 3. Morbidity from female sterilization by endoscopic techniques (for explanations, see footnotes for Table 1) Author(s) Reference Year of Cuty Type of TmnNoofcssTotal corn-No. publication o ntry operation Timing No. of cases plication rate Gutierrez-Najar (29) 1972 Mexico culdoscopy post partum + post abortum 302 0.0 Little (30) 1972 USA culdoscopy interval 181 3.3 Steptoe (31) 1970 United Kingdom laparoscopy interval mainly 1 350 0.3 Liston et al. (32) 1970 United Kingdom laparoscopy interval mainly 760 0.9 Jordan et al. (33) 1971 United Kingdom laparoscopy interval 910 2.7 Whiteley (34) 1972 United Kingdom laparoscopy post partum 100 4.0 Chun et al. (35) 1970 Hong Kong laparoscopy interval 120 5.0 Tietze & Lewit (18) 1972 USA laparoscopy post abortum 266 5.3 Peterson & Behrmann (36) 1971 USA laparoscopy interval mainly 200 18.0 a a This rate includes 10 cases with transient cardiac arrhythmia and 5 cases with failure to produce a pneumoperitoneum (later operated on by vaginal tubal ligation). Table 4. Comparison of morbidity rates for tubal sterilization (TS) by vaginal and abdominal routes, by the same authors. In three of the studies there is also a comparison of interval operation alone and sterilization in combination with other procedures (abortion, appendectomy, delivery, caesarean section) Author(s) Reference Year of Timing ofNo. publication operation interval Akhter (17) 1973 TS + other procedures White (15) 1966 interval J interval McMaster & (37) 1971 J Ansari TS + other procedures Tietze & Lewit (18) 1972 post abortum(suction) interval Edwards & (38) 1973 Hakansson post partum at caesarean Morbidity rate (%) vaginal route abdominal route No. of complica- No. of complica- patients tions (%) patients tions (%) 60 25 20.0 28.0 91 13.3 J 9.6 90 1 11.1 122 43 7.3 13.9 361 3.0 0.0 494 6.7 131 13.7 380 20.3 293 11.5 7.4 850 7.8 21.4 that vaginal tubal sterilization carries a slightly higher risk of complications, at least when interval operations are considered. The complications are mainly minor ones, however, and might be balanced by the usually shorter hospital stay and other advan- tages. When methods are discussed in connexion with the combination of abortion plus sterilization, the method of interrupting the pregnancy must also be considered. Here the prospective and large JPSA study (18) gives a fairly clear picture. Table 5 shows (in healthy patients only) the basal complication rate ABORTION Table 5. The influence of simultaneously performed tubal sterilization (TS) on the complication rate after induced abortion in healthy women. Data extracted from the JPSA study by Tietze & Lewit (18) Procedure a No. of women Complicationrate b Suction without TS 50 352 4.2 with TS 777 14.2 Dilatation and curettage without TS 3 077 6.0 with TS 48 (16.7) Saline without TS 13 946 23.4 with TS 108 25.0 Hysterotomy without TS 96 33.3 with TS 683 33.4 a A high proportion of the sterilization procedures for suction curettage + TS were laparoscopic sterilizations. b Total complication rate per 100 aborted women. for different abortion procedures and the additional risk from tubal sterilization. If hysterotomy is used for abortion, then tubal sterilization does not add further morbidity and, in fact, hysterotomy in itself seems to carry a higher risk than tubal sterilization in itself, as no other combination reaches that level. Patient criteria Age and parity should influence morbidity to some extent. Sterilizing operations are mainly car- ried out in the latter half of the reproductive period and in multiparous women. According to Tietze & Lewit (18) there is a substantial increase in risk also for " abortion only " in healthy women above 30 years of age and in women with 3 or more prior births when they are aborted after 12 weeks' gesta- tion; for early gestation there was no such increase. In the age group 30-39 years, 11.2% of women underwent sterilization at the time of the abortion; in the age group above 40 years, 15% did so, as also 19% of women with 4 or more births. In the whole JPSA material, however, simultaneous sterilization was carried out in only 3.7 %. The patient's health status also influences mor- bidity (and mortality) to a considerable extent, but this parameter is seldom taken into account in the investigations found in the literature. Tietze & Lewit (18), in their JPSA study, found that sterilization was carried out five times as often in patients with preexisting medical complications as in healthy patients. Table 6 shows the effect of these preexisting medical complications and of the addition of sterili- zation on the complication rate. From the preceding discussion one may dare to draw the conclusion that the substantial increase in the complication rate when sterilization is added to abortion in healthy patients (from 4.2 to 25.9 per 100 abortions before the 13th week of pregnancy) is largely due to a difference in procedure. The addition of preexisting medical complications should not imply the addition Table 6. Influence of health status on complication rates from abortion only and from abortion + sterilization (A+S). Data extracted from the JPSA study (18) Abortion onlya A + S PECb A + S and PECand abortion <13 >13 <13 >13 <13 >13 <13 >13 weeks weeks weeks weeks weeks weeks weeks weeks Number of patients 50622 17 675 1 242 1 002 1 767 799 249 232 Major complication rates c 0.4 1.6 6.1 8.2 1.4 4.6 14.9 13.8 Total complication rates C 4.2 20.6 25.9 35.8 12.7 29.9 43.0 56.5 a Abortion only = abortion in healthy women, sterilization not performed. b PEC = pre-existing complications. c Complication rates given as the number of women with complications per 100 abortion operations. 145 KARIN G. B. EDSTROM of any special procedures, but it increases morbidity by 66%. Timing of the operation This is the aspect of greatest interest here, espe- cially to see if it is possible from existing data to establish the risks of the interval operation and of the addition of parturition, Caesarean section, and various abortion procedures. The JPSA study (18) gives an excellent picture of the relative risks of different abortion procedures as such and of the addition of sterilization. Very few studies have been found where a direct comparison between interval and other sterilizations was made, and the data are usually scanty. Turner & Hooper (39) found 0.8% thromboembolic episodes among 743 post-partum sterilizations, compared with 0 among 499 interval sterilizations and 0.2% among 18 000 vaginal deliv- eries. This is consistent with the findings of Mabray et al. (19), who found 20.7% morbidity in a group of 734 post-partum sterilizations and 14% in a matched control group of 100 vaginal deliveries. The data in Table 4 give a similar picture of the different com- binations. In no study, however, is the combination of abortion plus sterilization compared with either interval sterilization or post-partum sterilization. Another aspect of timing is the length of the interval between delivery (or abortion) and steriliza- tion and whether this influences morbidity or not. Traditionally, post-partum sterilization should be performed within 2 days of delivery. Thereafter, the risk of infection is believed to increase, and steriliza- tion is preferably postponed for a few weeks. Mabray et al. (19), however, did not find any significant difference in morbidity for different 12-h intervals up to 144 h after delivery. Recent bacterio- logical and histological studies of tubes from post- partum and interval sterilizations also do not find any correlation between the occurrence of infection or the presence of bacteria and the interval during the first week post partum (40-42). Thus, there is at present no evidence in the literature suggesting that the risk of adding the sterilization procedure at the time of the abortion is greater than that of exposing the patients to two separate procedures. If the decision to perform sterilization is well-founded, this operation would anyway have to be performed later on, and the disadvantages of a two-step procedure must be taken into consideration. One obvious disadvantage (apart from the practical advantage for both the woman and the doctor) is the fact that some women do not return for the sterilizing procedure and are then exposed to the continued medical risks combined with either the use of contraceptive methods or new pregnancies. The proportion of these women is not known, however. Another disadvantage of interpos- ing an interval is the risk of a new pregnancy before the second operation. In four studies concerning endoscopic interval sterilizations (30, 32, 43, 44) 18 patients out of 2 616 were found to have been in an early stage of pregnancy at the time of operation-a rate of 0.7%. After abortion the risk of pregnancy arises earlier than after delivery. Boyd & Holmstrom (45) found signs of ovulation as early as 10 days after abortion, and in their study the mean length of the proliferative phase was 22 days. An interval of 4-6 weeks would then allow for an ovulation before the operation in most of the patients if they are not protected by oral contraceptives. CONCLUSIONS The statement that abortion should preferably not be combined with sterilization has been reviewed and discussed here. In view of the important conse- quences of this concept and the lack of valid proof of its correctness it seems to be important to carry out a study designed to obtain a valid answer to the question. This should be a large-scale prospective case-control study to compare the complication rate associated with simultaneous abortion and steriliza- tion with that associated with the two procedures separated by an interval. The groups should be selected at random and include only healthy women requesting both abortion and sterilization. From the data reviewed above it is also evident that the study should be standardized as far as possible as regards the methods used for abortion and for tubal ligation, the stage of pregnancy at abortion, and the length of the interval. Such a study was outlined by the Task Force on Sequelae and Complications of Induced Abortion at a meeting held in Newcastle-upon-Tyne, England, in September 1973. It will be carried out in various countries and within the WHO Expanded Programme of Research, Development and Re- search Training in Human Reproduction, and it is hoped that, within a couple of years, it will give an answer to the question whether an imposed interval between the two operations really decreases total morbidity and hence mortality. 146 ABORTION 147 RESUME RISQUES RELATIFS DE LA STtRILISATION PRATIQUEE SEULE ET DE LA S#TRILISATION ASSOCIE A L'AVORTEMENT Beaucoup de m6decins affirment que la securite de la femme est plus grande lorsque la ligature des trompes est pratiquee un certain temps apres I'avortement (opera- tion i intervalle) plut6t que simultanement. Or, sur le plan pratique, cette attitude a des consequences desa- gr6ables pour les patientes. Nous avons tent6 dans cet article d'6valuer les donn6es fournies par la litterature sur ce sujet et de voir si cette opinion 6tait r6ellement fond6e. Le taux de mortalite lie a la sterilisation (25 a 30 pour 100 000 op6rations, avec ou sans intervalle) est a peu pres le meme que celui des autres methodes contracep- tives, sur une periode de dix ans (30 pour 100 000 utili- satrices de la contraception orale et 30 a 50 pour 100 000 interruptions de grossesse). Toutefois, il n'est pas pos- sible d'evaluer nettement la mortalite ou la morbidit6 liWes a la sterilisation dissociee ou associee a I'avortement car les etudes publiees sont trop restreintes ou traitent a la fois de sterilisations post-partum, a intervalle et post- avortement. En outre, la definition de la mortalite varie considerablement d'un auteur a l'autre tout comme le taux de surveillance. I1 semble cependant que la methode de sterilisation utilisee influe d'une facon marquee sur le taux des compli- cations. Celui-ci est plus eleve pour la laparotomie que pour l'endoscopie avec electrocoagulation. De meme, dans l'operation combinee, le choix de la methode abortive influe sur les suites postoperatoires. L'evacuation par aspi- ration pratiquee au debut de la grossesse et combinee a la sterilisation par laparoscopie ou culdoscopie est la methode pour laquelle on enregistre le taux de compli- cations le plus faible, tandis que l'hysterotomie ou l'hys- terectomie par laparotornie entratnent le taux le plus 6lev6. Avant 40 ans, I'age n'a pas d'influence importante sur la morbidite postop6ratoire. D'autre part, les demandes de sterilisation en relation avec I'avortement sont quatre fois plus frequentes chez les femmes de plus de 40 ans et chez les multipares que dans la population totale des femmes avortees. L'etat de sant6 des patientes avant I'avortement joue un r6le beaucoup plus important. Les travaux bien connus de Tietze & Lewit (18) font apparaitre que les complications m6dicales pre-existantes chez les femmes qui subissent simultanement l'avorte- ment et la st6rilisation augmentent de 66% le taux de complications par rapport aux femmes en bonne sant6. Pour des raisons pratiques, les femmes qui ont choisi la sterilisation comme methode contraceptive desirent generalement que les deux operations soient faites en meme temps. Si la sterilisation est retardee, le risque de nouvelle grossesse non d6siree s'accroit. Par ailleurs, un certain nombre de femmes ne veulent plus se presenter ni pour une operation ni pour d'autres mesures contra- ceptives. L'examen de la litterature ne permet pas de conclure a l'avantage medical d'une dissociation de I'avortement et de la sterilisation; des preuves plus directes restent necessaires. Le Groupe d'action sur l'6va- luation des sequelles et des complications de l'avortement provoque pr6pare actuellement le plan d'une etude multi- disciplinaire qui foumira ces informations. Le present article offre une description et un examen du plan de cette etude. REFERENCES 1. HUNTNGFORD, P. Mortality and morbidity of abor- tion. Lancet, 1: 1012-1013 (1971). 2. KESTELMAN, P. Mortality and morbidity of abortion. Lancet, 2: 368-369 (1971). 3. GOLDSTEIN, P. & STEwART G. Trends in therapeutic abortion in San Francisco. Americanjournal ofpublic health, 62: 695-699 (1972). 4. ANKLESARIA, S. B. In: Purandare, B. N. & Jahveri, C. L., ed. Proceedings ofthe International Seminar on Maternal Mortality, Family Planning and Biology of Reproduction. Bombay, 1970, pp. 181-194. 5. PRESSER, H. B. Voluntary sterilization: a world view. Reports on population/family planning, 5:1-36 (1970). 6. TIETZE, C. The current status of fertility control. Law and contemporary problems, 25: 426-444 (1960). 7. nETzE, C. Mortality with contraception and induced abortion. Studies in family planning, 45: 6-8 (1969). 8. T1ETZE, C. ET AL. Mortality with legal abortion in New York City 1970-72. Journal of the American Medical Association, 225 (5): 507-509 (1973). 9. THOMPSON, B. & ILLSLEY, R. Sterilization-social and psychological aspects. IPPF medical bulletin, 6 (1): 1-2 (1972). 10. Lu, T. & CHUN, D. A long-term follow-up study of 1055 cases of postpartum tubal ligations. Journal of obstetrics and gynaecology of the British Common- wealth, 74: 875-880 (1967). 11. W1mTEHOUSE, D. B. Tubal ligation: a follow-up study. Advances in fertility control, 4 (1): 22-26 (1969). 12. PHATAK, L. Evaluation of the optimal time for tubal ligation. In: Richart, R. M. & Prager, D. J. ed. Human sterilization. Springfield, Illinois, Charles C. Thomas, 1972, pp. 86-92. 148 KARIN G. B. EDSTROM 13. KROENER, W. F. Surgical sterilization by fimbri- ectomy. American journal of obstetrics and gynecol- ogy, 104: 247-254 (1969). 14. HOFMEISTER, F. J. Postpartum sterilization. Clinical obstetrics and gynecology, 12: 676-690 (1970). 15. WIrrE, C. A. Tubal sterilization: a fifteen year survey. American journal of obstetrics and gynecol- ogy, 95: 31-39 (1966). 16. HAYES, D. M. & WOLFE, W. M. Tubal sterilization in an indigent population: report of 14 years experi- ence. American journal of obstetrics and gynecology, 106: 1044-1050 (1970). 17. AKTHER, M. S. Vaginal versus abdominal tubal ligation. American journal of obstetrics and gynecol- ogy, 115: 491-496 (1973). 18. TIETZE, C. & LEWIT, S. Joint Program for the Study of Abortion (JPSA): early medical complications of legal abortion. Studies on family planning 3 (6): 97- 122 (1972). 19. MABRAY, C. R. ET AL. Tubal sterilization morbidity on a charity hospital service. Obstetrics and gynecol- ogy 36: 204-207 (1970). 20. BLACK, W. P. & SCLARE, A. B. Sterilization by tubal ligation: a follow-up study. Journal of obstetrics and gynaecology ofthe British Commonwealth, 75: 219-224 (1968). 21. JoHN, A. H. & DUNSTER, G. D. Sterilization by posterior colpotomy. Journal ofobstetrics and gynae- cology of the British Commonwealth, 79: 381-382 (1972). 22. YUZPE, A. A. ET AL. Tubal sterilization: methodol- ogy, postoperative management and follow-up of 2934 cases. Canadian Medical Association journal, 107: 115-117 (1972). 23. SMITH, R. A. & SYmmoNDs, R. E. Vaginal salpingect- omy (fimbriectomy) for sterilization. Obstetrics and gynecology, 38: 400-402 (1971). 24. SOGOLOW, S. R. Vaginal tubal ligation at time of vacuum curettage for abortion. Obstetrics and gyne- cology, 38: 888-892 (1971). 25. LAUFE, L. E. & SUMMERSON, S. Internal vaginal tubal ligation. In: Duncan, C. W., Falb, R. D., & Spiedel, J. J., ed. Female sterilization: prognosis for a simplified outpatient procedure. New York, Aca- demic Press, 1972, pp. 73-76. 26. LAUFE, L. E. & KREUTNER, A. K. Vaginal hysterec- tomy: a modality for therapeutic abortion and sterilization. American journal ofobstetrics and gyne- cology, 110: 1096-1099 (1971). 27. ROACH, C. J. ET AL. Vaginal hysterotomy for steril- ization. American journal of obstetrics and gynecol- ogy, 114: 670-673 (1972). 28. v. NAGELL, J. R. V. Jr & RODDICK, J. W. Vaginal hysterectomy as a sterilizing procedure. American journal of obstetrics and gynecology, 111: 703-707 (1971). 29. GUTIERREZ-NAJAR, A. J. Tubal ligation by culdo- scopy. In: Advances in planned parenthood, VII. Amsterdam, Excerpta Medica, 1972, pp. 207-210. 30. LrrTLE, W. Culdoscopic outpatient sterilization (a preliminary report). In: Duncan, C. W., Falb, R. D., & Spiedel, J. J., ed. Female sterilization: prognosis for a simplified outpatient procedure. New York, Academic Press, 1972, pp. 77-82. 31. STEPTOE, P. Recent advances in surgical methods of control of fertility and infertility. British medical bulletin, 26 (1): 60-64 (1970). 32. LISTON, W. ET AL. Female sterilization by tubal electrocoagulation under laparoscopic control. Lancet, 1: 382-383 (1970). 33. JORDAN, J. A. ET AL. Laparoscopic sterilization and follow-up hysterosalpingogram. Journal of obstetrics and gynaecology of the British Commonwealth, 78: 460-466 (1971). 34. WHMTELEY, P. F. Laparoscopic sterilization in the puerperium. Journal of obstetrics and gynaecology of the British Commonwealth, 79: 166-167 (1972). 35. CHUN, D. ET AL. Laparoscopic tubal sterilization. IPPF medical bulletin, 4 (3): 1-2 (1970). 36. PETERSON, E. P. & BEHRMAN, S. J. Laparoscopic sterilization. American journal ofobstetrics and gyne- cology, 110: 24-31 (1971). 37. McMASTER, R. H. & ANSARI, A. H. Vaginal tubal ligation. Obstetrics and gynecology, 38: 44-50 (1971). 38. EDWARDS, L. E. & HAKANSSON, E. Y. Changing status of tubal sterilization. American journal of obstetrics and gynecology, 115: 347-353 (1973). 39. TURNER, G. & HOOPER, N. Sterilization and thromboembolism. Journal ofobstetrics andgynaeco- logy ofthe British Commonwealth, 78: 737-740 (1971). 40. MUSTAFA, M. A. & PINKERTON, J. H. M. Bacteriol- ogy of fallopian tube in relation to puerperal steril- ization. Journal of obstetrics and gynaecology of the British Commonwealth, 77: 171-173 (1970). 41. SPORE, W. W. ET AL. Bacteriology of postpartum oviducts and endometrium. American journal of obstetrics and gynecology, 107: 572-577 (1970). 42. RUBIN, A. & CZERNOBILSKY, B. Tubal ligation: a bacteriological, histological and clinical study. Obstetrics and gynecology, 36: 199-203 (1970). 43. THOMPSON, B. & WHEELESS, C. R. Outpatient steril- ization by laparoscopy, a report of 666 patients. Obstetrics and gynecology, 38: 912-915 (1971). 44. WHEELESS, C. R. Jr. Outpatient sterilization by laparoscopy under local anaesthesia in less devel- oped countries. In: Duncan, C. W., Falb, R. D., & Spiedel, J. J., ed. Female sterilization: prognosis for simplified outpatient procedures. New York, Academic Press, 1972, pp. 125-129. 45. BOYD, E. J. & HOLMSTROM, E. G. Ovulation follow- ing therapeutic abortion. American journal of ob- stetrics and gynecology, 113: 469-471 (1972).