Mortality associated with medical termination of pregnancy.
India liberalised abortion with the Medical Termination of Pregnancy Act, 1971. The King Edward Memorial Hospital, Bombay, has been performing abortions since April 1, 1972. This article reviews the abortion programme at this institution with special emphasis on the mortality associated with legal abortions so as to define measures for reducing the mortality.
The cases under review were amongst those that underwent medical termination of pregnancy (M.T.P.) for a period of six years from April 1, 1972 to March 31, 1978. During this period a total of seven thousand seven hundred and fifty abortions were carried out with a total of seven deaths associated with the abortions. During the same period, thirteen patients admitted as emergency cases died as a result of procedures performed outside our institution.
The data shows that the delivery to M.T.P. ratio which was 10:1 in 1972 has slowly fallen to 2.92:1 in 1978, showing the tremendous increase in the number of abortions.
It is seen from [Table 1] that the mortality associated with delivery is distinctly more than that associated with abortion demonstrating that abortion is a relatively safe procedure.
Of the seven deaths at our institution [Table 2], there were no first trimester M.T.P. deaths. Of the seven second trimester deaths, four were associated with intra-amniotic hypertonic saline, two with extra-ovular mannitol and one with extra-ovular hypertonic saline. Thus, saline-associated deaths top the list of mortality with abortions. All these seven patients died despite availability and use (when required) of various intensive care measures, such as vital sign and C.V.P. monitoring, laboratory facilities, emergency operation theatre facility, blood transfusions, fibrinogen, antibiotics, respirators and facilities for dialysis.
Among the transferred cases [Table 3], there were eight deaths associated with dilatation and curettage (D & C) or suction evacuation (S.E.), three with intra-amniotic hypertonic saline and two with other methods-one was intra-amniotic distilled water and the other was placement of an extra-ovular catheter.
Only five deaths occurred with the uterus size of twelve weeks or smaller while fifteen were associated with the uterus size larger than twelve weeks [Table 4].
As a result of the increasing demand for legal abortion all over the country, it has been found appropriate to review the risk of mortality associated with these procedures. At our institution seven deaths have occurred over a six year period during which 7,750 medical terminations of pregnancy were performed giving a mortality rate of 0.093 per one thousand procedures.
Surveying the deaths at our institution, we see no death with M.T.P. in the first trimester. First trimester is indeed a safe period for M.T.P. with a mortality ratio of only 0.4 per 100,000 abortions for uterine size less than eight weeks as reported by Cates et al. According to the same authors, this mortality rises by five times if the uterine size is nine to ten weeks. Menstrual regulation, a relatively newer technique is a very safe procedure for early first trimester M.T.Y. as seen by no mortality in over 1,533 procedures performed by us. Pachauri and Fortney9 reporting on 6,549 cases of menstrual regulation found no deaths associated with the procedure.
Rapid dilatation with evacuation of the products by sharp curettage or suction have been traditional abortion methods. These probably carry a little risk of death. Fortunately, we have had no deaths from 3,689 procedures. Berger et al. reviewing the New York State mortality from abortions for 1970-72 find a maternal mortality of 2.1 per 100,000 for sharp curettage and 2.8 per 100,000 for suction evacuation. Teitze and Bangaarts12 further strengthen the case of first trimester abortion by finding a death to case ratio of three or less per 100,000 abortions whilst the lowest recorded maternal mortality rates are in the range of 8-18 per 100,000 live births. In fact, in countries such as the U.S.S.R., Japan and East Europe, there are more abortions performed than there are live births!
There were eight deaths associated with dilatation and curettage or suction evacuation from the transferred cases. Of these deaths, five procedures were carried out on uteri larger than twelve weeks. The danger of performing these procedures on larger sized uteri has been stressed earlier by Krishna et al. The greatest dangers of these methods have been haemorrhage and shock, sepsis and genital trauma. From amongst the cases seen by us, five deaths were due to sepsis, two were due to perforative peritonitis whilst one was due to suspected pulmonary embolism. Many of these procedures were performed by junior surgeons.
In the second trimester abortion group, all our deaths were due to procedures done in this period of gestation. The multitude of methods in vogue for second trimester M.T.P. speaks for the far-from ideal methods available today. Of the seven deaths at our institution, five were due to saline one due to extra-ovular saline and four due to intra-amniotic hypertonic saline. Death due to coagulopathy occurred in two of these five cases; whilst in one, there was additional evidence of a uterine (fundal) rupture. Of the two other cases, one died of sepsis whilst the other has suspected bacterial endocarditis. A post-mortem examination was not done in the last case. Coagulopathy is a known complication of saline abortions. There is a significant change in coagulation factors, such as prothrombin time, fibrinogen, platelets, factors V and VIII in virtually every patient receiving hypertonic saline.,  A maximum charge in the above factors occurs at around nine hours post-injection of hypertonic saline. However, the above coagulation factors reach their normal levels by the, time abortion occurs. Nevertheless, several case reports describe occasionril serious and life threatening instances of coagulation defects following saline induced abortions., Coagulopathy was responsible for 60% of our deaths associated with saline abortions. The case that died of bacterial endocarditis had had a tubal sterilization procedure in addition. The fifth case had an extra-ovular saline procedure and died of fulminant sepsis. Any extra-ovular method is open to criticism as we keep a foreign body (catheter) in a potentially infected area for a stipulated period of time. Use of agents with antiseptic properties have not significantly lowered the sepsis rate with extra-ovular method. Strict asepsis and prophylactic antibiotics may help to reduce sepsis associated with these procedures.
Mannitol as an extra-ovular procedure was responsible for two of our deaths. One died of gas gangrene, whilst the other was a case of failed first trimester M.T.P. She was induced at fourteen weeks gestation by this method (extra-ovular mannitol) and she aborted. The use of intravenous pitocin expedited the abortion process. She developed hypofibrinogenaemia and acute renal shutdown and died in spite of fibrinogen, blood and haemodialysis.
Of the referred cases, three belonged to the saline group. All these cases had initially been induced by the intraamniotic hydrocortisone method. When this failed, the saline method was tried. Of the two `other' cases, one had intraamniotic distilled water and died of septicaemia and myocarditis whilst the remaining case was induced with the catheter method so popular in Japan. She too died of sepsis, 900 ml of pus being present in her peritoneal cavity at the time of post-mortem examination.
It is quite evident from the above that saline abortions are dangerous and should not be performed. Our death rate from saline is 5.50 per 1,000 cases. Wagatsuma reported a death rate with saline of 1.96 per thousand cases in 1965. More recently, Schiffer has reported a rate of 0.20 per 1000 cases. Burnett et al2 reporting on the Joint Program for the study of abortion cases, give a death rate of 0.14 per thousand cases with saline. Prostaglandin in our hands have proved quite safe with no deaths from seven hundred fifty abortions. They are, however, expensive and not easily available. Moreover, Cates et al. from their cases state that the relative safety of intraamniotic prostaglandin F2 alpha over intra-amniotic hypertonic saline remains to be established.
Catheters or bougies, popular in Japan, have several disadvantages. These include, according to Grimes et al.  long induction to abortion interval, need for a second procedure and a high complication rate.
We have performed hysterotomy with tubal sterilisation quite successfully in two hundred ninety eight cases so far without any deaths. Berger et al. report a mortality of 271.2 per 100,000 cases associated with hysterotomy. The high risk cases unsuitable for other methods of abortion are taken up for hysterotomy and hence the high mortality. In fact, Cates et al. (1972-1975) report a mortality of 42.6 per 100,000 cases. In our hands hysterotomy with tubal sterilization has a low morbidity and nil mortality rate. Perhaps, we have not done that many cases to prove or disprove the high mortality rate reported earlier.
Indeed, second trimester abortion is a relatively dangerous procedure. Young age, good general health, early abortion, strict asepsis, experience of the surgeon, strict quality control over the method used, use of prophylactic antibiotics, inpatient hospitalization are some of the factors that may lead to diminished mortality and morbidity associated with M.T.P.
Seven thousand seven hundred and fifty abortions were performed at the K.E.M. Hospital, Bombay, over a period of six years with seven deaths for a mortality rate of 90.3 per 100,000 abortions. No deaths occurred in first trimester abortions.
Prostaglandins appear to be relatively safe for second trimester abortion; while in properly selected cases, hysterotomy has a low mortality and morbidity. Saline with its high morbidity and mortality should be used with discretion as a method for second trimester abortion.
Sepsis and coagulopathy were the two important complications associated with high mortality.
We thank Dr. C. K. Deshpande, Dean, K.E.M. Hospital and Dr. V. N. Purandare, Head, Department of Obstetrics and Gynaecology, K.E.M. Hospital, Bombay, for allowing us to publish the hospital data.