Leke, Robert J. MD*; Goyaux, Nathalie PhD†; Matsuda, Tomohiro PhD‡; Thonneau, Patrick F. MD§
Whereas ectopic pregnancy is well documented in industrialized countries, little research has been done on this gynecologic emergency in developing countries, especially in African regions. Liskin,1 reviewing ectopic pregnancy incidence from the 1960s until the middle of the 1980s, reported the highest incidence rates in African countries (between 0.5% and 2.3% of live births), whereas low incidence rates were observed in Asia and the Middle East in the same period (between 0.4% and 0.6% of live births).
More recently, in our review of studies on ectopic pregnancy incidence rates in African countries, we found that incidence may well have increased in recent decades.2 In Madagascar, a hospital-based ectopic pregnancy incidence of 2.9% (live births) was observed in the island of Nosy Be between 1993 and 1995.3 In Nigeria (Ile-Ife teaching hospital), the hospital-based incidence of ectopic pregnancy quadrupled between 1977 and 1987 (0.4–1.7% of live births),4 and in Gabon (University Medical Center of Libreville) it doubled in a 12-year period from 1% in 1977 to 2.3% in 1989.5
In Cameroon, at Yaounde University Hospital, the incidence of ectopic pregnancy increased from 0.9% to 1.7% between 1984 and 1992.6. Cameroon is located in Central Africa, between Nigeria, Chad, and Gabon. The total population is estimated at 18 million inhabitants. The capital, Yaounde, lies in the center of the country with an estimated urban population of 1,250,000. Government centers, national and international institutions, banks, and universities are based in Yaounde, which is a relatively compact conurbation.
To obtain a global estimate of the population-based ectopic pregnancy incidence in African countries and also to analyze the use of medical facilities, we conducted a prospective study of ectopic pregnancy in Yaounde, the capital of Cameroon (Central Africa).
MATERIALS AND METHODS
To collect exhaustive data on all cases of ectopic pregnancy during the study period (January to December 2000), we first investigated and enrolled all surgical facilities (public and private sectors) where women with ectopic pregnancy would be taken in charge. In the 7 surgical facilities (Hopital Central, General Hospital, University Hospital, CNPS Hospital, Djoungolo Hospital, Military Hospital, Biyem Assi Hospital), gynecologists and surgical teams were enrolled and informed about the objectives of the study. A training session on the recruitment of women with ectopic pregnancy (through surgical findings) and completion of the questionnaire was organized with all gynecologic staff of the 7 surgical institutions.
During the study period of the year 2000, 356 women admitted for an ectopic pregnancy to surgical health facilities of the city of Yaounde were enrolled. Thirty-six women did not usually reside in the Yaounde area and were excluded from the analysis. An institutional review board in Yaounde (Cameroon) approved this study.
For each woman admitted for ectopic pregnancy, we collected sociodemographic information and reproductive history using a questionnaire. A first version of the questionnaire was tested in a limited number of ectopic pregnancy cases to determine the global feasibility and the quality of answers. After analysis, a final version of the questionnaire was elaborated and used by all investigators (gynecologists and surgical teams, interviewers) as a standardized tool for data collection.
Each woman was asked about lower abdominal pain before hospital admission and the date of onset of these symptoms. We also asked each woman about use of medical facilities or ultrasonography, again before her hospital admission. This was done in a face-to-face interview during the woman's hospital stay. To obtain valid information on health care behavior, particular attention was given to the conditions of the interview (confidential) and to the ability of interviewers to empathize with the women admitted.7 Female interviewers were recruited among health professionals (nurses, midwives, health workers) who had previous experience in research activities. All interviewers received specific standardized training sessions before the start of the study. The questionnaire was written in French, but interviewers were also selected for their ability to translate questions into local dialects.
At the same time, we collected for each enrolled woman medical and obstetrical data from gynecologic and surgical files and also from emergency admission records (some women admitted in emergency may be directly hospitalized and treated in emergency units without any contact with gynecologic services). The following variables were collected: medical history (previous gynecologic disorders, history of previous ectopic pregnancy, previous chlamydiae infection), clinical findings at hospital entry, type of surgery (laparotomy, celioscopy), existence of hemoperitoneum, type of treatment (radical or conservative), postsurgical complications (infection or hemorrhage), and final vital status.
After a complete reassessment of all cases of ectopic pregnancy by the research staff (composed of African and French obstetricians), statistical analysis was performed with STATA 6.0 (Stata Corporation, College Station, TX). The χ2 test and 95% confidence intervals were used for estimated proportions.
During the year 2000, we recorded 320 cases of ectopic pregnancy in health facilities in the city of Yaounde. Based on the last census (1987) and on the results of the Demographic and Health Survey in Cameroon (1998), it is estimated that 40,100 (36,090–44,110) live births occurred during the study period (January to December 2000) in the city of Yaounde.8 Thus, the population-based incidence rate of ectopic pregnancy in the city of Yaounde was 0.79 (0.72, 0.88) per 100 live births in 2000.
By comparison, if we used only pregnancies registered in participating health facilities as the reference, the hospital-based incidence rate of ectopic pregnancy was 2.55 per 100 pregnancies (320 of 12,523).
Sociodemographic characteristics and medical histories of the women are listed in Table 1. The women's mean age was 23.4 (± 3.2) years, with 16% being more than 34 years old. For 13% it was their first pregnancy, and 32% already had 4 or more children.
One fourth (24%) of the women had previously had gynecologic disorders and 7% had already experienced an ectopic pregnancy. Six percent of the women mentioned previous Chlamydia trachomatis infection.
Laparotomy was performed in 98% of cases and celioscopy in 2%. Hemoperitoneum was observed in 93% of cases, but conservative treatment of the tube was noted in 24% of patients. The rate of postsurgical complications was relatively low at 5%.
Among the 320 patients admitted for an ectopic pregnancy, 3 maternal deaths were recorded, giving a mortality rate of 0.94% (0.32%, 2.72%). The localization of ectopic pregnancy was the isthmus and the fimbria. In 1 case, the woman was dead on admission and no details were given. For these 3 maternal deaths, the time between the development of symptoms and time of hospital admission was relatively short (between 1 and 2 days)
Voluntary sterilization (tubal ligation) was performed in 7% of women.
Figure 1 illustrates the women's attitude regarding symptoms and use of medical facilities. In the 236 women who answered that they had had “belly pain” before hospital admission, we defined 2 groups of women according to their course of action. Group A (n = 38) was composed of women who had sought care directly at hospital level, and group B (n = 198) of women who had first consulted in an urban facility (health center, medical clinic, private consultation) before admission to the hospital.
In group B, 63% of women (125 of 198) had undergone ultrasound scanning at this urban health facility. For 84 women, items regarding symptoms and consultations in health facilities had not been completed.
We then calculated for each woman in the 2 groups the time elapsing between the onset of lower abdominal pain and the date of hospital admission. The time elapsing between the first symptoms and admission to hospital was much shorter in group A than in group B (2 days versus 4 days).
Indeed, after having had a medical consultation in a health facility, women in group B waited 2 days (median) before hospital admission. Nevertheless, no significant difference was observed in complication rate and vital status between the 2 groups of women (3 of 38 [7.9%] in group A compared with 10 of 198 [5.1%] in group B; P = .454, Fisher exact test).
In this study performed in Yaounde, the capital of Cameroon, we observed a population-based ectopic pregnancy incidence rate of 0.8% (95% confidence interval 0.72%, 0.89%) of live births.
The ectopic pregnancy incidence rate found in Yaounde, the capital of Cameroon, is lower than that observed in Europe and North America, where it is currently estimated at between 1.2% and 1.9% and includes ectopic pregnancies that occur after in vitro fertilization.9–11
However, it is not valid to compare the incidence of ectopic pregnancy in industrialized countries with that in developing countries. In fact, in industrialized countries, where almost all births take place in maternity units and where ectopic pregnancy is treated in emergency health facilities, the total number of ectopic pregnancy cases can be calculated fairly precisely.12 Therefore, the numerator is often available because the number of ectopic pregnancy cases is registered (especially in cohort studies), and the denominator can also be calculated without difficulty from the number of live births.
In developing African countries, the situation is radically different. Recourse to health care is often the last step in a diversified therapeutic pathway, and access to hospitals remains limited for socioeconomic and health organization reasons. Generally, in Africa, recruitment of patients from hospitals is therefore highly selective: it concerns the women who have survived and those geographically and financially able to obtain access to a reference structure for surgical management. This being so, it is often difficult to perform valid study through cases only registered in reference hospitals. This is why, in our study, we systematically collected data regarding all cases of ectopic pregnancy registered in all health facilities in the city of Yaounde (private and public health facilities located in large and small institutions, gynecological and obstetrics units, and also surgery and emergency units). Even though we certainly did not identify all of the cases of ectopic pregnancy in the city of Yaounde during the study period, it is reasonable to think that our systematic approach permitted the collection of the majority of these cases.
In the African population, where there are no birth registries and limited census, the number of live births can be estimated only through demographic surveys. We used the last Demographic and Health Survey performed in Cameroon (1997) to derive the total number of live births and thus obtain an estimate of the denominator.
Even though we tried to limit potential bias (eg, participation of all institutions offering surgical facilities in the study area, careful attention to all hospital admissions of women for any gynecologic disorders to detect undiagnosed ectopic pregnancy), our observed ectopic pregnancy incidence rate must be considered a minimum level. Taking into account our incidence estimate, 0.8% of live births, and the potential underestimation, it is reasonable to think that the ectopic pregnancy incidence rate in Cameroon is lower than that currently observed in industrialized countries. Our descriptive study is unable to explain such a difference, and studies with adequate design (case-control and cohort studies; an African multicenter case-control study is currently in preparation) must be performed to identify the different ectopic pregnancy risk factors in African countries.13
In our study, the hospital-based incidence rate of ectopic pregnancy was 2.55 per 100 pregnancies (320 of 12,523). This was in the upper range of what we observed in other African studies conducted with the same design. A retrospective study performed in 1 reference gynecologic unit of Yaounde found a lower estimated ectopic pregnancy incidence (1.7%), but this study was performed nearly 10 years ago.6 This potential increase in ectopic pregnancy incidence observed in the 2 studies over a 10-year period is in agreement with our recent finding that, in developing countries, there has been a tendency toward an increased incidence of ectopic pregnancy in recent decades.2
In our study, ectopic pregnancy was diagnosed late, leading to significant delays in hospital admission. The very low use of ultrasound scans observed in our study is common to studies performed in African countries.14,15 In a study conducted in Ghana, Obed et al16 assessed the impact of the introduction of ultrasound scans as a diagnostic tool in a reference hospital in Accra. Ectopic pregnancy management was compared in patients who underwent transabdominal ultrasound scans and in those who did not. The proportion of unruptured ectopic pregnancy was higher (8.5% versus 0.3%) and the frequency of misdiagnosis lower in women who underwent scans than in women who did not. Surprisingly, in our study we did not find that ultrasonography was a relevant tool to establish ectopic pregnancy diagnosis, nor did we find that it could reduce the time to hospital admission. These discordant results could perhaps be explained by the fact that numerous ultrasound scans done in health facilities may have been performed by unskilled health professionals or in women with a very short duration of amenorrhea.
Finally, in our series, only 24% of patients received conservative surgical treatment. This finding was comparable with results in Benin and in the Ivory Coast.17,18.
Three maternal deaths were recorded during the study period, giving an ectopic pregnancy fatality rate of 1%. This is comparable with the rate reported in developing countries but 100 times higher than that reported in industrialized countries (eg, a case fatality rate of 0.04% was observed in the UK in 1994–1996).19–22 Nevertheless, we must be aware that data collection on maternal deaths linked to ectopic pregnancy complications may well be incomplete. For example, analyzing the circumstances of death of women admitted for ectopic pregnancy, Baffoe and Nkyekyer23 have shown that the principal difficulty is to establish the immediate cause of maternal deaths, with a high risk of misclassification between ectopic pregnancy and induced abortion (especially if the woman did not undergo surgery). Therefore, the ectopic pregnancy fatality rate of 1% observed in the town of Yaounde must be considered as a minimum level.
The small number of cases of maternal deaths (n = 3) did not permit valid multivariate analysis to identify risk factors. Nevertheless, the dramatically high percentage of women with hemoperitoneum indicates that delay in diagnosis and in surgical unit admission are certainly 2 crucial issues. These findings must incite gynecologists and public health advisers to promote information and education campaigns to encourage women to consult more rapidly for gynecologic disorders. At the same time, continuing education must be developed within medical and gynecologic networks to draw greater attention to the frequency and management of ectopic pregnancy, which must be considered a leading cause of maternal mortality in developing countries.
In conclusion, the 0.79% ectopic pregnancy incidence rate observed in this African country must be considered a minimum due to probable underestimation. Nevertheless, this rate is lower than that currently observed in industrialized countries. To explain such differences between industrialized and developing countries, cohort and case-control studies must be performed to identify specific ectopic pregnancy risk factors in African countries.
Late diagnosis, low percentage of conservative treatment, and subsequent maternal deaths are important findings that should encourage African gynecologists to promote ectopic pregnancy prevention programs and to improve the care given to women with ectopic pregnancy.
1. Liskin LS. Maternal morbidity in developing countries: a review and comments. Int J Gynaecol Obstet 1992;37:77–87.
2. Goyaux N, Leke R, Keita N, Thonneau P. Ectopic pregnancy in developing countries: a review. Acta Obstet Gynecol Scand 2003;82:305–12.
3. Ratinahirana S, Razanamparany PV, Radaniarison H, Ratsimanohatra E, Rakotozafy G. Current aspects of ectopic pregnancies in Nosy Be (Madagascar), from November 1993 to February 1995 [in French]. Cahiers Sante 1997;7:19–23.
4. Makinde OO, Ogunniyi SO. Ectopic pregnancy in a defined Nigerian population. Int J Gynaecol Obstet 1990;33:239–41.
5. Picaud A, Nlome-Nze AR, Faye A, Kouvahe V, Kpodar D, Ouzouaki F, Nikili MT, Saleh S. Progression of the incidence of ectopic pregnancy in Libreville, Gabon, from 1977 to 1989 [in French]. Med Afri Noire 1992;39:798–805.
6. Kouam L, Kamdom-Moyo J, Doh AS, Ngassa P. Treatment of ectopic pregnancies with laparotomy in under-equipped countries. A series of 144 cases from Yaounde, Cameroon) [in French]. J Gynecol Obstet Biol Reprod 1996;25:804–8.
7. Rasch V, Muhammad H, Urassa E, Bergström S. Self-reports of induced abortion: an empathetic setting can improve the quality of data. Am J Public Health 2000;90:1141–4.
8. Demographic and health survey in Cameroon (EDSC-II). Calverton (MD): Bureau Central des Recensements et des Etudes de Population de Yaounde, DHS Program of Macro International; March 1999.
9. Coste J, Bouyer J, Germain E, Ughetto S, Pouly J-L, Job-Spira N. Recent declining trend in ectopic pregnancy in France: evidence of two clinicoepidemiologic entities. Fertil Steril 2000;74:881–6.
10. Storeide O, Veholmen M, Eide M, Bergsjo P, Sandvei R. The incidence of ectopic pregnancy in Hordaland County, Norway, 1976–1993. Acta Obstet Gynecol Scand 1997;76:345–9.
11. Ectopic pregnancy: United States, 1990–1992. MMWR Morb Mortal Wkly Rep 1995;44:46–8.
12. Rajkhowa M, Glass MR, Rutherford AJ, Balen AH, Sharma V, Cuckle HS. Trends in the incidence of ectopic pregnancy in England and Wales from 1966 to 1996. Br J Obstet Gynaecol 2000;107:369–74.
13. Bjartling C, Osser S, Persson K. The frequency of salpingitis and ectopic pregnancy as epidemiologic markers of Chlamydia trachomatis. Acta Obstet Gynecol Scand 2000;79:123–8.
14. Dissanayake L, Lema VM, Wanjala SH, Wachira MW. The value of ultrasonography in the diagnosis of ectopic pregnancy at the Kenyatta National Hospital, Nairobi. East Afr Med J 1993;70:535–9.
15. Nko'o Amvene S, Biwole M, Kasia JM, Shasha W, Essane JL, Oyono E, et al. Ectopic pregnancy in tropical climate: an ultrasonographic study of 93 consecutive cases [in French; abstract in English]. J Radiol 1993;74:589–92.
16. Obed SA, Wilson JB, Elkins TE. Diagnosing unruptured ectopic pregnancy. Int J Gynaecol Obstet 1994;45:21–5.
17. De Muylder X. Ectopic pregnancy in Zimbabwe. Int J Gynaecol Obstet 1991;35:55–60.
18. Abdul IF. Ectopic pregnancy in Ilorin, Nigeria. Int J Gynaecol Obstet 1999;66:179–80.
19. Bugalho A, Strolego F, Pregazzi R, Osman N, Ching C. Extrauterine pregnancy in Mozambique. Int J Gynaecol Obstet 1991;34:239–42.
20. Martey JO, Djan JO, Twum S, Browne EN, Opoku SA. Maternal mortality and related factors in Ejisu District, Ghana. East Afr Med J 1994;71:656–60.
21. Dorfman SF. Maternal mortality in New York City, 1981–1983. Obstet Gynecol 1990;76:317–23.
22. De Swiet M. Maternal mortality: confidential enquiries into maternal deaths in the United Kingdom. Am J Obstet Gynecol 2000;182:760–6.
23. Baffoe S, Nkyekyer K. Ectopic pregnancy in Korle Bu Teaching Hospital, Ghana: a three-year review. Trop Doct 1999;29:18–22.