High prevalence of sexually transmitted infections among asymptomatic women opting for the intrauterine contraceptive device use in Mwanza, Tanzania: An urgent call for control interventions.

Objective: Placement of intrauterine contraceptive device (IUCD) in asymptomatic woman infected with sexually transmitted infection (STIs) can lead to pelvic inflammatory diseases (PID) and infertility if not well treated. The current study investigated the magnitude of sexually transmitted infections among women opting for IUCD use in the city of Mwanza, Tanzania.
Methods: A cross-sectional study involving 150 asymptomatic women was conducted from August to December 2017. Detection of Chlamydia trachomatis antigen from endocervical swabs was done using immunochromatographic rapid tests while sera were used for detection of Treponema pallidum, human immunodeficiency virus (HIV) and herpes simplex virus Type 2 (HSV-2) antibodies.
Results: The overall prevalence of STIs was 45/150 (30%, 95% CI: 22-37) while that of individual STIs were 27.3%, 5.3%, and 2.6% for C trachomatis, T pallidum, and HSV-2, respectively. History of dysuria (aOR 6.6; 95% CI 2.3-18.8; p < 0.001) and history of STIs (aOR 4.6; 95%CI 1.0-20.8; p = 0.049) independently predicted presence of STIs. Conclusions: Prevalence of STIs among women opted for IUCD use in the city of Mwanza, Tanzania is alarmingly high and is predicted by past history of dysuria and history of partner's STIs, calling for the need of screening of the STIs among high-risk women in low- and middle-income countries (LMICs) opting for IUCD use.

Introduction

Sexually transmitted infections (STIs) remain a global public health problem. The World Health Organization (WHO) estimates that about 500 million new STIs occur annually among people aged between 15 and 49 years. In sub-Saharan Africa, the incidence of STIs among population aged between 15 and 49 years is about 240/1000 which is the highest record in the world. It is estimated that about 8.3, 21.1 and 59.7 million new cases of Chlamydia trachomatis, Neisseria gonorrhoea and Trichomonas vaginalis infections, respectively, occur in sub-Saharan Africa annually. N. gonorrhoea, C. trachomatis and T. vaginalis have been found to increase human immunodeficiency virus (HIV) acquisition and transmission. In Mwanza Tanzania, the prevalence of N. gonorrhoea among adult patients attending care and treatment clinics (CTC) using polymerase chain reaction (PCR) was reported to be 2%. Moreover, a study which was done in 10 different antenatal clinics in Mwanza, Tanzania among adolescent pregnant women showed that 199 (49.4%) had at least one STI. The prevalence of STIs among HIV women opting for intrauterine contraceptive device (IUCD) in Uganda was reported to be 5.9%.

Approximately, 160 million women worldwide use IUCD making it the most popular contraceptive method after sterilization. The use of IUCD in the city of Mwanza with an estimated population of 1.2million people has been found to increase in the past 3 years. It was noted that in 2014, 2015 and 2016, a total of 11,170, 12,292 and 14,807 women opted IUCD use, respectively.[7-9] In an asymptomatic woman, the placement of IUCD can result in the transmission of pathogens responsible for STIs into the uterine cavity leading to development of pelvic inflammatory disease (PID), chronic pelvic pain and infertility. PID-associated intrauterine contraception is the commonest in the first 20 days of intrauterine contraception initiation.[11,12] The United States Center for Disease Control and Prevention (CDC)) and WHO recommend women at high risk of STIs to undergo STIs screening before the placement of IUCD.[13,14] However, in low- and middle-income countries (LMICs) such services are not readily available. Syndromic approach has been used to screen women before IUCD insertion, however, the approach has low sensitivity. Tanzania, like many other LMICs, the syndromic management for STIs is being implemented. Evidence shows that 30%–80% of the women with N. gonorrhoea, 85% with C. trachomatis and 80% with T. vaginalis are asymptomatic. This indicates that a large number of women might be living with asymptomatic STIs.

STIs symptoms have low sensitivity and specificity in detecting STIs signifying the importance of STIs screening to detect STIs before IUCD insertion. Despite the fact that the prevalence of these infections is high in LMICs, there are limited number of studies that focused on women opting for IUCD use. In a view of this, this study investigated the presence of common STIs among women undergoing syndromic screening for STI prior to placement of IUCD. These findings are crucial in revising current protocol in LMICs in order to reduce IUCD associated morbidities.

Methods

Study design, duration, target population and study area

A cross-sectional health facility-based study was conducted between August and December 2017 in the city of Mwanza. Enrolment was done at Makongoro and Uzazi na Malezi bora Tanzania (UMATI) clinics. The total population served by these clinics is about 56,442 with approximately 500 IUCD insertions in a year. The study included family planning clinic attendees aged 18 years and above who opted for IUCD use.

Sample size estimation and sampling technique

The sample size was obtained by Kish Leslie formula using the prevalence of 5.9%. The minimum sample size was 85 women, however, a total of 150 women were enrolled. The study participants were enrolled conveniently until the sample size was reached.

Selection criteria

The study included women who were asymptomatic for STIs and seeking for IUCD placement service. Women who presented with signs and symptoms of STIs after physical examination (e.g. abnormal vaginal discharge purulent with yellow or brownish colour with foul-smelling, genital sores or blisters, painful intercourse, genital itching, lower abdominal pain, painful urination, pregnant women, those having gynaecological conditions such as cancer of the cervix, gestation trophoblastic disease, endometrial and ovarian cancer) as per national guidelines for management of sexually transmitted and reproductive tract infections were excluded. All women who tested positive were managed as per the Tanzania Standard Treatment Guidelines.

Data/sample collection and sample processing

Pre-tested structured questionnaire was used to collect socio-demographic and other relevant information such as education level, age, religion, marital status, participant’s alcohol use, disclosure of HIV status to partner, use of ARV, condom use, CD4 count, parity and number of sexual partners per year. About 4–5 mL of venous blood sample was collected aseptically from each consented participant and placed in plain vacutainer tubes (Becton Dickson and Company, Kenya). Sera were separated and stored at -80ºc until processing.

Sera were used for detection of HSV-2 using immunochromatographic rapid tests as per manufacturer instructions (INVBIO Biomaterials Solutions, Beijing, China). Detection of HIV 1 and 2 was done as per Tanzania protocol for HIV screening. For T. pallidum (Syphilis) detection, a forward algorithm was used whereby a non-Treponemal test, T. pallidum antibodies (ARKRAY Healthcare Pvt. Ltd., Surat, India) followed by a treponemal test and T. pallidum hemagglutination test (TPHA) (Human Biochemical and Diagnostic, Wiesbaden, Germany) confirmation.

Endocervical swabs were collected and used to detect C. trachomatis infections as detailed in the manufacturer’s instructions (INVIBIO Biotech Co Ltd, Beijing, China), with sensitivity detection of 105 bacteria/ml and specificity of 97.5% to detect chlamydia antigen.

The Papanicolaou (PAP) smear was stained as previously described and examined under light microscope in low- and high-power objectives to observe pathological findings such as necrosis and presence of different inflammatory cells such as neutrophils, lymphocytes etc.

Data analysis

Data were collected and entered into computer Microsoft Excel 2007 and later analysed using STATA version 12. Continuous variables (age, gravidity etc.) were summarized using median with interquartile range (IQR) and categorical variables (residence, occupation marital status, etc.) were summarized as proportions. Test for association was done using chi- square test. Univariate and multivariate logistic regression model was used to determine independent predictors for STIs. A p-value of less than 0.05 at 95% confidence interval was considered statistically significant.

Results

Sociodemographic characteristics of asymptomatic women opted for IUCD use

A total of 150 asymptomatic women opted for IUCD placement were enrolled with the median age of 26 [IQR 23–32] years. The majority 129 (86.0%) aged between 18 and 35 years and most of them 144 (96.0%) were from urban areas. Three quarters 113 (75.3%) were Christians and more than three quarters 126 (84%) were married (Table 1).

Table 1.

Distribution of sociodemographic data among 150 asymptomatic women opted for IUCD use in Mwanza city.

Client particulars	Number (n)	Percent (%)
Age group
18–35 years	129	86.0
> 35 years	21	14.0
Residency
Rural	6	4.0
Urban	144	96.0
Religion
Christian	113	75.3
Muslim	36	24.0
Others	1	0.7
Marital status
Married	126	84
Not Married	24	16
Occupation
Gov. Employee	8	5.3
Not Employed	55	36.7
Business	66	44.0
Self Employed	21	14.0
Education level
Primary	90	60.0
Sec/Collage	53	35.3
Illiterate	7	4.7
Parity
Primipara	53	35.3
Multipara	76	50.7
Grand Multipara	19	12.7
Nulliparous	2	1.3
Health facility
Butimba	66	44.0
Makongoro	68	43.3
Umati	16	10.7
Sex frequency
Once or not at all/week	25	16.7
Twice or more/week	125	83.3

History of symptoms of STIs and sexual behaviours among women opted for IUCD use

Lower abdominal pain 35/150 (23.3%), vaginal discharges 31/150 (20.7%) and dysuria 23/150 (15.3%) were reported as history (at least 3 months before seeking placement) of STIs symptoms. The most commonly reported high risk sexual behaviours were multiple sexual partners 28/150 (18.7%) (Table 2).

Table 2.

History of symptoms of STIs and sexual behaviours among women opted for IUCD use in Mwanza city.

Past symptoms	Number	Percent (%)
History of vaginal discharge
Yes	31	20.7
No	119	79.3
History of dysuria
Yes	23	15.3
No	127	84.7
History of lower abdominal pain
Yes	35	23.3
No	115	76.7
History of dyspareunia
Yes	17	11.3
No	133	88.7
History of vaginal bleeding
Yes	5	3.3
No	145	96.7
History of discharge to partner
Yes	1	0.7
No	149	99.3
History of multiple sex partner
Yes	28	18.7
No	122	81.3
History of past STI infection
Yes	11	7.3
No	139	92.7
History of blisters
Yes	18	12.0
No	132	88.0

Prevalence of specific STIs among asymptomatic women opted for IUCD use in Mwanza City

Out of 150 enrolled asymptomatic women opted for IUCD use, the prevalence of HSV2 (HSV-2, IgM), T. pallidum and C. trachomatis were 3/115 (2.6%, 95% confidence interval (CI): 0.3–5.6), 8/150 (5.3%, 95% CI:1.7–9.0) and 34/150 (22.7%, 95% CI: 15.9–29.4), respectively. The overall prevalence of STIs was 45/150 (30.0%, 95% CI: 22.6–37.4). In addition, the sero-prevalence of specific HSV-2 IgG antibodies were found to be 34/150 (22.6%, 95% CI: 15.9–29.3). Among the participants, only (4%) were known to be HIV positive and were on antiretroviral therapy (ART).

Cytological changes observed among women opted for IUCD use in Mwanza city

Among cytological changes observed, the most common features were the presence of chronic inflammation indicated by the presence of lymphocytes which was observed in 56 (37.3%) women and acute inflammation as indicated by the presence of predominantly neutrophils which was observed in 14 (9.3%) women (Figure 1). However, none of the observed histopathological changes were found to be associated with any of studied STIs.

Figure 1.

(a): Neutrophils and superficial cells (b): Lymphocytes.

Factors associated with active STIs among asymptomatic women opted for IUCD use in

On univariate logistic regression analysis, history of vaginal discharge (odds ratio (OR) 2.8; 95% CI 1.2–6.3; p = 0.014), history of lower abdominal pain (OR 2.9; 95%CI 1.3–6.5; p = 0.007), history of per vaginal bleeding (OR 10.1; 95%CI 1.1–93.3; p = 0.041) were significantly associated with STIs among asymptomatic women opted for IUCD use. By multivariate logistic regression analysis, the odds of having history of dysuria (OR 6.6; 95% CI 2.3–18.8; p < 0.001) and the odds of having a partner with history of STIs (OR 4.6; 95%CI 1.0–20.8; p = 0.049) independently predicted the presence of active STIs among women opted for IUCD use in the city of Mwanza (Table 3).

Table 3.

Factors associated with active STIs among asymptomatic women opted for IUCD use in Mwanza city.

Client characteristics	STI		Univariate		Multivariate
	Yes	No
	n (%)	n (%)	OR [95% CI]	p-value	OR [95% CI]	p-value
Age in years
18–35	39 (30.2)	90 (69.8)	1.0
>35	6 (28.6)	15 (71.4)	0.9 [0.3–2.6]	0.878
Residence
Urban	42 (29.2)	102 (70.8)	1.0
Rural	3 (50.0)	3 (50.0)	0.4 [0.1–2.1]	0.289
Religion
Christian	37 (32.7)	53 (46.9)	1.0
Muslim	8 (22.2)	28 (77.8)	1.7 [0.7–4.1]	0.234
Others	0 (0.0)	1 (100.0)
Marital status
Not married	7 (29.2)	17 (70.8)	1.0
Married	38 (30.2)	88 (69.8)	0.9 [0.4–2.5]	0.923
Occupation
Employed	1 (12.5)	7 (87.5)	1.0
Self employed	3 (14.3)	18 (85.7)	1.2 [0.1–13.2]	0.901	1.1 [0.1–15.7]	0.936
Business	20 (30.3)	46 (69.7)	3.0 [0.4–26.4]	0.313	4.0 [0.4–40.9]	0.241
Not employed	21 (38.2)	34 (61.8)	4.3 [0.5–37.7]	0.185	5.7 [0.5–58.7]	0.145
Education level
Sec/collage	14 (26.4)	39 (73.6)	1.0
Illiterate/Primary	31 (32.0)	66 (68.0)	1.3 [0.6–2.8]	0.479
Parity
Null/Prime para	18 (32.7)	37 (67.3)	1.0
Multi/Grand para	27 (28.4)	68 (71.6)	0.8 [0.4–1.7]	0.579
Health facility
Butimba	17 (25.8)	49 (74.2)	1.0
Makongoro	23 (33.8)	45 (66.2)	1.3 [0.7–3.1]	0.309
UMATI	5 (31.2)	11 (68.8)	1.3 [0.4–4.3]	0.657
Sex frequency
Once or not at all	8 (22.0)	17 (78.0)	1.0
Twice or more/week	37 (29.6)	88 (70.4)	0.9 [0.4–2.3]	0.811
H/vaginal discharge
No	30 (25.2)	89 (74.8)	1.0
Yes	15 (48.4)	16 (51.6)	2.8 [1.2–6.3]	0.014
History of dysuria
No	30 (23.6)	97 (76.4)	1.0
Yes	15 (65.2)	8 (34.8)	6.1 [2.3–15.7]	<0.001	6.6 [2.3–18.8]	<0.001
H/lower abdominal pain
No	28 (24.3)	87 (75.7)	1.0
Yes	17 (48.6)	18 (51.4)	2.9 [1.3–6.5]	0.007
History of dyspareunia
No	38 (28.6)	95 (71.4)	1.0
Yes	17 (41.2)	10 (58.8)	1.8 [0.6–4.9]	0.290	–	–
H/vaginal bleeding
No	41 (28.3)	1.4 (71.7)	1.0
Yes	4 (80.0)	1 (20.0)	10 [1.1–93.5]	0.041	5.5 [0.5–62.6]	0.170
Discharge (sexual partner)
No	45 (30.2)	104 (69.8)	1.0
Yes	0 (0.0)	1 (100.0)	Omitted	–
Multiple sexual partner
No	35 (28.7)	87 (71.3)	1.0
Yes	10 (35.7)	18 (64.3)	1.4 [0.6-3.3]	0.466
Partners STI infection
No	38 (27.3)	101 (72.7)	1.0
Yes	7 (63.6)	4 (36.4)	4.7 [1.3–16.8]	0.019	4.6 [1.0–20.8]	0.049
History of blisters
No	40 (30.3)	92 (69.7)	1.0
Yes	5 (27.8)	13 (72.2)	0.9 [0.3–2.6]	0.826
HIV status
No	42 (29.6)	100 (70.4)	1.0
Yes	3 (37.5)	5 (62.5)	1.4 [0.3–6.3]	0.636

OR: odds ratio; CI: confidence interval.

Discussion

This study has reported the burden of STIs among women opted for IUCD use in Mwanza, Tanzania, the overall prevalence of active STIs was 30.0% which is higher than two previous studies[5,6] from East Africa. In the present study, C. trachomatis was the leading pathogen detected followed by T. pallidum. Regarding T. pallidum, the prevalence was comparable to a previous study among adolescent pregnant women in rural areas of Mwanza despite population differences implying the same magnitude of T. pallidum across different women population in the City of Mwanza. In comparison to a previous study in the northern region of Tanzania among pregnant women, the reported prevalence in the current study is significantly high. Variations in geographical distribution and risk factors among the study populations may explain observed differences.

Concerning HSV-2 acute infection, seroprevalence reported in the current study is almost similar to the recently published report among pregnant women in the rural areas of Mwanza. In comparison to IgG seroprevalence, the findings are comparable to studies in northern part of Tanzania and rural areas of Mwanza region.[4,21] This could be explained by the fact that, the risk factors in the study areas could be the same. In the contrary, the IgG seroprevalence reported in this study is low compared to 87% reported in Mbeya. The possible explanation could be differences in study population whereby the previous study enrolled bar attendants who were at higher risk of acquiring STIs than the participants in the current study.

Regarding C. trachomatis, in this study, the prevalence was significantly high compared to a similar study done in Uganda which reported a prevalence of 0.9% 5 and low compared to a previous study in the same settings. This could be explained by the fact that the previous study focused on infertile women.

Factors associated with asymptomatic STIs among women opted for IUCD

Among the factors studied, history of dysuria in the last 3 months before seeking IUCD placement was found to predict STIs among women opted for IUCD use. Similar findings were also observed in previous studies in the USA and India whereby a significant proportion of women with different STIs presented with dysuria.[16,24] This could be explained by the fact that dysuria has been reported as one of the commonest symptoms of STIs. History of STIs to the partner was also found to predict STIs which is similar to a previous report from the USA whereby history of partner STIs was linked to C. trachomatis infection. As observed in the current study, most of the STIs were asymptomatic and symptoms occurred when the infection has already passed to the partner. This emphasise the need to consider a policy of laboratory screening for women who opted for IUCD use before insertion of the device in LMICs where these infections are common to avoid unnecessary complications such as tubal factor infertility and chronic PID. The recommended changes are possible in LMICs because most of these infections can be screened by the point of care tests such as those used in the current study. These rapid immunochromatographic tests are less expensive, easy to use (no need of advanced technical skills and equipment) and produce reliable results.

Study limitations

The high prevalence of C. trachomatis could be overestimated due to diagnostic technique used in the current study which is less specific as compared to PCR. In addition, N. gonorrhoea was not tested because the prevalence of this pathogen has been found to be very low in the study setting.

Conclusion and recommendations

The prevalence of STIs is alarmingly high with C. trachomatis being the commonest STI among asymptomatic women opted for IUCD use. There is a paramount need to perform laboratory screening of STIs, especially among women with history of dysuria at least in the past 3 months before seeking placement and partners STIs before IUCD insertion. Further studies to investigate other STIs and outcome of these infections in this population are warranted in this area.