A 3-Year Retrospective Study of Unintended Pregnancy in a Developed Multi-Ethnic Asian Community: A Call for Better Healthcare System for Family Planning

doi:10.21203/rs.3.rs-1293677/v1

Download PDF

Research Article

A 3-Year Retrospective Study of Unintended Pregnancy in a Developed Multi-Ethnic Asian Community: A Call for Better Healthcare System for Family Planning

https://doi.org/10.21203/rs.3.rs-1293677/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background:

Termination of pregnancy poses avoidable risk to women. Women who seek referrals from primary care professionals (PCP) to gynaecologists to manage their unintended pregnancy (UP) are potentially at-risk. The study aimed to determine the prevalence, demographic profiles and associated factors in a developed community requesting for referrals from PCPs for UP.

Methods:

The sociodemographic and clinical data were extracted from the electronic medical records of pregnant multi-ethnic Asian women at eight Singapore public primary care clinics from July 2017 to June 2020. Their variables were computed and compared among women of different age bands using appropriate statistical tests. Logistic regression was used to identify the factors associated with UP referrals.

Results:

Among 9794 gravid women, 974 of them requested gynaecologist referrals to terminate UP over the 3-year period, constituting a prevalence of 9.94%. The mean age of women requesting such referrals was 29.7±7 years. 10.7% of them had more than one prior UP and 15.7% were foreigners. Majority of women were married, did not require social assistance nor had comorbidities. Low proportion of them (2.9%) were prescribed contraceptives. Multivariable logistic regression analysis showed that women of Indian ethnicity, single, aged <20 years and >40 years, were more likely to request referrals for UP.

Conclusions:

One in ten gravid women sought referrals for UP, especially the adolescents and older women, and Indian ethnicity. An accessible, community-based healthcare service to educate and counsel women on family planning is urgently needed to mitigate unnecessary risk from UP and maintain their reproductive health.

Unintended Pregnancy

Family Planning

Since 1960s, birth rates have declined in developed nations (Vollset et al., 2020). The lower birth rates are associated with increased accessibility to contraceptives, lifestyle choices associated with economic affluence, and children becoming an economic drain caused by rising costs to raise children. Higher education and professional careers causing women to have children later in life may lead to a true decline in their fertility levels due to ovarian ageing and reduced chances of conception (Nargund, 2009). Moreover, among women who are pregnant, a proportion of them opt to terminate it, which further reduce the birth rate. These women face increased risks related to the surgical procedure or adverse effects of related medications.

The first step to reduce the risks of these women is to understand the factors associated with unintended pregnancy (UP) in the local community. An unintended pregnancy can result in a woman who has already completed her family, or does not wish to carry on with the pregnancy at that specific time-point, although she may plan to have children in the future. UP is frequently linked to complex social issues, according to Singh et al (2002) and Lim et al (2012), and can impose emotional and socioeconomic burdens on women, their families, and communities (Banerjee et al., 2009). UP among adolescents is a common global health hazard associated with complications such as anemia, stillbirths, preterm deliveries, and low birth weights in subsequent pregnancies (Phupong & Suebnukarn, 2007), (Reime et al., 2008).

Family planning is the cornerstone to prevent UP, and to maintain the well-being of the women and their families. Benefits of family planning include improved child health and development, reduced pregnancy-related health risks for mothers and improved family financial planning (Bailey, 2013). Yet despite the array of contraceptives available, UP continues to occur, accounting for most of the abortions (Mosher et al., 2012). Alarmingly, studies show that more than 120 million people around the world do not use any form of contraception, even though they report that they are sexually active and do not wish to become pregnant (Strauss et al., 2004). It reflects an unmet need for family planning and urgency to address the factors which underpin UP.

Primary care professionals (PCPs) are usually the healthcare providers of family planning services in the local community. Reversible contraceptive methods including oral contraceptive pills (OCP), progesterone injections and intra-uterine devices (IUCD), are usually available in community or primary care clinics in developed nations (Kang et al., 2019). PCPs are also in the frontline to receive women’s requests for gynecologist referral to manage UP if oral therapy is unavailable.

In Singapore, despite national declining birth rate, abortion service is readily available after the implementation of the Abortion Act in 1974. Since then, the number of abortions performed each year has increased, peaking in 1985, when 35% of all pregnancies were terminated. Although the absolute number of abortions has reduced after the introduction of mandatory counselling in 1986, 24.6% of pregnancies were terminated. Today, approximately 12,000 pregnancies are terminated each year. With an average of about 37,000 to 39,000 deliveries a year, this translates to about 1 termination out of 4 pregnancies annually (Singh et al., 1996). Notably, national statistics showed an increase in proportion of abortions in foreign women in Singapore. In primary care where opportunity avails to mitigate UP, the number of local and foreign women seeking PCP referrals for termination of pregnancy remains unknown.

Hence, this SAFE (Surveillance on Abortion and Family planning in primary carE) study aimed to determine the prevalence, sociodemographic characteristics and factors associated with women seeking gynecologist referral to manage UP. Understanding the magnitude of the referrals and the factors influencing these women will provide impetus to design better and person-centric interventions to improve their reproductive health literacy and modify their behavior to reduce UP.

Study sites

The study was a 3-years retrospective electronic medical records’ review of patients with referrals for UP from SingHealth Polyclinics (SHP) (SingHealth Polyclinics, 2021). The institution is a network of eight public fee-for-service primary care clinics (polyclinics) located across the central and eastern parts of Singapore. Located in the Pasir Ris, Tampines, Bedok, Punggol, Sengkang, Marine Parade, Outram and Bukit Merah estates, these polyclinics provide comprehensive primary care services to between 450 and 1000 outpatients across all ages during office hours in each working day (SingHeath Polyclinics, 2021). SHP managed 2.2 million patient visits and rendered over 4.6 million services in 2020. Those who require specialist care will be referred by the polyclinic physicians to the secondary and tertiary hospitals for further management. The services are subsidised for local nationals and permanent residents on a tiered system. Foreigners are required to pay in full for their services at the polyclinics.

Data of study subjects

Data was extracted from the SHP electronic medical record system (Sunrise Clinical Manager®, SCM). The study population included women aged 15–54 years old who visited the 8 polyclinics between 1 July 2017 to 30 June 2020 with ICD-10 diagnosis codes in the EMR for “pregnancy”. Women with a diagnosis code for "termination of pregnancy" are the subjects of interest in this study. Both demographic and clinical data such as comorbidity were also extracted from SCM.

The study population comprises the multi-ethnic Asian women in Singapore, including citizens and permanent residents, plus foreigners such as foreign employees, temporary visitor pass-holders. The majority of Singaporeans are of Han Chinese ethnicity and the minority ethnic groups include Malays, Indians and Others (such as Eurasians).

Comorbidities are based on the Charlson Comorbidity Index (CCI), which is computed from 19 medical conditions (myocardial infarction, congestive cardiac failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic obstructive pulmonary disease, connective tissue disease, peptic ulcer disease, mild liver disease, moderate-severe liver disease, diabetes mellitus, diabetes with end-organ damage, hemiplegia, chronic kidney disease, solid tumour, cancer without metastasis, leukaemia, lymphoma, AIDS).

Data such as nationality, marital status, payment for contraception product and services, social and financial support program enrolment, such as the Community Health Assist Scheme (CHAS) and Medifund status were captured separately in the separate Outpatient Administrative System (OAS) which primarily operationalizes polyclinic business transactions and appointments. Providing information on marital status in OAS by the polyclinic clients is optional. CHAS card holders indicate lower socioeconomic status.

Data processing and audit

A research informatics staff from the Research Department in SHP ensure that the data extraction methodology was sound and extracted the data via Electronic Health Intelligence System (eHINTS). An appointed trusted third party (TTP) from the separate SHP Health Informatics unit assisted in the data de-identification. Next, the TTP pass the de-identified data to the research team for analysis.

Statistical analysis

A key objective of this study was to estimate proportion of referral for UP. A sample size of 9000 is needed to produce 95% Confidence Interval with width of 0.012 and sample proportion of 9% prevalence of referral. Referral indication in the EMR for termination of UP was the primary outcome and was treated as binary data with categories: “referred” and “not-referred”. All demographic and socio-economic characteristics were summarized based on race and age groups for referred patients.

Variables were also summarized based on status of referral. Continuous and categorical variables were described as mean (standard deviation [SD]) and frequency (percentages) respectively. Difference between age group/ race and other continuous and categorical variables were tested using analysis of variance (ANOVA) and Chi-Square test, respectively. All variables were computed based on referral status. To find difference between categorical and continuous variables with respect to referral status was tested using two sample t–test and Chi-Square test respectively. Univariate and multivariable logistic regression models were used to identify factors associated with referral for unintended pregnancies.

The following variables were assessed in the univariate logistic regression models: age, marital status, ethnicity, CHAS status and polyclinic region. Factors with p-value < 0.2 level in univariate analyses were entered into the multivariable model. Stepwise, forward and backward variable selection methods were used to finalize multivariable models. Quantitative association from logistic regression model were expressed as odds ratios (OR) along with 95% CI (95%CI). All tests were two sided at and p value < 0.05 was set as statistical significance. All analyses were carried out in STATA 16.

Ethics approval and funding

The study was supported by the Duke NUS Medical School Academic Medicine Ethos Award for QXES. The SingHealth Centralized Institution Review Board approved the study (CIRB: 2020/3012)

Demographical characteristics of the study population

During the study period, 9794 gravid women comprising women attending their antenatal visit and women requesting referrals for UP presented in 8 polyclinics (Table 1). A total of 974 women (9.9%) requested referrals for UP. Of these women, 105 (10.7%) had records of repeated referrals of unintended pregnancies. The mean age of women referred for UP was 29.6 (SD 7.37) years.

Table 1

Profile of women attending their antenatal visit and women requesting referrals for unintended pregnancy
Variable	Total N = 9794	Status of pregnancy referral		P-value (Chi-Square / T-test)
Variable	Total N = 9794	Intended N = 8820	Unintended N = 974	P-value (Chi-Square / T-test)
Age				0.0043
Mean (SD)	0030.2 (5.37)	0030.3 (5.09)	029.6 (7.37)
Min, Max	0014, 56	0014, 56	016, 51
Age category, n(%)				< 0.0001
< 20 years	0273 (2.8)	0176 (2.0)	097 (10.0)
20–40 years	9213 (94.1)	8421 (95.5)	792 (81.3)
> 40 years	0308 (3.1)	0223 (2.5)	085 (8.7)
Ethnicity, n(%)
Chinese	2926 (29.9)	2623 (29.7)	303 (31.1)	-
Malay	3411 (34.8)	3084 (35.0)	327 (33.6)	0.3080
Indian	639 (6.5)	541 (6.1)	98 (10.1)	0.0003
Others	1074 (11.0)	981 (11.1)	93 (9.5)	0.1119
Foreigners	1744 (17.8)	1591 (18.0)	153 (15.7)	0.0783
Marital status, n(%)				< 0.0001
N#	8016	7204	812
Married	7149 (89.2)	6655 (92.4)	494 (60.9)
Single	0867 (10.8)	0549 (7.6)	318 (39.1)
Nationality, n(%)				0.0713
Singapore Citizen/PR	8050 (82.2)	7229 (82.0)	821 (84.3)
Foreigner	1744 (17.8)	1591 (18.0)	153 (15.7)
Previous TOP History Age				-
N#	105	0	105
Mean (SD)	0025.1 (5.26)	NA	025.1 (5.26)
Median (IQR)	0024.0 (7.0)	NA	024.0 (7.0)
Min, Max	0016, 38	NA	016, 38
Financial Assistance Status, n(%)				0.0148
Yes	2107 (21.5)	1868 (21.2)	239 (24.6)
No	7686 (78.5)	6952 (78.8)	734 (75.4)
Location of Polyclinic, n(%)				0.1688
Eastern	4337 (44.3)	3909 (44.3)	428 (43.9)
North-Eastern	3895 (39.8)	3524 (40.0)	371 (38.1)
South-Eastern	1029 (15.9)	1387 (15.7)	175 (18.0)
Comorbidity*, n(%)				0.0753
No	9360 (95.6)	8440 (95.7)	920 (94.5)	9360 (95.6)
Yes	0434 (4.4)	0380 (4.3)	054 (5.5)	0434 (4.4)
* Comorbidities are selected based on the Charlson Comorbidity Index (CCI).
PR = Permanent Resident, SD = Standard Deviation, IQR = Interquartile Range, TOP = Termination of Pregnancy, CHAS = Community Health Assist Scheme

Ethnicity

The proportions of local Chinese, Malays, Indians, and others (Singaporean and Permanent Residents) in the study population were 31.1%, 33.6%, 10.1% and 9.5% respectively, while 15.7% comprised foreigners from various ethnicities (Bangladeshi, Burmese, Chinese, Eurasian, Filipino, Indian, Indonesian, Japanese, Malay, Nepalese, Pakistani, Sikh, Thai, and Vietnamese). The results show a year-on-year increase of the referrals over the three-year period for the Chinese, Malay and Indian groups with the latter much more so in particular (Fig. 1). From 2017 to 2018, there was a 26.0% and 28.9% increase in the Chinese and Malay groups respectively while there was a 47.6% increase in the Indian group. In the following year, there was a further 33.30% and 4.3% increase in the Chinese and Malay groups respectively while the Indian group saw a 48.4% increase as compared to the previous year. Indeed, Indian gravid women also showed significantly higher odds of requesting UP referral compared to Chinese (OR = 2.09, 95%CI: 1.57–2.78, p < 0.001) (Table 5).

Table 5

Univariate and multivariable logistic regression analysis to find associated risk factors for referral requests for unintended pregnancy.
Variable	Unadjusted Odds Ratio (95% CI)	P-value	Adjusted Odds Ratio (95% CI)	P-value
Age (years)		< 0.0001		< 0.0001+
<20	5.86 (4.526, 7.587)	< 0.0001	1.87 (1.368, 2.558)	< 0.0001
20–40	1.00		1.00
>40	4.05 (3.124, 5.257)	< 0.0001	5.076 (3.726, 6.915)	< 0.0001
Marital status, n(%)		< 0.0001		< 0.0001
Single	7.78 (6.591, 9.181)	< 0.0001	7.59 (6.315, 9.124)	< 0.0001
Married	1.00		1.00
Ethnicity, n(%)		< 0.0001		< 0.0001+
Chinese	1.00		1.00
Malay	0.92 (0.778, 1.082)	0.3080	1.07 (0.883, 1.301)	0.4847
Indian	1.57 (1.226, 2.005)	0.0003	2.09 (1.568, 2.781)	< 0.0001
Others	0.82 (0.643, 1.047)	0.1119	1.06 (0.805, 1.394)	0.6808
Foreigners	0.83 (0.679, 1.021)	0.0783	1.16 (0.905, 1.49)	0.2408
CHAS Status, n(%)
CHAS	1.21 (1.038, 1.415)	0.0149
Non-CHAS	1.00
Polyclinic region		0.1693+
Eastern	0.87 (0.72, 1.045)	0.1355
North-Eastern	0.83 (0.69, 1.009)	0.0620
South-Eastern	1.00
CHAS = Community Health Assist Scheme. + refers type 3 value.

Marital, socioeconomic and clinical status

For marital status, among 812 women, 60.9% were married while single women constituted 39.1%; 23.9% of them were of lower socioeconomic status indicated on their CHAS status (Table 2). The majority (94.5%) of women did not have medical comorbidities (Table 4).

Table 2

sub-analyses of ethnicities of women requesting referrals for unintended pregnancies
Demographic variables N(%)	Chinese N = 303	Malay N = 327	Indian N = 98	Others N = 93	Foreigners* N = 153	Total N = 974	P-value (Chi-Square / ANOVA
Age (years)							< 0.0001
Mean (SD)	29.4 (8.1)	028.3 (6.7)	29.9 (7.0)	28.9 (6.8)	033.5 (6.5)	029.7 (7.4)
Min, Max	016, 48	017, 51	17, 44	17, 45	016, 46	016, 51
Marital status, n(%)							< 0.0001
N#	252	277	85	86	112	812
Married	124 (49.2)	172 (62.1)	48 (56.5)	52 (60.5)	98 (87.5)	494 (60.8)
Single	128 (50.8)	105 (37.9)	37 (43.5)	34 (39.5)	14 (12.5)	318 (39.2)
CHAS Status, n(%)							< 0.0001
CHAS	069 (22.8)	113 (34.6)	27 (27.6)	24 (25.8)	000	233 (23.9)
Non-CHAS	234 (77.2)	214 (65.4)	71 (72.4)	69 (74.2)	153 (100 )	741 (76.1)
Polyclinic region†, n(%)							< 0.0001
Eastern	115 (38.0)	179 (54.7)	35 (35.7)	46 (49.5)	050 (32.7)	425 (43.6)
North-Eastern	132 (43.6)	097 (29.7)	40 (40.8)	32 (34.4)	073 (47.7)	374 (38.4)
South-Eastern	056 (18.5)	051 (15.6)	23 (23.5)	15 (16.1)	030 (19.6)	175 (18.0)
* Foreigners were analysed as a group in comparison with local multi-ethnic Asian women † Polyclinics in Eastern region (Pasir Ris, Tampines, Bedok), North-Eastern region (Seng Kang, Punggol), South-Eastern region of Singapore (Bukit Merah, Outram, Marine Parade). PR = Permanent resident, SD = Standard Deviation, IQR = Interquartile Range, CHAS = Community Health Assist Scheme, ANOVA = Analysis of Variance.

Table 4

Sub-analyses of women across three age groups requesting referrals for unintended pregnancies
Variable	< 20 years N = 89	20–40 years N = 800	> 40 years N = 85	Total N = 974	P-value
Nationality, n(%)					< 0.0001
Singapore Citizen/PR	85 (95.5)	675 (84.4)	61 (71.8)	821 (84.3)
Foreigner	04 (4.5)	125 (15.6)	24 (28.2)	153 (15.7)
Marital status, n(%)					< 0.0001
N#	81	664	67	812
Married	6 (7.4)	427 (64.3)	61 (91.0)	494 (60.8)
Single	75 (92.6)	237 (35.7)	06 (9.0)	318 (39.2)
Ethnicity, n(%)					< 0.0001
Chinese	43 (48.3)	224 (28.0)	36 (42.4)	303 (31.1)
Malay	29 (32.6)	286 (35.8)	12 (14.1)	327 (33.6)
Indian	04 (4.5)	085 (10.6)	09 (10.6)	098 (10.1)
Others	09 (10.1)	080 (10.0)	04 (4.7)	093 (9.5)
Foreigner	04 (4.5)	125 (15.6)	24 (28.2)	153 (15.7)
Comorbidity, n(%)					0.0002
No	85 (95.5)	763 (95.4)	72 (84.7)	920 (94.5)
Yes	04 (4.5)	037 (4.6)	13 (15.3)	054 (5.5)
PR = Permanent resident. P – values are based on Chi-Square test.

Geographical variation

Comparing the referral variations across the 8 polyclinics, those located in the Eastern region of Singapore (Pasir Ris, Tampines, Bedok) received the highest proportion of women requesting UP referrals (43.6%), followed by polyclinics in the North-eastern region (Seng Kang, Punggol) (38.4%), and South-eastern region of Singapore (Bukit Merah, Marine Parade, Outram) (18.0%) (Table 2).

Contraception usage

In the referred group of women, 946 (97.1%) of them were not on any form of contraceptives dispensed from the polyclinic pharmacies. Of the 28 (2.9%) women on family planning, 23 were prescribed oral contraceptive pills (OCP), and 5 had inserted intrauterine contraceptive devices (IUCD). Following the referral for UP, 12 of these 28 women continued family planning (7 were prescribed OCP and 5 had IUCD) (Fig. 3).

Sub-analysis of ethnicity of women who were referred for unintended pregnancies

Foreign women referred for UP were analysed as a minority group in the study population. Demographical variables that were significantly associated with ethnicity include age, marital status, socioeconomic status (CHAS), and site of polyclinic consultation and referral (Table 2). The foreign women were older with a mean age of 33.5 years. Pairwise age comparison between various ethnic groups revealed that foreign women were significantly older than local women (Table 3).

Table 3

Pairwise age comparison between ethnic groups of women requesting referrals for unintended pregnancies (P-value)
	Chinese	Foreigners	Indian	Malay	Others
Chinese		< 0.0001	0.5335	0.0567	0.5352
Foreigners	< .0001		0.0001	< 0.0001	< 0.0001
Indian	0.5335	0.0001		0.0515	0.3137
Malay	0.0567	< 0.0001	0.0515		0.5039
Others	0.5352	< 0.0001	0.3137	0.5039
Note: Age as a continuous variable.

Over half of the local Chinese (50.8%) were single women when they were referred for UP. More women of Malay ethnicity who were referred were CHAS cardholders (34.6%) and sought referrals (54.7%) from polyclinics located in the eastern region (Pasir Ris, Tampines, Bedok). Higher proportions of local Chinese (43.6%), Indian (40.8%), and foreign women (47.7%) sought referrals from polyclinics in the North-eastern region of Singapore (Seng Kang, Punggol).

Sub-analysis of referred women across the age groups

Among referred women, 89 (9.1%) of them were < 20 years of age, 800 (82.1%) women were aged between 20–40 years, and 85 (8.7%) were > 40 years of age (Table 4). Variables that were significantly associated with age include nationality, marital status, ethnicity, and medical comorbidities. For marital status, higher proportion of women < 20 years old were singles (92.6%), while women between 20–40 years old (64.3%) and those > 40 years old were married (91%). Proportionately more referred local Chinese women (48.3%) were younger (< 20 years old) and 42.4% of them were older (> 40 years old) compared to those aged between 20–40 years (28%). Among women with medical comorbidities, most were older women (15.3%) aged > 40 years old compared with 4.5% and 4.6% in the two younger age groups respectively.

Risk factors for requesting referral for unintended pregnancy

Univariate logistic regression analysis revealed that age, marital status, ethnicity, and CHAS status were significantly associated with referral for UP (Table 5). Women < 20 years old (OR = 5.86, 95%CI = 4.53–7.59) and > 40 years old (OR = 4.05, 95%CI = 3.12–5.26) showed higher odds of requesting referrals for UP. Single women (OR = 7.78, 95%CI = 6.59–9.18), local Indian (OR = 1.57, 95%CI = 1.27–2.01) and lower socioeconomic status (CHAS cardholders) (OR = 1.21, 95%CI = 1.04–1.42) showed higher odds of requesting referrals for unintended pregnancy.

Multivariable logistic regression analysis revealed that age, marital status, and ethnicity were significantly associated with referral for UP (Table 5). Women < 20 years old (OR = 1.87, 95%CI = 1.37–2.56) and > 40 years old (OR = 5.08, 95%CI = 3.73–6.92) showed increased odds of UP referrals when compared to women between 20–40 years of age. Single women (OR = 7.59, 95%CI = 6.32–9.12), local women of Indian ethnicity (OR = 2.09, 95%CI = 1.57–2.78) were associated with increased odds of referrals for UP.

To our knowledge, this is the first study in Singapore to determine the prevalence and demographic profiles of multi-ethnic Asian women seeking referrals from public primary care clinics to specialists for UP. The referral prevalence for UP over a 3-year period was 9.9%. Hence, one in ten gravid women sought to terminate their pregnancies, subjecting themselves to morbidity and even mortality risks. About 10–50% of women developed life-threatening complications and long-term consequences such as infection and secondary subfertility due to abortions (World Health Organization, 2008).

Among these 974 women, 10.7% had prior referral for UP. The prevalence is higher than the 5.3% of women presented with repeated abortions reported by a local tertiary hospital from 1996 to 2000, and 2005 to 2009 (Lim et al., 2012), (Banerjee et al., 2009). The high prevalence of repeated unintended pregnancies in our study is of concern. A more effective care model beyond the current process of mandatory abortion counselling and existing family planning services is needed to reduce repeated abortions. A Danish study has revealed that women seeking a third abortion tended to use less efficient contraceptive methods or none at all than those who had 1st or 2nd termination of pregnancy (Osler et al., 1997). In contrast, the risks of repeated abortions decreased when IUCD or sterilization were used as contraception (Miinmaki et al., 2009). IUCD (Fig. 2) is available in the polyclinics but its uptake at the polyclinics is low (4.2%), despite its high efficacy, low side effects, high user satisfaction and cost-effectiveness of this long-acting contraception (Oakley, 1994). Gosavi et al. reported that local women at a tertiary hospital had poor awareness and knowledge of contraception, including long-acting reversible contraceptive (LARC) which minimally interrupt their life.¹⁸ In addition, the uptake and perspectives of local women toward sterilization as a permanent contraceptive procedure carried out in hospital is little known and merits further research.

Contraception is key to reduce UP. The vast majority (97.1%) of referred women did not have any record of taking contraception themselves. Singapore has a fee-for-service primary healthcare system, in which the local residents can seek consultation with multiple care providers. Hence, they could have obtained OCP from other healthcare providers such as private general practitioners. Their spouse or partner could have used barrier method in family planning, which would not be documented in their electronic medical records. Apparent failed contraception is reflected by the 16 referred women who had prior OCP based on their prescription records. 5 out of 12 women was prescribed IUCD after referral as a cost-effective and satisfactory form of contraception as mentioned above.

Among the 28 women who were prescribed contraceptives, 3 were young women below 20 years of age and 195 were above 40 years. A local 2018 survey reported poor knowledge about contraceptives among local youths (Association of Women for Action and Research, 2021). About 60% of those aged 18 years and below who engaged in sexual activity did not take any precautions to avoid pregnancy or sexually transmitted disease. They tended to seek sexual health information from friends and partners. Parents were ranked as a least popular portal of communication due to scepticism of their openness for this topic discussion. A local gender equality community organisation is testing a pilot program “Birds & Bees” to coach parents who intended to initiate conversations about sexual matters at home but the results have yet to be reported (Association of Women for Action and Research, 2021).

The older women may misperceive declined fertility after the age of 40 years and reduced alert on contraception may result in UP (Godfrey et al., 2011). LARC such as IUCD, implant and injectable contraceptives are suitable methods to avoid UP. LARC recommendations should ideally be shared with the gravid mothers during the antenatal and immediate postnatal period when family planning is likely to be within their concern and priority.

Local Indian women were significantly associated with higher likelihood (AOR = 2.09, p < 0.001) of UP referrals (Table 5). In a recent publication by Sheila Desai et al in New York City, among country of origin groups, Indian women had the highest rate of abortions compared to other Asians (Sheila et al., 2021). In India, negative attitudes towards female babies persist, where people appear to prefer a son at any cost and female foeticide remains prevalent (Gupta, 2017). Antenatal ultrasound assessment is routinely performed in Singapore which will reveal the gender of the foetus to the parents. The effect of the awareness of the foetal gender among local Indian women on their decision to continue or terminate their pregnancy remains unknown.

Single women comprised 39.2% of the study population and were more likely to request UP referrals (OR = 7.59) than married women. Many younger women (92.6%) were singles (Table 4). Their relationship with their partners may be less stable, resulting in UP arising from unprotected sexual intercourse. Due to increasing opportunities for education, employment and career development, young and unmarried women may have lower threshold for UP.

Foreigners comprised 15.7% of the women referred for UP. The Ministry of Health (2013) had reported that foreigners constituted 26.5% of abortions in 2012, almost doubled from 13.5% in 2003. More than 30 percent of the total population of Singapore are currently foreigners (Popul.²⁶ Any rising trend in abortions among foreigners is disconcerting to local health authority. Literature alludes to the more severe negative health effects of migration on women than men (Adanu & Johnson, 2009). Contraceptive use by foreign women can be constrained by various factors including socio-economic status, language barriers, poor health literacy, lack of healthcare access and awareness of the health services of the host country (Karnjuš et al., 2009), (Shah et al., 2018). Their risks of UP are elevated due to these barriers (Rechel et al., 2011).

Our study has important limitations. The retrospective data report referrals by gravid women for abortion in a cluster of polyclinics, but do not reflect the actualization of the procedure at the tertiary hospitals. The results are likely to be under-estimated as referral data from private general practitioners or walk-in to private obstetricians are unavailable. The results do not show causality nor provide details for the UP, including the psychosocial context of the decision. The data is based on coding, which is operator-dependent and potentially subjected to errors. The use of condoms (Fig. 2) is prevalent, being a widely used contraception which is readily available at community pharmacies and convenience stores beyond the polyclinics but is not recorded in the polyclinic database.

Implications for Practice and/or Policy

Polyclinics are sited within 30 minutes of travel by public transport in residents within the respective estates. They are well-positioned to offer accessible and affordable family planning services to the women of reproductive age. Polyclinics sited in the Eastern and North-Eastern regions of Singapore attended to more women seeking UP referrals due to higher density of younger families residing in the newer public housing estates in these localities. This finding suggests a review of polyclinic capacity to meet the contraception needs of the women due to varying geographical hotspots on the island-state. Timely and effective intervention to the approximately 1 in 10 gravid women visiting the polyclinics for UP can potentially deter the next UP. Efforts should be directed to upstream preventive measures when the women consult the polyclinics for their post-natal care, when the topic of family planning can be delivered by the PCPs. Nonetheless, Poon Z et al have revealed gaps and challenges faced by PCPs in optimizing postnatal care and offering contraception advice to women in the polyclinics. A trial is being planned to assess the feasibility of operationalizing telemedicine-based postnatal care and contraception service at the polyclinics and its acceptability by women in the community. It will be piloted in polyclinics, which are hotspots for UP referrals at the eastern region of the island state.

Almost one in ten women sought gynaecologist referral in polyclinics to manage UP. About 10% of these women had repeated abortions. The associated factors include age, such as those younger than 20 years or older than 40 years, singlehood, Indian ethnicity and lower socioeconomic status. A new care model to deliver effective family planning advices to these women is being planned and aims to reduce UP and maintain their physical and reproductive health.

Ethics approval and consent to participate. ((Human, Animals, Plants and Source) – The SingHealth Centralized Institution Review Board approved the study (CIRB: 2020/3012).

a. Ethical approval with full name of ethics committee. SingHealth Centralized Institution Review Board, A/Prof Agnes Ng, Chairman of Singhealth Centralised Institutional Review Board E approved it. The rest of the ethics committee was not involved as this was a retrospective de-identified data analysis determination for the IRB.

b. Institutional/national/international guidelines followed for the study. Yes

c. Statement on informed consent from the participants/legal guardians (if participants are below age 18) for involvement in the study. Study is retrospective and data obtained from electronic medical records are de-identified before commencing with analysis.

IRB Determination will be uploaded under the section Related Files in your portal. As mentioned by the SingHealth Centralised Institutional Review Board, a clinical database (individually identifiable) can be used for patient care and management purposes but not for research as consent for research has not been obtained. No IRB approval is required for use in patient care and management. If there is an intent to use the clinical database (individually identifiable) for research, application to the IRB with consent from the subjects OR application for a waiver of consent in order to use the database in the individually identifiable manner (High Bar waiver criteria must be met if data was obtained or compiled after 1 Nov 2017). Alternatively, the third party de-identification process can be used to render the database non-identifiable to be used in research.

Consent for publication – Not Applicable

Availability of data and materials – The sociodemographic and clinical data were extracted from the electronic medical records of pregnant multi-ethnic Asian women at eight Singapore public primary care clinics from July 2017 to June 2020.

Competing interests - None

Funding:

This work was supported by Duke-NUS Medical School Academic Medicine Ethos Award for QXES [grant number AM-ETHOS01/FY2020C2/35-A69]. The SingHealth Centralized Institution Review Board approved the study (CIRB: 2020/3012).

Authors' contributions – Quak XES conducted the study, analysed the data and drafted the paper. Sultana R analysed the data. Aau WK extracted data from the electronic medical record system. Goh CC interpreted the results on the oral contraceptive use by women in the polyclinics. Tan NC interpreted and discussed the results and co-drafted the manuscript. All authors reviewed the draft and approved the manuscript.

Acknowledgements – We would like to thank Winnie Ong and Winnie Seah from SingHealth Polyclinics and Family Medicine Academic Clinical Programme, for their administrative support and institution review board.

Authors' information –

Quak Xin En Stephanie [Quak XES]

The first author is an aspiring medical student from Duke NUS Medical School who is supported by the SingHealth-Duke NUS Academic Medicine Enhancing Training, Healthcare, Outcomes & Standards Award to conduct the study.

Rehena Sultana [Sultana R]

Rehena is a biostatistician at the Center for Quantitative Medicine in Duke-NUS Medical School, Singapore.

Aau Wai Keong

Wai Keong is Assistant Manager cum Research Informatics Officer at the Department of Research in SingHealth Polyclinics, Singapore.

Goh Chin Chin

Chin Chin is a Pharmacy Practice Manager at SingHealth Polyclinics.

Tan Ngiap Chuan [Tan NC]

Clinical Associate Professor (Dr) Tan Ngiap Chuan is Academic Vice Chair, Research, at SingHealth–Duke NUS Family Medicine Academic Clinical Programme, as well as Director of Department of Research at SingHealth Polyclinics. He is also a practicing family physician cum senior consultant who looks after patients across all age groups in the polyclinic, including pregnant women and post-natal mothers.

Vollset, S. E., Goren, E., Yuan, C. W., et al (2020). Fertility, mortality, migration, and population scenarios for 195 countries and territories from 2017 to 2100: a forecasting analysis for the Global Burden of Disease Study. The Lancet, 396(10258):1285–306.
Nargund, G. (2009). Declining birth rate in Developed Countries: A radical policy re-think is required. Facts, views & vision in ObGyn, 1(3):191.
Singh, K., Fong, Y. F., Loh, S. Y. (2002). Profile of women presenting for abortions in Singapore at the National University Hospital. Contraception, 66:41–6.
Lim, L., Wong, H., Yong, E., Singh, K. (2012). Profiles of women presenting for abortions in Singapore: focus on teenage abortions and late abortions. Eur J Obstet Gynecol Reprod Biol, 160:219–22.
Banerjee, B., Pandey, G., Dutt D., Sengupta, B., Mondal, M., Deb, S. (2009). Teenage pregnancy: a socially inflicted health hazard. Indian J Community Med, 34:227–31.
Phupong, V., & Suebnukarn, K. (2007). Obstetric outcomes in nulliparous young adolescents. Southeast Asian J Trop Med Public Health, 38:141–5.
Reime, B., Schu ̈cking, B. A., Wenzlaff, P. (2008). Reproductive outcomes in adolescents who had a previous birth or an induced abortion compared to adolescents’ first pregnancies. BMC Pregnancy Childbirth, 8:4.
Bailey, M. J. (2013). Fifty years of family planning: new evidence on the long-run effects of increasing access to contraception. National Bureau of Economic Research, Oct 4.
Mosher, W. D., Jones, J., Abma, J. C. (2012). Intended and unintended births in the United States: 1982–2010. Natl Health Stat Report, 55:1.
Strauss, III J. F., Nass, S. J., editors (2004). New frontiers in contraceptive research: a blueprint for action. National Academies Press, Apr 12.
Kang, M., Guetterman, T., Prussack, J., Ursu, A., Wu, J. (2019). Contraceptive Care for Women With Medical Conditions: A Qualitative Study to Identify Potential Best Practices for Primary Care Physicians. Family medicine, 51(7):559–66.
Singh, K., Prasad, R. N. V., Ratnam, S. S. (1996). Abortion trends in Singapore: a 25-year review. J Pediatr Adoles Gynecol, 9:27–30.
SingHealth Polyclinics. (2021). https://polyclinic.singhealth.com.sg
Word Health Organization. (2012). Unsafe abortion incidence and mortality: global and regional levels in 2008 and trends during 1990–2008. (No. WHO/RHR/12.01). World Health Organization.
Osler, M., David, H. P., Morgall, J. M. (1997). Multiple induced abortions: Danish experience. Patient Educ Couns, 31:83–1
Miinmaki, M., Pouta, A., Bioigu, A., et al. (2009). Frequency and risk factors for repeated abortions after surgical compared with medical termination of pregnancy. Obstetrics & Gynecology, 13(4):845–52. https://doi.org/10.1097/AOG.ob013e3819cae06
Oakley, D. (1994). Rethinking patient counselling techniques for changing contraceptive behavior. Am J Obstet Gynecol, 170:1585–1590.
Gosavi, A., Ma, Y., Wong, H., & Singh, K. (2016). Knowledge and factors determining choice of contraception among Singaporean women. Singapore medical journal, 57(11), 610.
Association of Women for Action and Research (AWARE), Singapore. (2021). Many young people unsure of contraceptive use, and feel they can’t talk to parents about sexual health. https://www.aware.org.sg/2018/11/many-young-people-unsure-of-contraceptive-use-and-feel-they-cant-talk-to-parents-about-sexual-health/
Association of Women for Action and Research (AWARE), Singapore. (2021). Birds & Bees: A Workshop for Parents. https://www.aware.org.sg/birds-and-bees/
Godfrey, E. M., Chin, N. P., Fielding, S. L., Fiscella, K., Dozier, A. (2011). Contraceptive methods and use by women aged 35 and over: A qualitative study of perspectives. BMC women's health, Dec;11(1):1–9.
Sheila, D., Mary, H., Heidi, E. J. (2021). Differences in abortion rates between Asian populations by country of origin an nativity status in New York City, 2011–2015. Int J Environ Res Public Health, 18(2):6182. https://doi.org/10.3390/ijerph18126182
Gupta, A. (2017). Female foeticide in India. UNICEF. http://unicef.in/PressRelease/227/Female-foeticide-in-India
Ministry of Health, Singapore. Abortion statistics. http://www.moh.gov.sg/content/moh_web/home/pressRoom/Parliamentary_QA/2013/abortion-statistics.html
National Population and Talent Division. (2021). Population in Brief 2021. https://www.population.gov.sg/files/media-centre/publications/Population-in-brief-2021.pdf
Adanu, R. M. & Johnson, T. R. (2009). Migration and women's health. Int J Gynaecol Obstet, Aug;106(2):179–81. https://doi.org/10.1016/j.ijgo.2009.03.036
Karnjuš, I., Prosen, M., Bogataj, U., et al. (2009). Use of Modern Contraceptive Methods among Migrant Women Living in Slovenia (p. 81). Women, Migrations and Health.
Shah, R., Kiriya, J., Shibanuma, A., Jimba, M. (2018). Use of modern contraceptive methods and its association with QOL among Nepalese female migrants living in Japan. PloS one, May 16;13(5):e0197243.
Rechel, B., Mladovsky, P., Devillé, W., Rijks, B., Petrova-Benedict, R., McKee, M. (2011). Migration and health in the European Union: an introduction. Maidenhead: Open University Press.
Poon, Z., Lee, E. C. W., Ang, L. P., `, N. C. (2021). Experiences of primary care physicians managing postpartum care: a qualitative research study. BMC Fam Pract, Jun 30;22(1):139. https://doi.org/10.1186/s12875-021-01494-w

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

A 3-Year Retrospective Study of Unintended Pregnancy in a Developed Multi-Ethnic Asian Community: A Call for Better Healthcare System for Family Planning

Status:

Version 1

Abstract

Figures

Introduction

Materials And Methods

Study sites

Data of study subjects

Data processing and audit

Statistical analysis

Ethics approval and funding

Results

Demographical characteristics of the study population

Ethnicity

Marital, socioeconomic and clinical status

Geographical variation

Contraception usage

Sub-analysis of ethnicity of women who were referred for unintended pregnancies

Sub-analysis of referred women across the age groups

Risk factors for requesting referral for unintended pregnancy

Discussion

Implications for Practice and/or Policy

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1