Interpregnancy interval in a rural Guatemalan population: results from a quality improvement database

Abstract

Objective:

This analysis describes the interpregnancy interval (time from livebirth to subsequent conception) in a convenience sample of women living in Southwest Guatemala and the association of antepartum characteristics and postpartum outcomes with a short interpregnancy interval (< 24 months).

Methods:

This is an observational study of a convenience sample of women enrolled in the Madres Sanas community antenatal/postnatal nursing program supported by the Center for Human Development in Southwest Trifinio, Guatemala, between October 1, 2018 and October 1, 2019. We observed the distribution of interpregnancy intervals among the population of women with a reported date of last live birth, and used bivariate comparisons to compare women with a short interpregnancy interval (< 24 months) to those with an optimal interval (≥ 24 months) by antepartum, obstetric and delivery, and postpartum outcomes.

Results:

171 parous women enrolled in the Madres Sanas program between October 1, 2018 and October 1, 2019, and reported the date of their last live birth. One hundred-forty-one (82.5%) women delivered and 130 of those women (92.2%) were seen for their 40-day postpartum visit. The mean interval was 37.1 months with a 22.1-month standard deviation. The median interval was 33.7 months with an interquartile range of 19.6 – 49.5 months. Among these women, 113 (66.1%) the interpregnancy interval was at least 24 months. The only covariate of all sociodemographic, obstetric and antepartum, delivery, and postpartum characteristics that differed between women who achieved an interval (≥ 24 months) compared to those that did not (< 24 months), was age (median 22.9, interquartile range (IQR) [19.1,27.0] vs median 24.8, IQR [21.6,27.9], respectively, p = 0.006). A regression model found that with each increasing year of age, the interpregnancy interval increases by 1.08 months, p = 0.025.

Conclusion:

Among parous women, two-thirds of women space pregnancies at least 24 months. Older women were more likely to have a longer interval between live births.

Synopsis/Precis:

Two thirds of Southwest Trifinian women in rural Guatemala achieve an optimal interpregnancy interval and have a high likelihood of modern postpartum contraceptive uptake.

Introduction:

The World Health Organization (WHO) recommends that after a live birth, women should wait at least 24 months before attempting to conceive again in order to reduce the risk of adverse maternal, perinatal, and infant outcomes.(World Health Organization) In a study conducted in 18 Latin American countries, interpregnancy intervals (from live birth to subsequent conception) less than 12 months were associated with increased risk of low birth weight, preterm birth, small for gestational age, fetal death, and early neonatal death.(Agustin Conde-Agudelo, Rosas-Bermúdez, & Kafury-Goeta, 2006) In Guatemala, prior research suggested that intervals less than 18 months are associated with intrauterine growth restriction of the fetus, and intervals less than 14 months are associated with increased risk of intrapartum and neonatal deaths.(Agustin Conde-Agudelo et al., 2006)

The University of Colorado has founded a health service organization called the Center for Human Development (‘the Center’) in the Southwest Guatemalan lowlands in partnership with a local agribusiness that aimed to improve the health status of the community.(Asturias et al., 2016) The Center supports community-based maternal and child health programming and provides healthcare to workers who experience poverty and poor pregnancy outcomes.(Asturias et al., 2016)^,2 Over the course of pregnancy, quality improvement data on antepartum, intrapartum, and postpartum care of mother and baby is collected by community-health nurses in twelve communities through a program called Madres Sanas.(Asturias et al., 2016)^,2 Women are visited in their homes over the course of pregnancy and postpartum by nurses, and their children are followed for pediatric care through three years of age.(Asturias et al., 2016)^,2 We currently have four planned antepartum and two postpartum visits provided by skilled nurses through the Madres Sanas program with the option for additional unplanned visits.(Asturias et al., 2016)^,2 Education on postpartum contraception is initiated at the first prenatal visit and continued throughout pregnancy. Women are seen for their final maternal visit in their homes around 40 days postpartum but can be enrolled at any point during pregnancy, delivery, or the postpartum period.(Asturias et al., 2016)^,2

Beginning in October 2018, we began collecting data on interpregnancy interval. The primary aim of this secondary analysis was to describe the interpregnancy interval among a convenience sample of women enrolled in the Madres Sanas program. Our secondary aim was to identify antepartum, intrapartum, and postpartum differences between women who achieved a ≥24 month interval as compared to those who had a shorter interval. Finally, if any differences in patient characteristics or outcomes were identified, we aimed to determine which of those were associated with the longer interval.

Methods:

Population:

This study analyzed the prospectively collected quality improvement data documenting interpregnancy interval, demographics, and antepartum, obstetric and delivery, and postpartum characteristics among the convenience sample of women enrolled in Madres Sanas who were enrolled between October 1, 2018 and October 1, 2019. Only women with a history of live birth in their most recent prior pregnancy were included in the cohort.

Ethics Approval:

Individual consent was not obtained for this analysis, but the quality improvement database that is collected by the Madres Sanas nurses has ethics approval from the Colorado Multiple Institutional Review Board (COMIRB # 15-0909).

Outcomes:

This secondary data analysis of a quality improvement database has the primary outcome of interpregnancy interval. At the enrollment visit, women were asked about the date of delivery of their last liveborn infant. These data were used to calculate the interpregnancy interval, which is defined as the time from the prior delivery of a live birth to subsequent conception. The data was entered directly into the REDCap application on tablets, and transmitted securely to password-protected servers at the University of Colorado by the community health nurses.(REDCap, 2019) Date of conception was calculated based on the woman’s reported estimated due date by subtracting 38 weeks from the due date; this calculation assumes a 28-day menstrual cycle with conception on day 14. Data on delivery and postpartum characteristics are collected at the < 72 hour and 40-day postpartum visits, respectively. Due date in this population is calculated by last menstrual period in the majority of women (about 70%), by ultrasound in about 20% of cases, and by clinical exam in the remainder of women. We have been unable to validate gestational age to date, but this is a current focus of the program. Of note, babies are weighed with scales at the 72-hour visit to obtain birthweight.

We compared the population of women with an interpregnancy interval of <24 months to those with an interval of ≥24 months. We aimed to be able to identify demographic (collected at enrollment visit), obstetric (collected at enrollment visit), and intrapartum data (collected at < 72 hour visit) associated with achieving this longer interval as well as compare postpartum outcomes (collected at 40 day visit).

Analysis:

This analysis includes a description of the population of women observed during the study timeframe divided into the population of women who achieved an interpregnancy interval of greater than or equal to 24 months and those whose interval was shorter. In bivariate comparisons we analyzed sociodemographic, antepartum, intrapartum, and postpartum outcomes using Pearson’s chi-squared test for nominal categorical variables unless there was a low cell size, in which case Fisher’s exact test was used. The Kruskal-Wallis test was used for comparison of continuous variables. Variables with a p < 0.05 were considered statistically significant. STATA software version 15.2 (StataCorp LP, College Station, TX, USA) was used for analysis.

Results:

During the period of October 1, 2018 to October 1, 2019, 171 women in the Madres Sanas program who reported a prior live birth were enrolled, as shown in Figure 1. The population is divided into women (n = 58, 33.9%) who experienced a less than 24-month interpregnancy interval (left side of figure) and those who experienced a longer interpregnancy interval on the right (n = 113, 66.1%). In the shorter interpregnancy interval group, 46/58 women (79.3%) had delivered, and 46/46 (100.0 % of those delivered) were seen for their 40-day postpartum visit. In the longer interpregnancy interval group, 95/113 (66.1%) women had delivered, 84/95 (88.4%) of whom had been seen for their 40-day postpartum visit. Figure 2 is a map of the region and shows the relationships of the communities to each other and to the location of the clinic from which the community-based programs are facilitated. The communities are self-defined and do not represent districts or municipalities.

Figure 1.

Study population and where the woman is in her pregnancy course by interpregnancy interval

Figure 2.

Map of Communities in Region

Figure 3 is a histogram of the interpregnancy interval of all 171 women in our study population. The red line at 24 months demarcates the WHO-recommended interpregnancy interval to illustrate the density of the population that achieved a less than or greater than 24-month interpregnancy interval. The distribution is rightskewed, with the mean interpregnancy interval at 37.1 months and the median interval found to be 33.7 months. Table 1 subdivides the cohort by interpregnancy interval into 6-month cohorts. Two women had an interval less than 6 months, 10 women had a repeat pregnancy between 6 – 12 months, 26 women conceived between 12 and 18 months, and 20 women between a year and a half and two years. The remainder of the cohort had at least a 24-month interval between pregnancies.

Figure 3.

Histogram of Interpregnancy Interval Across Women Enrolled in Madres Sanas Program in Months

Table 1.

Interpregnancy Intervals of Cohort

Interpregnancy Interval (months)	N (%)
0 – 6	2 (1.2%)
6 – 12	10 (5.9%)
12 – 18	26 (15.2%)
18 – 24	20 (11.7%)
≥ 24	113 (66.1%)

Table 2 shows characteristics of the total study population in the first column, followed by a bivariate comparison of those women who did not achieve the 24-month interpregnancy interval (second column) compared to women who did space their pregnancies by 24 months (third column). Demographically, the total study population (column 1) was young (median age 24.1, IQR [20.8,27.8]), had been exposed to some level of education (82.9%), were almost all unemployed (97.1%), were often married (95.9%), and had a weekly household income less than or equal to 500 quetzals (81.4%; about $5 USD).

Table 2.

Bivariate comparisons of the association of maternal sociodemographic and obstetric characteristics, as well as pregnancy outcomes, by interpregnancy interval

	Total Population (n = 171)	Interpregnancy Interval ≤ 24 months (n = 58,33.9%)	Interpregnancy Interval > 24 months (n = 113, 66.1%)	P-value
Sociodemographic Characteristics
	0% missing	0% missing
Age in years (median IQR)	24.1 [20.8,27.8]	22.9 [19.1,27.0]	24.8 [21.6,27.9]	0.006 a
	n = 170, 0.6% missing	n = 170, 0.6% missing
Education				0.26b
None	29 (17.1%)	12 (20.7%)	17 (15.2%)
1 – 6 years	108 (63.5%)	37 (63.8%)	71 (63.4%)
7 – 10 years	25 (14.7%)	5 (8.6%)	20 (17.9%)
> 11 year	8 (4.7%)	4 (6.9%)	4 (3.6%)
	n = 170, 0.6% missing	n = 170, 0.6% missing
Employed				0.66b
Yes	5 (2.9%)	1 (1.7%)	4 (3.6%)
No	165 (97.1%)	48 (8.3%)	108 (96.4%)
	0% missing	0% missing
Married				0.43b
Yes	164 (95.9%)	57 (98.3%)	107 (94.7%)
No	7 (4.1%)	1 (1.7%)	6 (5.3%)
	n = 161,5.8% missing	n = 161,5.8% missing
Weekly Income				0.34c
0 – 500	131 (81.4%)	47 (85.5%)	84 (79.3%)
> 501	30 (18.6%)	8 (14.5%)	22 (20.8%)
Obstetric and Antepartum Characteristics
	n = 170, 0.6% missing	n = 170, 0.6% missing
Took Birth Control Before Conception				0.33c
Yes	59 (34.7%)	23 (39.7%)	36 (32.1%)
No	111 (65.3%)	35 (60.3%)	76 (67.9%)
	0% missing	0% missing
Parity				0.10c
1	72 (42.1%)	27 (46.6%)	45 (39.8%)
2	50 (29.2%)	11 (19.0%)	39 (34.5%)
3+	49 (28.7%)	20 (34.4%)	29 (25.7%)
	0% missing	0% missing
Trimester mother entered care				0.11c
First	46 (26.9%)	11 (19.0%)	35 (31.0%)
Second	99 (57.9%)	40 (69.0%)	59 (52.2%)
Third	26 (15.2%)	7 (12.0%)	19 (16.8%)
	0% missing	0% missing
Number of Madres Sanas PNVs*				0.95c
< 4	24 (14.0%)	8 (13.8%)	16 (14.2%)
4+	147 (86.0%)	50 (86.2%)	97 (85.8%)
	0% missing	0% missing
History of Prior Cesarean Birth
0	114 (66.7%)	37 (63.8%)	77 (68.1%)	0.57c
1+	57 (33.3%)	21 (36.2%)	36 (31.9%)
Delivery Characteristics
	Total Population (n = 141)	Interpregnancy Interval ≤ 24 months (n =46, 32.6%)	Interpregnancy Interval > 24 months (n = 95, 67.4%)	P-value
	0% missing	0% missing
Mode of Delivery				0.45c
Vaginal Birth	83 (58.9%)	25 (54.4%)	58 (61.1%)
Cesarean Birth	58 (41.1%)	21 (45.6%)	37 (38.9%)
	n = 139, 1.4% missing	n = 139, 1.4% missing
Location of Delivery				0.78c
Home or Other	49 (35.0%)	15 (33.3%)	34 (35.8%)
Facility (Clinic or Hospital)	91 (65.0%)	30 (66.7%)	61 (64.2%)
	n = 139, 1.4% missing	n = 139, 1.4% missing
Birth Attendant				0.83c
Comadrona (TBA, “unskilled”)	44 (31.7%)	14 (30.4%)	30 (32.3%)
Nurse or Physician (“skilled”)	95 (68.3%)	32 (69.6%)	63 (67.7%)
	n = 134, 5.0% missing	n = 134, 5.0% missing
Birthweight at Delivery				0.82c
≤ 2500g	17 (12.7%)	6 (13.6%)	11 (12.2%)
2500g	117 (87.3%)	38 (86.4%)	79 (87.8%)
Postpartum Characteristics
	Total Population (n = 130)	Interpregnancy Interval ≤ 24 months (n = 46, 35.4%)	Interpregnancy Interval > 24 months (n = 84, 64.6%)	P-value
	0% missing	0% missing
Desired Timeframe Until Next Pregnancy				0.55b
I don’t know	3 (2.3%)	1 (2.2%)	2 (2.3%)
Approximately 3 years	29 (22.3%)	8 (17.4%)	21 (25.0%)
> 3 years	26 (20.0%)	12 (26.1%)	14 (16.7%)
No more children desired	72 (55.4%)	25 (54.3%)	47 (56.0%)
	0% missing	0% missing
Using a Postpartum Contraceptive Method by 40 days Postpartum				0.36c
Yes	55 (42.3%)	17 (37.0%)	38 (45.2%)
No	75 (57.7%)	29 (63.0%)	46 (54.8%)

^*PNVs: prenatal visits, WHO: World Health Organization, TBA: traditional birth attendant

^a:Kruskal-Wallis test

^b:Fisher’s exact test

^c:Chi-squared test

Obstetrically, 71.3% of women had a parity of 1–2 with the remaining women para three or more. The highest proportion of women entered care in their second trimester (57.9%) and were not taking birth control at the time of conception (65.3%). Most were seen for their planned four antepartum visits by the Madres Sanas nurses (86.0%) and 33.3% of women had a history of prior cesarean birth.

Regarding their deliveries, 41.1% of the population gave birth by cesarean, 65.0% delivered in a facility (clinic or hospital), 68.3% were delivered by someone other than a traditional birth attendant (nurse or physician), and the majority of women (87.3%) had a baby with a birthweight of 2500 grams or more.

Of the women who were seen for their 40-day postpartum visit in the study timeframe (n = 130), 55.4% reported that they did not desire any more children, and 42.3% had already initiated a postpartum contraceptive method at the time of evaluation.

In bivariate comparisons (column 4 of Table 1), only age was significantly different between women who had an interval of less than versus greater than or equal to 24 months (median 22.9, IQR [19.1,27.0] vs median 24.8, IQR [21.6,27.9], p = 0.006). Given that no other demographic, obstetric, antepartum, intrapartum, or postpartum characteristics were different between the two groups, we performed a regression of age on interpregnancy interval (Figure 4), and found that for each increasing year of age, the interpregnancy interval increases by 1.08 months (p = 0.025, data not shown).

Figure 4.

Regression Analysis of Relationship of Age and Interpregnancy Interval in Months

Discussion:

Our analysis of interpregnancy interval from a convenience sample of women living in the Southwest Trifinio region of rural Guatemala, based on self-reported data obtained from a quality improvement database, resulted in the following findings: 1) almost two-thirds of women in the community are achieving a 24-month interpregnancy interval (65.5%), 2) in this small sample, only age was associated with the longer interpregnancy interval, with older women being more likely to separate pregnancies by 24 months, and 3) initiation rates of postpartum contraceptive uptake by 40 days postpartum was 47.1% in the population under study, and does not differ between those who had a longer interval than those who had a shorter interval. Historically, women have only had access to condoms, pills, and injectable contraceptives in this community with prior research on barriers to long-acting reversible contraceptives suggesting that lack of spousal approval, difficulty accessing contraceptive methods, lack of knowledge, and fear of adverse side effects contribute to the use of short-acting methods.(Schultz, Larrea, Celada, & Heinrichs, 2018)

Many low- and middle-income country settings experience high fertility rates and poor pregnancy planning/spacing, which is why the WHO has made the proactive recommendation of aiming for a 24-month interval between delivery of one pregnancy and conception of the subsequent pregnancy.(World Health Organization) Interestingly, in the community of interest, it appears that the majority of women surveyed spaced their pregnancies by 24 months (65.5%), which is likely due to the contraceptive educational focus of the Madres Sanas program given the prevalence of malnutrition in the area. We know that a specific aim of the Madres Sanas program, which was initiated in 2011, was to improve maternal and perinatal outcomes through, among other methods, counselling on proper pregnancy spacing.(Asturias et al., 2016; Schultz et al., 2018) Other literature has shown that in similar settings in Latin America, the median interpregnancy interval is 27.8 months, which suggests that proper pregnancy spacing may be preferable in this region.(A. Conde-Agudelo, Belizan, Norton, & Rosas-Bermudez, 2005) The median interval in Guatemala has previously been shown to be 27.1 months.(USAID, 2011) We hypothesize the achievement of a 24-month interpregnancy interval in a majority of women in this community is due to the diligence of the Madres Sanas nurses in educating women about the benefits of proper pregnancy spacing since 2011, which reflects the policy of the WHO that postpartum contraception is an effective method of attaining proper pregnancy spacing, and the availability of condoms, pills, and injectable contraceptives as health posts in the region.(World Health Organization, 2013)

An interesting finding of our analysis is that in our convenience sample, younger women were more likely to have a shorter interpregnancy interval as compared to older women (median 23.1, IQR [19.0,26.9] vs median 24.8, IQR [22.2,27.9], p = 0.005). We hypothesize that age and parity are correlated, and older women may have achieved their desired family size and be more interested in contraceptive use at that time. We do know from some historical data analysis in our community that older women are more likely to use sterilization for contraception. This method is available about an hour away in the closest hospital to the region, but this high barrier to access may account for why older women are more likely to seek definitive contraception. Our finding is novel in comparison to the inverse correlation found between interpregnancy interval and age in data from the United States (US); we hypothesize the US trend may be due to delayed childbearing in the high-income setting.(Haight, Hogue, Raskind-Hood, & Ahrens, 2019; Schummers et al., 2018) Collecting qualitative data to contextualize this finding would be useful to better understand both how and why older women may exhibit a longer interval and the converse in younger women in the community. Younger women may want more children but with a longer interval; this may represent a gap in use of long-acting reversible contraceptives that should be addressed. Another interesting finding that we are also studying is cesarean birth rates in the region and how this contributes to family planning decisions. We have previously noted rising cesarean birth rates in the community, with an overall rate of 41.1% in this cohort.(Harrison et al., 2020) This a significant area of research for our group, and we are planning prospective research on cesarean birth optimization.

We did not identify differences in pregnancy outcomes by interpregnancy interval, which have been previously associated with short interpregnancy intervals (such as birthweight). Our data, probably due to the small sample size, do not therefore support the present recommendation to increase intervals in this community. However, if a quality improvement intervention was to be developed based on our results, focusing on improving pregnancy spacing in the youngest women would be the only supported concept. We do have an ongoing program called “Big Decisions” that is focused on adolescent education regarding sexuality and contraceptive use in local high schools, which, although it is not clinical or health outcomes research, is focused on young women and men in the community with regards to their sexual education. Additionally, other research on neurodevelopmental and nutritional outcomes are underway in this community, so further analyses of how interpregnancy interval is associated with these outcomes in infancy and early childhood would be of interest as an area forfuture research. A focus on pregnancy spacing may be important to improve longer-term maternal and perinatal outcomes as well as a focus on reducing unintended and potentially mistime pregnancies, according to the reproductive preferences expressed by women (Table 1).

This study is limited by the small sample size, the convenience sample study design, the fact that outcomes are obtained by maternal self-report, that we do not have any data on pregnancy intention, and that the study timeframe is short without the availability of temporal trend data. Our null hypothesis finding may be due to the fact that women in the program have low health literacy and are unable to report complications they experienced. We are currently investing in quality improvement efforts to improve data quality both around adverse pregnancy outcomes as well as gestational age dating. Strengths of the study include the fact that the data come from an established quality improvement database that undergoes regular quality assurance reviews to maintain the quality of the data, as well as the fact that the nurses who provide care are the same individuals who collect the data and provide feedback on the utility and interpretability of the data collection forms. Additionally, the data are collected prospectively.

In conclusion, our secondary analysis of a quality improvement database containing data on a convenience sample of women who enrolled in the Madres Sanas antepartum and postpartum care program in the Southwest Trifinio in 2018-2019 found that two-thirds of women achieve the WHO recommended 24-month interpregnancy interval from livebirth to subsequent conception. The only antepartum, intrapartum, or postpartum characteristic associated with greater interpregnancy interval was older maternal age. No adverse outcomes were detected to occur more frequently among women with a shorter interpregnancy interval. Finally, we note that modern contraceptive use of condoms, pills, and injectables in this population by 40 days postpartum is around 50%.