Keywords
migration - antenatal care - Vietnam - perinatal data - refugee
Introduction and Background
Introduction and Background
According to the 2023 micro-census conducted by the German Federal Statistical Office,
about 215000 people of Vietnamese descent live in Germany, 136000 of whom have a personal
experience of migration (the so-called first generation) [1]. This Vietnamese diaspora is not spread evenly across Germany; about 20000 people
of Vietnamese descent live in Berlin alone. The percentage of ethnic Vietnamese women
presenting to the Charité Maternity Hospital in the district Berlin-Mitte is even
higher, at around 7% of pregnant women [2].
The migration background of ethnic Vietnamese women in the old Federal Republic of
Germany and former West-Berlin is, in most cases, a consequence of the Vietnam war
which was waged between North Vietnam and South Vietnam from 1955 to 1975. After the
victory of communist North Vietnam, more than one million people fled from Vietnam
across the South China Sea. Around 38000 of these so-called Boat People were taken
in by the Federal Republic of Germany between 1978 and the mid-1980s, most of them
from South Vietnam [3]. Ethnic Vietnamese women from the new German federal states and East Berlin have
a different immigration history; their parents or other family members were usually
recruited as contract workers from North Vietnam by the government of the GDR [4]. The social background of pregnant women of Vietnamese origin with a registered
address in a refugee shelter is
not clear, especially because in Germany, granting asylum to people from Vietnam is
currently limited to exceptional cases [5].
The aim of this retrospective study was to compare antenatal and obstetric care and
the birth outcomes of women with and women without a Vietnamese migration background.
To estimate the impact of a special, socially stressful situation, in a second step
we differentiated between women of Vietnamese origin with and those without a registered
address in a refugee shelter. Up to now, the care provided to Vietnamese women with
a recent history as refugees was neglected in publications on the medical care provided
to immigrants, and our data collection, which included refugee women, is probably
the first for German-speaking countries.
For the primary combined endpoint for the pregnancy outcome, we chose what we termed
an “ideal normal birth outcome,” which we defined as spontaneous or operative vaginal
delivery of a singleton pregnancy from week 37 + 0 of gestation onwards, no higher
degree perineal tear, a 5-minute Apgar score ≥ 8, an arterial cord blood pH of 7.20,
a peripartum blood loss < 500 ml, and no transfer of the newborn to a neonatal department.
Methods
Study population
In the register of births of the maternity hospital of Charité University Hospital,
Campus Mitte, we identified all patients who gave birth between 1.1.2016 and 31.12.2019
with a Vietnamese first name or surname. The registered addresses of all women were
checked and a registered accommodation in a refugee shelter was recorded.
The women of Vietnamese origin were compared in a 1:3 ratio with women of the same
age and parity and non-Vietnamese names who had given birth in the same hospital in
the same period.
The non-Vietnamese women were randomly selected from our patient population; their
nationality or ethnicity was not recorded. Women below the age of 18 and women for
whom obstetric data was lacking were excluded from our evaluation. The perinatal data
of all patients were obtained from the ViewPoint 5 documentation system (Solingen,
Germany) and the digital SAP patient file (Walldorf, Germany).
Once each patient had been classified as either of Vietnamese (V) and non-Vietnamese
(NV) origin, the patient data were anonymized.
Statistical analysis
All calculations were carried out with SPSS (Statistical Package for Social Sciences
23, IBM Corp., Armonk, NY, USA).
Descriptive analysis was carried out for continuous and categorical variables, with
continuous variables presented as mean and standard deviation (± SD) and frequency
reported in percent (%) and missing values. A non-normal distribution was assumed
for all continuous variables, and Mann-Whitney U-test was used for analysis; similarly,
Fisher’s exact test was used for categorical variables. The two-sided significance
level was set to 0.05. After the first part of the study, which compared two groups
of patients either with or without Vietnamese ethnicity and compared Vietnamese migrants
with and those without a registered address in a refugee shelter, multivariable regression
analysis was carried out to evaluate the impact of different clinical parameters on
caesarean section as the mode of delivery, transfer of the neonate to a neonatal department,
and incidence of higher degree (3rd or 4th degree) perineal tears. To do this, the
relevant influencing factors were
identified and the corresponding odds ratios (ORs) and confidence intervals (95% CI)
were calculated using a multivariable generalized linear model (GLM).
Ethical approval and data protection
An application for ethical approval (EA2/24/20) was filed prior to the retrospective
collection and analysis of data, and the study was approved by the Ethics Committee
of the Charité. The guidelines of the Charité on good clinical practice were observed
and the Berlin Data Protection Act was complied with.
Results
Patient characteristics and antenatal care
A total of 534 patients with a Vietnamese first name and surname presented to the
Maternity Hospital in Campus Mitte of the Charité between 1 January 2016 and 31 December
2019, which corresponds to 6.7% (with monthly fluctuations of between 2.2 and 15.0%)
of the total number of births. The perinatal data of 64 patients were incomplete with
regards to the birth outcome, and these patients were excluded from the study. The
comparison group consisted of 1410 patients of non-Vietnamese origin, of whom 11 women
(0.8%) were registered as living in a refugee shelter. In the group of ethnically
Vietnamese women, this figure was 209 (44.5%).
[Table 1] shows the characteristics of the patient population with and without a Vietnamese
migration background. The physical characteristics differed between groups, with a
mean height of 156 cm recorded for one group of patients (V) compared to 166 cm for
the other group (NV), p < 0.001. The mean BMI of Vietnamese women was lower (21 [V]
vs. 24 kg/m2 [NV], p < 0.001) and the percentage of underweight women (BMI < 18.5 kg/m2) was significantly higher in the group of Vietnamese women (19.9%) compared to non-Vietnamese
women (6.4%, p < 0.001). Only 1.1% of Vietnamese women were overweight compared to
10.9% of NV women (p < 0.001).
Table 1
Basic characteristics of the patient groups of Vietnamese and non-Vietnamese origin
including the number of cases; Mann-Whitney U-test was used for nominal variables
and two-sided Fisher’s exact test for categorical variables (level of significance
p = 0.05).
|
|
Vietnamese migration background
|
n
|
No Vietnamese migration background
|
n
|
P
|
|
Maternal age (years)
|
28.14 ± 5.40
|
470
|
28.13 ± 5.36
|
1410
|
0.996
|
|
Height (cm)
|
156.38 ± 4.85
|
470
|
165.74 ± 6.85
|
1410
|
< 0.001*
|
|
Body mass index (kg/m2)
|
20.70 ± 2.69
|
462
|
23.96 ± 4.98
|
1387
|
< 0.001*
|
|
|
92 (19.9%)
|
|
89 (6.4%)
|
|
< 0.001*
|
|
|
365 (79.0%)
|
|
1151 (83%)
|
|
< 0.001*
|
|
|
5 (1.1%)
|
|
147 (10.6%)
|
|
< 0.001*
|
|
Parity
|
1.77 ± 0.91
|
470
|
1.66 ± 0.78
|
1410
|
1.0
|
|
Gravidity
|
2.0 ± 1.13
|
|
2.0 ± 1.13
|
|
1.0
|
|
Primiparous
|
203 (48.1%)
|
|
701 (49.7%)
|
|
0.558
|
|
First trimester screening in week 11 + 0–13 + 0 of gestation
|
23 (12.2%)
|
189
|
280 (26.9%)
|
1040
|
< 0.001*
|
|
Detailed fetal scan in week 18 + 0–22 + 0 of gestation
|
103 (54.5%)
|
189
|
834 (80.2%)
|
1040
|
< 0.001*
|
|
Number of antenatal care visits
|
8.38 ± 3.0
|
455
|
10.54 ± 3.21
|
1284
|
< 0.001*
|
|
|
49 (10.8%)
|
|
47 (3.7%)
|
|
< 0.001*
|
|
|
121 (26.6%)
|
|
148 (11.5%)
|
|
< 0.001*
|
|
|
224 (49.2%)
|
|
627 (48.8%)
|
|
0.913
|
|
|
61 (13.4%)
|
|
462 (36%)
|
|
< 0.001*
|
|
First antenatal care appointment
|
|
462
|
|
1410
|
|
|
|
122 (26.4%)
|
|
796 (59.8%)
|
|
< 0.001*
|
|
|
221 (47.8%)
|
|
466 (35.0%)
|
|
< 0.001*
|
|
|
119 (25.8%)
|
|
69 (5.2%)
|
|
< 0.001*
|
|
HBsAg positive
|
30 (6.4%)
|
470
|
6 (0.4%)
|
1347
|
< 0.001*
|
|
Gestational diabetes
|
|
469
|
|
1343
|
|
|
|
63 (13.43%)
|
|
128 (9.53%)
|
|
0.013*
|
|
|
118 (25.16%)
|
|
199 (14.81%)
|
|
< 0.001*
|
|
|
11 (17.46%)
|
63
|
19 (14.84%)
|
128
|
0.675
|
When we looked at the number of antenatal care appointments, it was clear that women
of Vietnamese origin attended fewer appointments overall and that just under ¾ of
these appointments only began after the end of the 1st trimester of pregnancy, whereas
60% of all non-Vietnamese patients already had their first antenatal examination before
the 11th week of gestation (GW) as part of their antenatal care (p < 0.001) ([Fig. 1]). More than twice as many non-Vietnamese women (NV: 26.9%) had a first trimester
screening compared to Vietnamese women (V: 12.2%) (p < 0.001).
Fig. 1
Antenatal care of women with and without a Vietnamese migration background (in %).
It is also worth noting that the rate of hepatitis B surface antigen (HBsAg) was higher
in Vietnamese women (V: 6.4% vs. NV: 0.4%, p < 0.001) and the percentage of women
with gestational diabetes was also higher in this group (V: 13.4% vs. NV: 9.5%, p = 0.013).
Delivery outcomes
An overview of birth outcomes is shown in [Table 2] and [Fig. 2]. As regards the “ideal pregnancy outcome” according to our defined endpoints (i.e.,
delivery after 37 + 0 GW, 5-minute Apgar score ≥ 8, arterial cord blood pH ≥ 7.20,
no transfer of the newborn to the neonatal department, and spontaneous birth or vacuum
extraction without a 3rd or 4th degree perineal tear), there was no significant difference
between women with a Vietnamese migration background (V) and women in the non-Vietnamese
comparison group (NV) (V: 44.5% vs. NV: 38.1%, p = 0.1). However, 75.1% of women of
Vietnamese origin had spontaneous delivery or vacuum extraction without a 3rd or 4th
degree perineal tear, whereas this only applied to 67.2% of women of non-Vietnamese
origin (p < 0.001).
Table 2
Comparison of mode of delivery and childbirth-related injuries in Vietnamese and non-Vietnamese
patients. Mann-Whitney U-test was used for nominal variables and two-sided Fisher’s
exact test for categorical variables (level of significance p = 0.05).
|
|
NV
|
V
|
P
NV vs. V
|
|
All caesarean sections
|
453 (32.1%)
|
103 (21.19%)
|
< 0.001*
|
|
Primary caesarean section
|
262 (57.8%)
|
37 (35.92%)
|
< 0.001*
|
|
Secondary caesarean section
|
191 (42.2%)
|
66 (64.07%)
|
< 0.001*
|
|
Emergency caesarean section
|
31 (6.9%)
|
6 (5.82%)
|
0.829
|
|
Planned caesarean section
|
289 (63.80%)
|
60 (58.25%)
|
0.310
|
|
Repeat caesarean section
|
167 (57.78%)
|
40 (66.70%)
|
0.248
|
|
Breech presentation/transverse presentation
|
54 (18.69%)
|
11 (18.33%)
|
1.0
|
|
Pre-existing maternal condition
|
21 (7.27%)
|
2 (3.33%)
|
0.392
|
|
Macrosomia
|
11 (3.81%)
|
2 (3.33%)
|
1.0
|
|
Fetal growth restriction/placental insufficiency
|
5 (1.73%)
|
–
|
0.592
|
|
Placenta previa
|
4 (1.38%)
|
3 (5.00%)
|
0.101
|
|
Status post myoma enucleation
|
2 (0.69%)
|
–
|
1.0
|
|
Uterine fibroids
|
1 (0.346%)
|
–
|
1.0
|
|
Fetal indication
|
1 (0.346%)
|
–
|
1.0
|
|
Preventive caesarean section
|
23 (7.96%)
|
2 (3.33%)
|
0.277
|
|
Reason for preventive caesarean section
|
|
|
|
|
|
9 (39.13%)
|
1 (50%)
|
1.0
|
|
|
3 (13.04%)
|
–
|
1.0
|
|
|
3 (13.04%)
|
1 (50%)
|
0.300
|
|
|
3 (13.04%)
|
–
|
1.0
|
|
|
1 (4.34%)
|
–
|
1.0
|
|
|
2 (8.70%)
|
–
|
1.0
|
|
|
1 (4.34%)
|
–
|
1.0
|
|
|
1 (4.34%)
|
–
|
1.0
|
|
|
164 (36.20%)
|
43 (41.75%)
|
0.310
|
|
|
73 (44.52%)
|
13 (30.23%)
|
0.117
|
|
|
31 (18.90%)
|
19 (44.19%)
|
0.001*
|
|
|
14 (45.26%)
|
11 (57.89%)
|
0.560
|
|
|
17 (54.83%)
|
8 (42.10%)
|
0.560
|
|
|
1 (0.61%)
|
–
|
1.0
|
|
|
2 (1.22%)
|
3 (6.98%)
|
0.062
|
|
|
21 (12.80%)
|
4 (9.30%)
|
0.608
|
|
|
8 (1.77%)
|
1 (2.32%)
|
0.688
|
|
|
9 (5.49%)
|
2 (4.65%)
|
1.0
|
|
|
12 (7.32%)
|
–
|
0.076
|
|
|
27 (16.46%)
|
1 (2.33%)
|
0.012*
|
|
Spontaneous delivery
|
811 (55.17%)
|
326 (69.4%)
|
< 0.001*
|
|
Vacuum-assisted delivery
|
146 (10.35%)
|
41 (8.7%)
|
< 0.001*
|
|
With epidural anesthetic
|
411 (42.9%)
|
82 (22.3%)
|
< 0.001*
|
|
With injuries
|
703 (73.46%)
|
262 (71.38%)
|
0.448
|
|
|
129 (13.48%)
|
48 (13.08%)
|
0.928
|
|
|
170 (17.7%)
|
85 (23.17%)
|
0.029*
|
|
|
8 (0.83%)
|
13 (3.54%)
|
< 0.001*
|
|
|
1 (0.10%)
|
–
|
1
|
|
|
232 (24.24%)
|
99 (26.98%)
|
0.199
|
|
|
90 (62.64%)
|
21 (51.2%)
|
0.281
|
|
|
142 (17.51%)
|
78 (23.93%)
|
0.016*
|
|
|
7 (0.73%)
|
4 (1.09%)
|
0.510
|
|
|
137 (14.32%)
|
18 (4.90%)
|
< 0.001*
|
|
|
241 (25.81%)
|
37 (10.08%)
|
< 0.001*
|
|
Gestational age at delivery in weeks
|
39 + 1 ± 2.12
|
39 + 1 ± 1.57
|
0.047*
|
|
Birth weight in g
|
3310.11 ± 607.166
|
3226.21 ± 440.487
|
< 0.001*
|
|
Infant height in cm
|
50.59 ± 3.30
|
50.18 ± 2.41
|
< 0.001*
|
|
Infant head circumference in cm
|
34.5 ± 2.05
|
34.26 ± 1.76
|
< 0.001*
|
|
|
165 (11.7%)
|
71 (15.1%)
|
0.064
|
|
|
1156 (82.0%)
|
387 (82.3%)
|
0.890
|
|
|
89 (6.3%)
|
12 (2.6%)
|
0.0013*
|
|
Arterial cord blood pH
|
7.23 ± 0.080
|
7.2404 ± 0.07
|
0.356
|
|
|
7 (0.5%)
|
–
|
0.203
|
|
|
51 (3.6%)
|
7 (1.5%)
|
0.0202*
|
|
|
329 (23.5%)
|
113 (24 %)
|
0.203
|
|
|
1014 (72.4%)
|
350 (74.5%)
|
0.310
|
|
5-minute Apgar score
|
9.33 ± 0.981
|
9.46 ± 0.92
|
0.003*
|
|
5-minute Apgar score ≤ 7
|
76 (5.4%)
|
20 (4.3%)
|
0.397
|
|
Transfer to neonatal department
|
224 (16.0%)
|
56 (11.9%)
|
0.036*
|
|
At ≥ 37 + 0 weeks of gestation
|
148 (11.3%)
|
39 (8.8%)
|
0.182
|
|
Peripartum blood loss in ml
|
400.54 ± 303.238
|
434.38 ± 417.05
|
0.229
|
|
Blood loss
|
|
|
|
|
|
893 (64.6%)
|
307 (65.9%)
|
0.654
|
|
|
441 (31.9%)
|
136 (29.2%)
|
0.298
|
|
|
29 (2.1%)
|
12 (2.6%)
|
0.585
|
|
|
19 (1.4%)
|
11 (2.4%)
|
0.144
|
|
Hemoglobin concentrations at discharge (mg/dl)
|
10.95 ± 1.47
|
10.88 ± 1.57
|
0.639
|
|
Received packed red blood cells
|
6 (0.4%)
|
8 (1.7%)
|
0.0101*
|
Fig. 2
Pregnancy outcomes of women with and without a Vietnamese migration background (in
%).
Indications for and rate of caesarean sections
The caesarean section rate for women with a Vietnamese migration background was 21.2%
(V) compared to 32.1% for non-Vietnamese (NV) women (p < 0.001), which was around
10% lower than in the comparison group, although the indications for unplanned caesarean
sections differed; the percentage of patients of Vietnamese origin who required a
caesarean section for failure to progress in labor was lower than in the comparison
group (V: 44.19% vs. NV: 18.90%, p < 0.001). Women without a Vietnamese migration
background required a caesarean section more often for failed induction of labor (NV:
16.46% vs. V: 2.33%, p = 0.012).
EDA rate, incidence of perineal tears and episiotomies
With a rate of 42.9% compared to 22.3%, almost twice as many non-Vietnamese patients
received an epidural during vaginal delivery (p < 0.001).
After vaginal delivery, the rate of 3rd degree perineal tears was three times higher
in patients of Vietnamese origin (V: 3.55% vs. NV: 0.94%, p < 0.001), and the rate
of episiotomies during spontaneous delivery was also higher (V: 23.9% vs. NV: 17.5%,
p = 0.016).
Neonatal birth outcomes
Children born to women of Vietnamese origin were born more often between week 37 + 0
and 40 + 0 of gestation and required transfer to the neonatal department less often
(V: 11.9% vs. NV: 16%, p < 0.0036). The 5-minute Apgar score was slightly better for
children born to women of Vietnamese origin (V: 9.46 vs. = NV: 9.33, p = 0.003). Likewise,
detection of acidosis in arterial cord blood pH occurred less often in neonates born
to women of Vietnamese origin (NV: 3.6% vs. V: 1.5%, p = 0.0202). Macrosomia also
occurred less often in children born to women of Vietnamese origin (V: 2.6% vs. NV:
6.3%, p = 0.0013), although it must be noted that percentile curves for “Caucasian”
children were used [6].
Difference between women of Vietnamese origin registered in a refugee shelter and
those not registered in refugee shelter
The characteristics of Vietnamese patients registered in a refugee shelter and women
of Vietnamese origin not registered in a refugee shelter are listed in [Table 3]. Details on the respective pregnancy outcomes are given in Table S1. Vietnamese women with a registered address in a refugee shelter were three years
younger, on average, compared to Vietnamese women who did not live in a refugee shelter
(26.46 vs. 29.49 years, p < 0.001), had fewer children and had given birth less often
(mean: 1.54 vs. 1.96, p < 0.001). With a rate of 4.5% vs. 16.3% (p = 0.0196), they
were far less likely to attend a first trimester screening appointment and had a detailed
fetal ultrasound screening (second trimester screening in weeks 18 + 0–22 + 0 of gestation)
less than half as often (28.8% vs. 68.3%, p < 0.001) and in 40.7% of cases vs. 2.3%
(p < 0.001) only presented to an antenatal screening visit after week 20 of
gestation. Women living in a refugee shelter required an unplanned caesarean section
significantly more often (56.76% vs. 33.33%, p = 0.024); at 13.5% vs. 1.5%, the percentage
of emergency caesareans in this group was high (p = 0.022). The resulting maternal
and infant birth outcomes show that infants born to mothers living in a refugee shelter
had the same results with regards to Apgar scores and arterial cord blood pH-values
but required transfer to a neonatal department less often (6.69% vs. 16.09%, p = 0.0016)
despite a higher SGA rate (small for gestational age, neonatal weight ≤ 10th percentile)
of 20.09% (vs. 11.1%, p = 0.0092). Peripartum blood loss of more than 500 ml also
occurred more often in Vietnamese women living in a refugee shelter than in the comparison
groups (70.82% vs. 60.91%, p = 0.032).
Table 3
Basic characteristics of the group of patients of Vietnamese origin with and without
a registered address in a shelter for refugees, including the number of cases; Mann-Whitney
U-test was used for nominal variables and two-sided Fisher’s exact test for categorical
variables (level of significance p = 0.05).
|
|
Vietnamese migration background and a registered address in a shelter for refugees
|
n
|
Vietnamese migration background and a private address
|
n
|
p
|
|
Maternal age (years)
|
26.46 ± 4.5
|
209
|
29.48 ± 5.7
|
261
|
< 0.001*
|
|
Height (cm)
|
155.87 ± 4.60
|
209
|
156.79 ± 5.01
|
261
|
0.066
|
|
Body mass index (BMI) (kg/m2)
|
20.66 ± 2.71
|
205
|
20.73 ± 2.69
|
257
|
0.645
|
|
Underweight (< 18.5 kg/m2)
|
43 (20.98%)
|
|
50 (19.46%)
|
|
0.727
|
|
Normal weight (18.5–< 30 kg/m2)
|
159 (77.56%)
|
|
205 (79.77%)
|
|
0.569
|
|
Overweight (≥ 30 kg/m2)
|
3 (1.4%)
|
|
2 (0.8%)
|
|
0.659
|
|
Parity
|
1.54 ± 0.73
|
209
|
1.96 ± 0.995
|
261
|
< 0.001*
|
|
Gravidity
|
1.67 ± 0.89
|
|
2.26 ± 1.23
|
|
< 0.001*
|
|
Primiparous
|
99 (57.9%)
|
|
104 (39.8%)
|
|
0.1115
|
|
First trimester screening in week 11 + 0–13 + 0 of gestation
|
3 (4.5%)
|
66
|
20 (16.3%)
|
123
|
0.0196*
|
|
Detailed fetal scan in week 18 + 0–22 + 0 of gestation
|
19 (28.8%)
|
66
|
84 (68.3%)
|
123
|
< 0.001*
|
|
Number of antenatal care visits
|
|
206
|
8.97 ± 3.0
|
249
|
< 0.001*
|
|
|
7.67 ± 2.82
|
|
21 (8.4%)
|
|
0.0944
|
|
|
28 (13.6%)
|
|
21 (19.7%)
|
|
< 0.001*
|
|
|
72 (35%)
|
|
21 (53.8%)
|
|
< 0.001*
|
|
|
90 (43.7%)
|
|
45 (18.1%)
|
|
0.0014*
|
|
First antenatal care visit
|
16 (7.8%)
|
204
|
|
258
|
|
|
|
30 (14.7%)
|
|
92 (35.65%)
|
|
< 0.001*
|
|
|
91 (44.6%)
|
|
130 (50.39%)
|
|
0.224
|
|
|
83 (40.69%)
|
|
36 (2.3%)
|
|
< 0.001*
|
|
HBsAg positive
|
13 (6.2%)
|
209
|
17 (6.5%)
|
261
|
1.0
|
|
Gestational diabetes
|
|
208
|
|
261
|
|
|
|
24 (11.5%)
|
|
39 (14.9%)
|
|
0.340
|
|
|
62 (29.8%)
|
|
56 (21.5%)
|
|
0.042*
|
|
|
–
|
24
|
11 (28.20%)
|
39
|
0.004*
|
Factors influencing caesarean section rate, neonatal transfer to a neonatal department,
and 3rd degree perineal tear
A comparison of the characteristics of women with and without a Vietnamese migration
background and the subgroup analysis of Vietnamese patients living in a refugee shelter
and Vietnamese patients not living in a refugee shelter showed significant differences
between the compared groups, including differences in maternal height, BMI, the presence
of gestational diabetes, neonatal birth weights, and attendance at antenatal screening
visits as independent variables. Any significant impact of relevant independent variables
on birth outcomes therefore had to be investigated using multivariable regression
analysis, especially the dependent variables “caesarean section rate”, “rate of neonatal
transfers to a neonatal department” and the “rate of perineal tears.” [Fig. 3] and Table S2 shows the results: according to our analysis, living in a refugee shelter had the
greatest protective effect against the need to have a caesarean
section, with an odds ratio (OR) of 0.51 (95% CI: 0.36–0.73); greater maternal height
was also a protective factor with an OR of 0.97 (95% CI: 0.96–0.99). The effect of
a Vietnamese migration background just missed being significantly protective with
an OR of 0.68 (95% CI: 0.43–1.01).
Fig. 3
Factors influencing caesarean section according to multivariate regression analysis;
the odds ratios (OR) and 95% confidence intervals (CI) are shown.
According to our analysis, a higher gestational age at delivery (OR 0.93, 95% CI:
0.91–0.95), a higher 5-minute Apgar score (OR 0.45, 95% CI: 0.38–0.52) and maternal
residence in a shelter for refugees (OR 0.45, 95% CI: 0.23–0.89) had a significant
protective effect on the transfer rates of neonates to the neonatal department.
When we investigated the factors influencing the occurrence of a 3rd or 4th degree
perineal tear, multivariable regression analysis showed that a Vietnamese migration
background was the only significant risk factor with an OR of 5.39 (95% CI: 1.36–21.30).
Discussion
Summary of the most important findings
The investigated women with a Vietnamese migration background attended fewer antenatal
screening appointments compared to the control group, but the pregnancy outcomes for
women with a Vietnamese migration background were just as good and the caesarean section
rates were lower as were the transfer rates of neonates to the neonatal department.
Regression analysis showed that the independent protective variable against caesarean
section was not a Vietnamese migration background as such, but maternal residence
status with the mother living in a shelter for refugees. Interestingly, alongside
the Apgar score, refugee status was an independent protective factor against transfer
of the neonate to the neonatal department, and not the Vietnamese migration background
as such.
Antenatal care of women with a Vietnamese migration background
The percentage of Vietnamese patients living, or at least registered as living, in
a refugee shelter was surprisingly high at 44%. In the European Union, the percentage
of persons with refugee status according to the data of the UNHCR (United Nations
High Commissioner for Refugees) was 0.6% in 2022, which almost corresponds to the
percentage of patients in our investigated control group registered as staying in
a refugee shelter without a Vietnamese migration background (0.8%)
As 95% of the Vietnamese women registered to a refugee shelter only attended antenatal
care appointments from week 11 of gestation and just under 41% only attended antenatal
screening appointments from week 20 of gestation, future studies should investigate
whether these pregnant women are not offered adequate medical screening early on or
whether they have only been living in Germany since a short time and the necessary
medical examinations have been carried out abroad. Even if first trimester screening
is not part of the Maternity Protection Directive (Mutterschaftsrichtlinie) issued by the Joint Federal Committee of Germany despite this being proposed by
the Fetal Medicine Foundation, first trimester screening has the potential to detect
40–70% of serious malformations early on; up to 95% of aneuploidies are detected [7]
[8]
[9]
The generally higher risk of Asian female patients should also be considered, meaning
that screening for gestational diabetes should be carried out even if Asian women
present quite late for their first gynecological examination and, if necessary, these
women should be given nutritional counselling in their national language [10]
[11]
[12]. In all cases, our study found that identified pregnancy outcomes for the group
of patients with a Vietnamese migration background, especially for women living in
a refugee shelter, were not worse, despite the lower attendance at antenatal care
appointments. These results agree with those of the comparative study of Vietnamese
and non-Vietnamese women in Berlin of Boxall et al. Similarly, in their study of perinatal
outcomes which compared women with and those without a refugee
status, Rosenberg-Jeß et al. came to the conclusion that there was no difference in
pregnancy outcomes despite the lower levels of antenatal care [13]
[14]. In their review published in 2023, Ramadan et al. summarized international studies
published in the last 50 years which examined the perinatal outcomes of women with
refugee status in OECD countries. Out of more than 1000 publications, the authors
only identified nine publications which studied the perinatal health outcomes of refugees;
one of these studies was an Australian survey which analyzed the perinatal outcomes
of Vietnamese refugees. The authors also reported lower caesarean section rates and
a lower need for peripartum analgesia. Studies of refugee women from other countries
showed the same or lower caesarean section rates compared to local women in the respective
country [15]. The systematic review by Heslehurst, which compared 29 reviews on the perinatal
care of women with a refugee status and female asylum seekers, has pointed to the
sometimes contradictory findings with regards to preterm birth rate, fetal growth
restriction, perinatal mortality and mode of delivery [16]. The authors explain this by pointing out the heterogeneity of the study populations
and the differences in countries of origin and host countries.
The finding that perinatal outcomes were better despite poorer socioeconomic conditions
has been called the “Latina paradox,” in reference to the perinatal outcomes of female
Hispanic immigrants to the USA compared to women already living in the USA [17]. Cited reasons for this paradox include differences in eating habits and nicotine
consumption and cultural differences. When the “Latina paradox” is transferred to
other populations, it is known as the “healthy migrant effect,” i.e., despite poorer
social and/or financial conditions, immigrants are in better health than the populations
in their country of origin and the host country [17]. The healthy migrant effect in our analysis is underlined by the fact that the reasons
cited for having a planned caesarean section did not include pre-existing maternal
or fetal medical conditions or growth restriction. Women who
have the health and social resources to migrate from Vietnam to Germany could therefore
also have better pregnancy outcomes. Elshahat et al. accounted for the healthy migrant
effect by suggesting that immigrants have greater resilience and adaptability to adverse
living conditions [18]. As regards our study population, it could be speculated whether Vietnamese women
who only recently immigrated to Germany might have different cultural concepts about
antenatal care and giving birth compared to women with a Vietnamese migration background
who were born in Germany or had lived in Germany for some time and had therefore adapted
to “German conditions.”
Mode of delivery: caesarean section
In contrast to pre-existing medical conditions which require caesarean section, “soft”
indications for caesarean section such as preventive caesarean section or caesarean
section for failure to progress in labor were not diagnosed in individual women with
a Vietnamese migration background or Vietnamese women registered to a refugee shelter.
Future studies will be required to determine whether a poorer overall state of health
or more frequent antenatal counselling and medical interventions during antenatal
care might have led to the high rates of caesarean section in non-Vietnamese women.
With regards to women of Vietnamese origin, cultural preferences as well as more limited
communication with these women due to the language barrier could also have played
a role; this could not be conclusively determined in our study.
The state of Queensland has provided an online summary of the preferences and cultural
practices associated with birth in the Vietnamese diaspora in Australia [19]. The document also reports that Vietnamese women are skeptical about having a caesarean
section and fear complications such as blood loss. At the same time, other publications
show that the caesarean section rate in the urban areas of Vietnam has risen to more
than 40% and is associated with higher incomes, high or low infant birth weights,
and higher maternal age [20].
Our high percentage (15%) of emergency caesarean sections in Vietnamese women who
lived in a refugee shelter is also replicated in the literature; Gagnon et al. suggested
that low income and a lack of health care insurance are risk factors [21]. Miani et al. confirmed a connection between low income and emergency caesarean
section [22].
As the patients of Vietnamese origin in our study population were about 10 cm shorter
and slim, cephalopelvic disproportion was also expected. In fact, failure to progress
in labor as an indication for caesarean section was reported more than twice as many
times for the group of patients with a Vietnamese migration background (V: 44.19%
vs. NV: 18.90%); however, the incidence of planned caesarean sections for macrosomia
was almost the same in both groups (NV: 3.8% and V: 3.3%).
Mode of delivery: vaginal delivery
EDA rates and the occurrence of higher degree perineal tears differed considerably
between women with and women without a Vietnamese migration background.
Only half as many women with a Vietnamese migration background received EDA during
vaginal delivery compared to women with a different migration background or women
of German origin, although there were no differences with regards to refugee status.
This difference could be due to insufficient medical information because of the language
barrier but could also be ascribed to cultural preferences and ideas about EDA catheter
placement, as Dao et al. showed in a study carried out in Switzerland [15]
[23]. In a survey carried out in Hanoi, Nguyen reported that for the 50% of women who
had EDA during vaginal delivery, maternal age > 35 years, multiparity, higher income
and higher educational status as well as an urban place of origin were predictors
for the request for EDA [24]. These are findings that concur with
the lower EDA rate found in our study for primiparous women with a lower mean age
who were often refugees.
The literature largely confirms that higher degree perineal tears are three times
more common in women with a Vietnamese migration background. Although a similar study
by Boxall et al., which investigated the perinatal data of Vietnamese women, only
reported a higher rate of episiotomies, something we had also noted in our cohort,
and reported no change in the rate of third-degree perineal tears, this was contrary
to findings in other international publications [13]. Just like our regression analysis showed (OR for 3rd or 4th degree perineal tear,
Vietnamese migration background 5.4, 95% CI: 1.4–21.4), in an analysis of patients
living in Australia Davies-Tuck et al. also found that south Asian ethnicity was an
independent risk factor for a higher degree perineal tear with an OR of 3.1 (95% CI:
2.3–4.0) [25]. The recently published systematic review by Park et al.,
which looked at the results of 27 studies published over the last 30 years, confirmed
that Asian ethnicity is a risk factor for the occurrence of higher degree perineal
tears in non-Asian countries but that this higher rate of perineal tears is only partly
reproduced in Asian countries, suggesting, in the opinion of the authors, that social
factors, the language barrier, and racism could also be potential influencing factors
[26]. Purely biological factors such as small maternal height or length of the perineum
were not identified in the review as significant factors for a higher incidence of
perineal tears.
Limitations
This study is a retrospective exploratory analysis, in which patients with a Vietnamese
migration background were identified based on an analysis of their names, and were
compared with a heterogeneous group of parturient women of non-Vietnamese origin.
Even though no data on the specific ethnicity of patients was collected, this onomastic
approach used to identify people from Vietnam has accuracy rates of more than 99%
[27]
[28].
The women in our study who served as the comparison group for the women of Vietnamese
origin consisted of a group of women different nationalities and origins. The potential
impact of a possible language barrier could therefore not be determined for either
the group of Vietnamese women or for the control group, as no information on the German
language skills of the investigated patients with and without a Vietnamese migration
background was available. There was also no information about patients’ social status,
level of education and financial situation.
Conclusion for medical practice
The good pregnancy outcomes despite the more limited antenatal care reported for women
with a Vietnamese migration background with and without a registered address in a
refugee shelter should prompt reflections on how to make better use of resources when
providing gynecological care to pregnant women. The inverted pyramid proposed by Nicolaides,
which selects high and low-risk patients early on and only envisages a few routine
examinations for healthy pregnant women in the third trimester of pregnancy, could
serve as an example [29].
In addition to providing interpreters to avoid language barriers and showing sufficient
empathy for the patient’s individual wishes, gynecological counseling about the choice
of delivery mode also requires that patients are properly informed about the incidence
of risks during delivery. A closer study into the reasons for the not insignificant
number of failed labor inductions and preventive and repeat caesarean sections carried
out in non-Vietnamese women could potentially reduce the overall rate of caesarean
sections in Germany in future.
Supplement
Table S1: Comparison of mode of delivery and birth injuries of patients with a Vietnamese migration
background registered as living in a refugee shelter and those not living in a refugee
shelter; Mann-Whitney U-test was used for nominal variables and two-sided Fisher’s
exact test for categorical variables (level of significance: p = 0.05).
Table S2: Regression analysis: a) factors influencing caesarean sections, b) transfer to the neonatal department, c) third-degree perineal tear.
Note
The investigation received no funding. Due to its retrospective character, we did
not perform a public registration in advance.
