Long‐term neurological morbidity among children delivered by vacuum extraction ‐ a national cohort study

Abstract

Introduction

This is the first nationwide cohort study of vacuum extraction (VE) and long‐term neurological morbidity. We hypothesized that VE per se, and not only complicated labor, can cause intracranial bleedings, which could further cause neurological long‐term morbidity. The aim of this study was to investigate the risk of neonatal mortality, cerebral palsy (CP), and epilepsy among children delivered by VE in a long‐term perspective.

Material and methods

The study population included 1 509 589 term singleton children planned for vaginal birth in Sweden (January 1, 1999 to December 31, 2017). We investigated the risk of neonatal death (ND), CP, and epilepsy among children delivered by VE (successful or failed) and compared their risks with those born by spontaneous vaginal birth and emergency cesarean section (ECS). We used logistic regression to study the adjusted associations with each outcome. The follow‐up time was from birth until December 31, 2019.

Results

The percentage and total number of children with the outcomes were ND (0.04%, n = 616), CP (0.12%, n = 1822), and epilepsy (0.74%, n = 11 190). Compared with children delivered by ECS, those born by VE had no increased risk of ND, but there was an increased risk for those born after failed VE (adj OR 2.23 [1.33–3.72]). The risk of CP was similar among children born by VE and those born spontaneously vaginally. Further, the risk of CP was similar among children born after failed VE compared with ECS. The risk of epilepsy was not increased among children born by VE (successful/failed), compared with those who had spontaneous vaginal birth or ECS.

Conclusions

The outcomes ND, CP, and epilepsy are rare. In this nationwide cohort study, children born after successful VE had no increased risk of ND, CP or epilepsy compared with those delivered by ECS, but there was an increased risk of ND among those born by failed VE. Concerning the studied outcomes, VE appears to be a safe obstetric intervention; however, it requires a thorough risk assessment and awareness of when to convert to ECS.

This national cohort study of 1.5 million children planned for a vaginal birth investigates the risk of neonatal mortality, cerebral palsy, and epilepsy among children delivered by vacuum extraction in a long‐term perspective.Our results indicate that VE is a safe obstetric intervention concerning the studied outcomes, however, it implies a thorough risk assessment.

Abbreviations

AGA

average for gestational age

CP

cerebral palsy

ECS

emergency cesarean section

LGA

large for gestational age

ND

neonatal death

SGA

small for gestational age

SVB

spontaneous vaginal birth

VE

vacuum extraction

Key message.

Children born after successful vacuum extraction had no increased risk of neonatal death, cerebral palsy or epilepsy compared with those delivered by emergency cesarean section. Vacuum extraction is an equivalent or preferable alternative when needed to expedite birth; however, it requires a thorough risk assessment.

1. INTRODUCTION

Vacuum extraction (VE) is a common obstetrical procedure in the Western world, and in many countries, it has replaced the use of forceps. ¹ , ² In Sweden, VE is used in approximately 6% of all births, 14% among primiparas, and 2.8% among multiparas. Approximately 18% of women undergo a cesarean section, of those, 9.5% are emergency cesarean sections (ECS). ³ Although VE is used to prevent neonatal complications, it has been associated with rare but serious iatrogenic morbidity, such as extra‐ and intracranial hemorrhage and skull fractures. ⁴ , ⁵ , ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ The prevalence of intracranial hemorrhage in term infants varies widely, from 8% to 26% due to differences in the study population, follow‐up time, and the sensitivity in diagnostics. ¹¹ , ¹² , ¹³ , ¹⁴ Minor intracranial bleedings, detected by magnetic resonance in asymptomatic infants, have not been associated with adverse outcomes. ¹³ , ¹⁵ However, greater bleeding (symptomatic or asymptomatic) might have an association with neonatal mortality and long‐term neurological morbidities like cerebral palsy (CP) and epilepsy.

Despite the risk of neonatal cerebral complications, few studies have investigated the association between VE on long‐term consequences, and there is a need for more evidence. ¹⁶ Previous studies on VE have limitations of small sample size and retrospective design, with no comparison made between complications in VE and ECS deliveries. Further, the use of VE has decreased in some countries in favor of ECS. ¹⁷ Since ECS in the second stage of labor is a difficult procedure with increased risk for adverse outcomes for the mother like post‐partum hemorrhage, sepsis, and uterine tear ¹⁸ , ¹⁹ , ²⁰ it is important to clarify the risks and benefits of both VE and ECS.

Our hypothesis is that VE per se, can cause intracranial hemorrhage and that these in turn could lead to neurological damage. The aim of this study was to investigate the risk of neonatal death and long‐term neurological morbidity, that is, CP and epilepsy among children delivered at term by VE (successful and failed), and to compare their risks with those born spontaneously and by ECS.

2. MATERIAL AND METHODS

This historical cohort study used data from nationwide Swedish registries which were linked using the unique personal identification number assigned to all Swedish residents at birth or on a residence permit. Information from the Medical Birth Register ²¹ was cross‐linked with data from the National Patient Registry. ²²

The Swedish Medical Birth Register includes prospectively collected information on demographic data, reproductive history, and complications during pregnancy, childbirth, and the neonatal period for more than 98% of all births in Sweden. It also includes maternal characteristics which are obtained in a standardized manner from the woman's first antenatal visit, throughout pregnancy and birth. The National Patient Registry contains diagnostic codes on hospital inpatient care since 1987, and outpatient care from 2001 onwards, including dates for diagnoses. ²² , ²³

The study population was retrieved from Swedish Medical Birth Register and included all singleton children born in Sweden from 1999 to 2017. Pre‐ and post‐term babies were excluded in order to obtain a study population in the normal range of gestational age, that is, week 37.0–41.6. All fetuses were alive when entering labor. The onset of labor was spontaneous or induced, and modes of delivery were divided into VE, ECS, and spontaneous vaginal birth (SVB). Births with VE followed by ECS were analyzed separately and were defined as “failed VE” (Flow chart, Figure 1). The follow‐up time was from birth until December 31, 2019, which implied a follow‐up time between 2 and 21 years depending on the date of birth. Breech presentation, elective cesarean section, multiple births, congenital malformations (see attached file with diagnoses), antepartum stillbirths, and births where forceps were used (both combined with VE or ECS and exclusively) were excluded (n = 217 789), see flow chart, Figure 1.

FIGURE 1.

Flow chart describing the study population and mode of birth.

Diagnoses were defined according to the Swedish version of the International Classification of Diseases (ICD‐10) (Table S1). We investigated three main outcomes: Neonatal death, CP and epilepsy. Neonatal death was defined as death within 28 days after birth, CP was defined as any G80 after birth and epilepsy was defined as any G40 set after 28 days on two different occasions or an R56 code first, followed by a G40 at a later date. These outcomes were found by merging the registries described above, using the unique personal identification number assigned to all Swedish residents at birth.

Information on maternal characteristics, birth characteristics, and neonatal characteristics were collected from the Swedish Medical Birth Register. Parity was categorized into primipara or multipara. Measured weight and self‐reported height at the first antenatal visit were used to calculate early pregnancy body mass index (BMI, kg/m²). BMI was categorized according to WHO standards; underweight (below 18.5) normal (18.5–24.9), overweight (25–29.9), obese (≥30), or missing. ²⁴ Self‐reported information from the first antenatal visit regarding smoking was categorized as smoker or non‐smoker. Gestational age (GA), based on routine ultrasound dating performed before 18 postmenstrual weeks in 97%–98% of all pregnant women, was categorized into 37.0–38.6, 39.0–40.6, 41.0–41.6. Birthweight for gestational age was categorized into small for gestational age (SGA) (< −2SDs), appropriate for gestational age (AGA) (−2 SD + 2 SD), and large for gestational age (LGA) (> +2SDs). ²⁵ The fetal position at birth was divided into: occiput anterior, or positions other than occiput anterior. Fetal distress was defined by the diagnostic codes O680‐O683, O688, and O689. Intrauterine hypoxia, meaning hypoxia before the onset of labor, was defined as the diagnosis P200. For codes on further diagnoses, see Table S1.

2.1. Statistical analyses

We described the characteristics of the cohort according to the mode of birth with number (N) and percent (Table 1). We calculated the incidence risk for each outcome as the number of the outcome divided by the number of born children, presented as percentages.

TABLE 1.

. Characteristics of the study population and mode of birth.

	Mode of birth
	Vaginal		VE		Failed VE		ECS
	N	%	N	%	N	%	N	%
Covariates known at onset of labor
Maternal age (year)
≤19	20 883	1.6	1628	1.5	53	1.1	995	1.0
20–34	1 029 360	79.4	89 319	80.8	3877	79.4	71 862	73.1
>34	245 717	19.0	19 556	17.7	952	19.5	25 387	25.8
Maternal height (cm)
<162	316 180	24.4	31 885	28.9	1657	33.9	37 474	38.1
≥162–172	698 548	53.9	57 425	52.0	2423	49.6	45 995	46.8
>172	214 911	16.6	15 344	13.9	584	12.0	9333	9.5
Missing	66 321	5.1	5849	5.3	218	4.5	5442	5.5
BMI
<18.50	30 758	2.4	2861	2.6	81	1.7	1301	1.3
18.5–24.9	731 260	56.4	64 697	58.5	2682	54.9	44 366	45.2
25.0–29.9	288 290	22.2	23 688	21.4	1208	24.7	26 335	26.8
≥30.0	129 940	10.0	9134	8.3	511	10.5	16 903	17.2
Missing	115 712	8.9	10 123	9.2	400	8.2	9339	9.5
Parity
Primipara	501 588	38.7	87 232	78.9	4024	82.4	60 722	61.8
Multipara	794 372	61.3	23 271	21.1	858	17.6	37 522	38.2
Smoking at enrolment in maternity health care
Non‐smoker	1 135 996	88.3	98 184	89.4	4452	91.9	85 795	88.0
Smoker	91 809	7.1	6547	6.0	200	4.1	6897	7.1
Missing	58 118	4.5	5155	4.7	193	4.0	4829	5.0
Maternal CP	21	0.0	3	0.0	0	0.0	11	0.0
Maternal epilepsy	2266	0.2	254	0.2	12	0.2	249	0.3
Maternal diabetes (type 1/2/GDM)	16 246	1.3	1873	1.7	122	2.5	3720	3.8
Hypertension or PE	27 768	2.1	4351	3.9	192	3.9	7847	8.0
SGA or LGA
AGA	1 234 950	95.3	104 614	94.7	4550	93.2	87 186	88.7
SGA	19 684	1.5	2841	2.6	50	1.0	4088	4.2
LGA	39 288	3.0	2835	2.6	266	5.4	6759	6.9
Missing	2038	0.2	213	0.2	16	0.3	211	0.2
Gestational week
37+0–38+6	223 921	17.3	14 488	13.1	504	10.3	20 039	20.4
39+0–40+6	797 269	61.5	63 285	57.3	2617	53.6	48 273	49.1
41+0–41+6	274 770	21.2	32 730	29.6	1761	36.1	29 932	30.5
Intrapartum covariates
Intrauterine hypoxia	18	0.0	20	0.0	1	0.0	43	0.0
Fever during labor	4411	0.3	1982	1.8	132	2.7	3177	3.2
Chorioamnionitis	579	0.0	273	0.2	30	0.6	1473	1.5
Signs of fetal distress	16 984	1.3	51 591	46.7	2142	43.9	35 564	36.2
Prolonged labor	80 589	6.2	49 063	44.4	2867	58.7	38 690	39.4
Maternal exhaustion	6085	0.5	6858	6.2	292	6.0	2014	2.0
Indication for VE or ECS
Signs of fetal distress	13 565	1.0	39 064	35.4	1287	26.4	27 336	27.8
Prolonged labor	77 320	6.0	37 397	33.8	2051	42.0	30 690	31.2
Both	3419	0.3	12 527	11.3	855	17.5	8228	8.4
None of these	1 201 656	92.7	21 515	19.5	689	14.1	31 990	32.6
Shoulder dystocia	2459	0.2	1130	1.0	0	0.0	4	0.0
Occiput anterior position
No	43 946	3.4	11 568	10.5	1881	38.5	26 543	27.0
Missing	25 100	1.9	1880	1.7	664	13.6	15 153	15.4
Covariates known after birth
Apgar <7 after 5 min
	4928	0.4	2984	2.7	395	8.1	3350	3.4
Missing	7071	0.5	310	0.3	14	0.3	446	0.5
Sex
Girl	651 058	50.2	48 398	43.8	1931	39.6	44 301	45.1
Missing	2032	0.2	212	0.2	16	0.3	211	0.2
Infant birthweight (g)
<3000	130 465	10.1	12 047	10.9	269	5.5	12 585	12.8
3001–3500	445 879	34.4	37 274	33.7	1244	25.5	27 262	27.7
3501–4000	486 001	37.5	40 874	37.0	1972	40.4	32 483	33.1
4001–4500	192 503	14.9	16 844	15.2	1062	21.8	19 102	19.4
>4501	39 080	3.0	3252	2.9	319	6.5	6601	6.7
Meconium aspiration	1000	0.1	246	0.2	23	0.5	557	0.6
Intracranial bleeding	307	0.0	153	0.1	11	0.2	65	0.1
Neonatal convulsions	905	0.1	462	0.4	44	0.9	442	0.4
Disturbances of cerebral status/encephalopathy	613	0.0	551	0.5	63	1.3	524	0.5
Neonatal infections	4108	0.3	624	0.6	44	0.9	665	0.7
Jaundice	29 066	2.2	5868	5.3	326	6.7	2316	2.4
Hypoglycemia	18 441	1.4	2985	2.7	199	4.1	5855	6.0
Epilepsy	9254	0.7	921	0.8	47	1.0	968	1.0
CP	1276	0.1	219	0.2	19	0.4	308	0.3
Neonatal death	339	0.0	96	0.1	17	0.3	164	0.2
Year of birth
1999–2005	426 847	32.9	37 797	34.2	1314	26.9	31 148	31.7
2006–2010	347 772	26.8	33 318	30.2	1450	29.7	27 836	28.3
2011–2015	369 102	28.5	29 718	26.9	1529	31.3	28 188	28.7
2016–2017	152 239	11.7	9670	8.8	589	12.1	11 072	11.3

Abbreviations: CP, cerebral palsy; ECS, emergency cesarean section; GDM, gestational diabetes mellitus; LGA, large for gestational age; PE, pre‐eclampsia; SGA, small for gestational age; VE, vacuum extraction.

To study the associations between the mode of birth and each of the outcomes, we first compared children delivered by ECS, VE, and Failed VE with spontaneous vaginal birth as reference (Figures 2, 3, 4). Second, to compare VE and Failed VE with ECS, we repeated the analysis with ECS as a reference (Figure 5).

FIGURE 2.

Associations between neonatal death and mode of birth. Adjusted odds ratio (Adj OR) are adjusted for all covariates in Figure 2 known before birth. CP, cerebral palsy; ECS, emergency cesarean section; GDM, gestational diabetes mellitus; LGA, large for gestational age, PE, pre‐eclampsia; SGA, small for gestational age; VE, vacuum extraction.

FIGURE 3.

Associations between cerebral palsy (CP) and mode of birth. Adjusted odds ratio (Adj OR) are adjusted for all covariates in Figure 3 known before birth. CP, cerebral palsy; ECS, emergency cesarean section; GDM, gestational diabetes mellitus; LGA, large for gestational age; PE, pre‐eclampsia; SGA, small for gestational age; VE, vacuum extraction.

FIGURE 4.

Associations between epilepsy and mode of birth. Adjusted odds ratio (Adj OR) are adjusted for all covariates in Figure 4 known before birth. CP, cerebral palsy; ECS, emergency cesarean section; GDM, gestational diabetes mellitus; LGA, large for gestational age; PE, pre‐eclampsia; SGA, small for gestational age; VE, vacuum extraction.

FIGURE 5.

Associations between the three outcomes, Neonatal death, cerebral palsy (CP), epilepsy and vacuum extraction (VE). The adjusted odds ratio (Adj OR) are adjusted for the same covariates as in the Figures 2, 3, 4.

We used logistic regression to estimate the associations. Two models were used for each outcome, one crude and one adjusted. For each outcome, we adjusted for covariates, known before birth, in Figures 2, 3, 4 as categorical variables presented in each table. We adjusted for covariates associated with both the mode of birth and the outcomes (age, maternal height, BMI, parity, diabetes, pre‐eclampsia or hypertension, SGA/LGA, fever during labor, signs of fetal distress, intrauterine hypoxia, other than occiput anterior position, maternal CP or epilepsy and year of birth). Maternal height was included since short maternal stature is associated with complications during labor and height may be considered when choosing between VE and ECS. SGA/LGA was included since ultrasound with weight estimations are done frequently if there are deviations in the babies´ size, hence this information is known during labor and may affect decision‐making when there is a need to expedite birth.

No imputation of data was done and missing data were entered as a separate category in the analysis. We used forest plots to summarize the associations between the covariates and each outcome and present the number of children and present the risk for each outcome. We present associations as odds ratios (OR) with 95% confidence intervals (CI) in forest plots. CI intervals that do not cover one were regarded as significant, at 5% significance level All analyses were done in IBM SPSS version 28 and R version 4.0.1.

2.2. Ethics statement

The study was approved by the Regional Ethical Committee in Stockholm, Dnr: 2013/1922–31/24 on November 27, 2013, amendment approved September 25, 2019.

3. RESULTS

During the 19 year study period, 1 509 589 births fulfilled the inclusion criteria and were included in the analyses. The proportion of children delivered by VE was 7.3% (n = 110 503), failed VE constituted of 0.3% (n = 4882), ECS 6.5% (n = 98 244), and SVB 85.8% (n = 1 295 960).

Characteristics of the study population are shown in Table 1. Primiparas were more common among women giving birth by VE (both failed and successful) and ECS, while multiparas were more common among those giving birth spontaneously vaginally. In the ECS group, more women were overweight or obese as compared to women in the VE and SVB groups. The proportions of infants born SGA or LGA were overrepresented in the ECS group, and LGA infants were overrepresented in the group of failed VE.

The percent of neonates diagnosed with intracranial bleeding was highest in the group of failed VE (0.22%), followed by successful VE (0.14%), ECS (0.07%), and the group of SVB (0.02%). Further, percentages of jaundice, neonatal convulsions, and other disturbances of cerebral status/encephalopathy were highest in the failed VE group.

3.1. Neonatal death

In total, there were 616 neonatal deaths in our cohort, of those, 96 (15.6%) were born by VE, 17(2.8%) by failed VE, 164 (26.6%) by ECS, and 339 (55.0%) by SVB (Figure 2). Compared with children born spontaneously vaginally, the risk of neonatal death (ND) was higher among children born by failed VE, adj OR 6.17 (3.59–10.60), by ECS adj OR 2.77 (2.13–3.61) and successful VE adj OR 1.61 (1.20–2.15). Compared with children delivered with ECS, the risk was higher among failed VE but lower compared with VE (Figure 5). The strongest associations with ND were intrauterine hypoxia and signs of fetal distress.

3.2. Cerebral palsy (CP)

Among the 1822 children diagnosed with CP, 219 (12.0%) were delivered by VE, 19 (1.0%) by failed VE, 308 (16.9%) by ECS, and 1276 (70.0%) by SVB (Figure 3). Compared with children born SVB, children born by failed VE had the highest risks for CP, adj OR 2.39 (95% CI: 1.49–3.84), followed by ECS, adj OR 1.80 (95% CI: 1.52–2.13). After adjusting for covariates known before birth, there was no difference in the risk of CP between children born SVB and those born by VE. Further, there was no difference between failed VE and ECS (Figure 5). The covariates having the highest associations with CP were intrauterine hypoxia, SGA, and signs of fetal distress.

3.3. Epilepsy

Of the 11 190 children who were diagnosed with epilepsy, 921 (8.2%) were born by VE, 47 (4.2%) by failed VE, 968 (8.7%) by ECS, and 9254 (82.7%) by SVB (Figure 4). Compared with children born SVB, those born by successful VE, adj OR 1.04 (95% CI: 0.96–1.13) and failed VE, adj OR 1.29 (95% CI: 0.96–1.73) had no increased risk of epilepsy. However, children born by ECS had a slightly increased risk; adj OR 1.21 (1.12–1.32). The covariates intrauterine hypoxia, and maternal epilepsy diagnosis had the strongest associations with epilepsy in offspring.

There were 802 children who had both CP and epilepsy diagnoses. Thus, among children with CP, nearly half of them also had epilepsy.

4. DISCUSSION

We found that the risk of ND was higher among children born by failed VE but lower among those born by successful VE compared with children delivered with ECS. However, compared with SVB, all three interventions were associated with an increased risk of ND. Regarding CP, the risk was similar among children born by VE compared with those who had SVB and lower compared with those delivered by ESC. Among children born after failed VE, the risk of CP was similar to those born by ESC. Further, compared with children born spontaneously vaginally, the risk of epilepsy was similar among children born by VE (successful or failed). Compared with children delivered with ECS, the risk of epilepsy was slightly lower among those born by VE.

The access to high‐quality health‐registries made it possible to perform a nationwide cohort study including 1.5 million term births and to investigate associations of rare neurological outcomes and events with different modes of birth in a long‐term follow‐up. Data was based on medical birth records and registries with standardized variables coded in the same way, without involving the patients, thus minimizing various types of bias (i.e., selection and recall bias). Moreover, maternity care is free of charge in Sweden, management routines, as well as gestational age determination are standardized, and 99% of women give birth in public hospitals. This minimizes the risk of confounding by unmeasured sociodemographic factors. Adjustments for covariates that may affect the decision of birth mode were done, including the diagnoses signs of fetal distress and intrauterine hypoxia. ²⁶ The latter, implicating hypoxia debuting before labor, were strongly associated with both CP, epilepsy and ND. The adjustments reduced the risk of confounding by indication, detected in the difference between the crude and adjusted models. Having missing values in some covariates was strongly associated with the outcomes, for example, missing birthweight and missing occiput position. The reasons are probably that the task of writing a complete birth record is not prioritized if the neonate is very compromised when born. The potential impact of the missing data on covariates is unknown. We did not perform any multiple imputations (MI) because the non‐testable assumption of missing completely at random that MI is based on is likely not fulfilled. ²⁷

A major limitation was that the data on mode of birth included ECS in both the first and second stages of labor, and not only the second stage, which had been the optimal comparison group. Hence the groups of ECS and VE may differ in some respects; neonates in the ECS group may be a more fragile group, showing intolerance for labor in the first stage of labor, and children in the VE group have been exposed to a longer duration of labor than the ECS‐group. Another limitation is that we had no information on the duration of labor and the use of oxytocin, aspects that are also associated with the outcomes. ²⁸ Further, the registries do not contain detailed information about important characteristics of the VE deliveries (i.e., type of VE instrument, station of the fetal head, the skill of the obstetrician, duration of VE, and cup detachment). Additionally, the risk of confounding by indication cannot be completely eliminated even if we adjusted for indications for VE and ECS such as fetal distress. Hence, the groups may still differ in aspects not possible to adjust for.

Several studies have reported severe neurological morbidity among children born by protracted or failed VE ⁴ , ²⁹ , ³⁰ , ³¹ but contradictive results exist. ³² VE converted to ECS is, per se, more complicated than successful VE. Complications include longer duration of extraction, number of pulls, or cup detachments. Converting to cesarean section also prolongs the time to birth which can worsen asphyxia. Additionally, when a subsequent ECS is conducted, the head of the fetus squeezes through the birth canal a second time, which might increase the trauma against the fetal head. ³³ The similar risks when comparing VE and ECS, indicate that the complicated birth rather than the extraction itself impacts the outcomes. These results are in line with some previous studies. ¹⁴ , ³⁴ The described characteristics related to the mode of birth, highlight the importance of assessing risk factors for failed VE (eg LGA, occiput posterior position, mid‐pelvic‐fetal station, short maternal stature, epidural anesthesia) and following clinical guidelines in the management of VE. ³³ , ³⁵

A review of guidelines for instrumental births in the Anglo‐Saxon world found a wide consensus regarding indication, contraindications, and prerequisites, ²⁰ indicating that generalizability is possible in similar contexts. In Sweden, national advice for clinical guidelines has been developed over the years. ³⁶

The chosen outcomes are rare but severe. CP is a lifelong disability that implies a decreased quality of life as well as high societal costs, while epilepsy has a more heterogeneous long‐term outcome, and remission among children is common. ³⁷ , ³⁸ This study contributes to knowledge of these outcomes which in some instances can have a preventable effect, that is, to carefully assess risk factors for failed VE, following guidelines for the management of VE, and converting to ECS in time. Furthermore, the increased risk for complications and adverse outcomes for the mother is not considered in this study. ¹⁷ , ¹⁸ , ³⁹

5. CONCLUSION

The absolute risks of ND, CP, and epilepsy are very low; however, the number of children born globally is high, which implies that the outcomes affect many children. The present study shows an increased risk of ND after failed VE compared with ECS, but not after successful VE. For the outcome CP, children born by successful VE had similar risk compared with those who had SVB and after failed VE, the risk of CP was similar to those delivered by ECS. Regarding epilepsy, VE (failed or successful) was not associated with increased risks compared to SVB or ECS.

Our results indicate that VE is a safe obstetric intervention concerning the studied outcomes; however, it implies a thorough risk assessment. Causality could not be established in this cohort study and the higher OR in both VE and ECS groups compared with SVB indicate that the complicated labor rather than the mode of the birth per se is associated with the outcomes.

AUTHOR CONTRIBUTIONS

CE initiated the study and retrieved the data and all authors designed the study. The statistical analyses were made by HP and HU. The manuscript was written by HU who received continuous input from all authors.

CONFLICT OF INTEREST STATEMENT

None.

Supporting information

Table S1.

Ulfsdottir H, Ekéus C, Tedroff K, Åberg K, Järnbert‐Pettersson H. Long‐term neurological morbidity among children delivered by vacuum extraction ‐ a national cohort study. Acta Obstet Gynecol Scand. 2023;102:843‐853. doi: 10.1111/aogs.14568