Placental Abruption and Child Mortality

We conducted a population-based case-control study using linked data from population-based birth and cause-of-death registers from Finland. All singleton births in Finland from 1987 to 2005 were included in the study. We used the medical birth register (MBR) and the hospital discharge register (HDR) maintained by the National Institute for Health and Welfare to identify all women with a diagnosis of placental abruption by using codes from the International Classification of Diseases, Ninth Revision (codes 641.20, 641.21, and 641.23 for 1987–1995) and the International Classification of Diseases, 10th Revision (codes O45.0, O45.8, and O45.9 from 1996 onward) in the HDR and a check mark for placental abruption (yes or no) in the MBR (October 1990–2005). A reference group originated from the same registers (HDR and MBR) and consisted of 3 women without placental abruption for each abruption, matched by maternal age, parity, multiplicity, year of birth, and hospital district area. If all controls were not found, the age criteria were loosened by ±1 year to achieve 3 controls for each case. The hospital discharge data were then linked to the MBR by using the mother’s unique personal identification number to identify all births with placental abruption. Multiple pregnancies were excluded from the final study sample. We identified 4190 women with a singleton birth and placental abruption and 12 570 matched referent women with a singleton birth without placental abruption. After excluding stillbirths, the final study population comprised 3888 children born after placental abruption (later referred to as index children) and 12 530 referent children.

The proportion of immigrants in the Finnish population is relatively small. Although data on the native country of the parturient are currently included in the MBR from 2000 onward, these data were not available for the whole study period (1987–2005).

Baseline data on interventions that mothers receive during pregnancy and delivery and on the newborn’s outcome during the first 7 days are collected in the MBR from all delivery units (since 1987). All live births and stillbirths involving infants who are either gestational age of at least 22 completed weeks (≥155 days) or have a birth weight ≥500 g are registered in the MBR. The data are compiled at the time of birth by using the mother’s maternity records. Less than 0.1% of all births are missing from the MBR. The data have been validated and correspond well with data available from hospital records.^15,16 In the HDR, data are collected on all inpatient episodes in all hospitals (since 1967), all outpatient surgical procedures in public hospitals (since 1994), and all outpatient visits in public hospitals (since 1998). The register contains information on admission and discharge date and diagnosis.

Both index and referent children were linked to the cause-of-death register (maintained by Statistics Finland) by their unique personal identification number to identify all deaths recorded up to the end of 2013. Causes of death were obtained from death certificates. The determination of the cause of death is based on medical or forensic evidence required for the death certificate. Forensic determination of the cause of death may be necessary if the death is not the result of an illness, if it is accidental or violent, or if it is caused by a treatment procedure or an occupational disease.¹⁷ In other cases, the death certificate is based on medical evidence. Data on underlying causes of death since 1969 have been collected in a computerized database. The coverage of cause-of-death statistics is 99.9%, and coding is based on International Classification of Diseases, Ninth Revision codes (used during 1987–1995) and International Classification of Diseases, 10th Revision codes (used since 1996). The causes of death were grouped by taking into account both the main and underlying causes of death. External causes of death included accidents, homicides, and suicides. All deaths of index and referent children were manually double-checked by 2 authors (O.R. and M.T.) to confirm the accurate cause of death.

The duration of gestation calculated from the last menstrual period was confirmed or corrected by ultrasound screening examinations at 11 to 13 or 18 to 20 weeks’ gestation. Recorded in the MBR is the best clinical estimate for gestational age at birth. Smoking habits were recorded during antenatal clinic visits. Women who smoked at least 1 cigarette per day through the pregnancy were defined as smokers. Birth weight and gestational age were used to study the outcome of newborns. Newborns who were small for gestational age (SGA) were defined as having a birth weight ≤−2 SD, and newborns who were large for gestational age (LGA) were defined as having a birth weight ≥+ 2 SD of the national sex-specific standard.¹⁸ Extremely preterm newborns were born before 28 weeks’ gestation, very preterm newborns were born from 28 0/7 to 31 6/7 weeks’ gestation, and moderately preterm newborns were born from 32 0/7 to 36 6/7 weeks’ gestation.

An interaction analysis was performed if a compound effect between variables was suspected. Cox proportional hazard regression method was used to compare the index children to the referent children. Age-adjusted hazard ratios (HRs) with 95% confidence intervals (CIs) were calculated. P values <.05 were considered statistically significant. The analyses were performed by IBM SPSS Statistics version 23 (IBM SPSS Statistics, IBM Corporation). The trend of infant mortality during the follow-up was tested as 2-sided by the Cochran-Armitage trend test by using StatXact version 4.0.1.

The study was approved by the National Institute for Health and Welfare (the register keeper), which also authorized the use of health register data in scientific research (permission: 1202/5.05.00/2013; date of amendment: June 7, 2017), as required by national data protection legislation. Statistics Finland authorized the use of cause-of-death register data (permission: TK-53-1035-13).

The protocol for this study was approved by the ethics committee of Helsinki University Hospital (permission: 334/13/03/03/2013; date of amendment: June 15, 2017). Because we used routinely collected anonymized electronic health records, no formal patient consent was required for this study.

There were 1 121 244 singleton deliveries in Finland during the study period, with the incidence of placental abruption being 374 out of 100 000 (0.4%). A total of 280 deaths among index children and 107 deaths among referent children were recorded by the end of 2013. The median follow-up time was 18 years (interquartile range: 13–22; range: 0–27).

Baseline characteristics of the mothers with and without placental abruption in relation to offspring mortality are presented in Table 1. The results reveal that despite the mother’s age at delivery, marital status, socioeconomic status, or parity, the overall mortality was increased among children born after placental abruption (HR: 8.70; 95% CI 6.96–10.9). With Figure 1, we show that the cumulative hazard of death among the index children was higher compared with the referent children. The difference is evident already in the early postnatal period and increases toward the end of the follow-up. With Figure 2, we present the infant mortality as a 2-year moving average by year of birth among the index and referent children during the study period. The trend of infant mortality among the index children was decreasing (P < .001).

Detailed data of the interaction of placental abruption, maternal smoking, and offspring mortality in 4 different subgroups are presented in Table 2. Nonabruption, nonsmoking mothers were set as the reference group. Overall offspring mortality and mortality in all age groups were increased not only among smoking mothers with abruption, but also among smoking nonabruption mothers. However, although the point estimate for offspring mortality was higher among smoking women with abruption compared with nonsmoking women with abruption, the CIs were overlapping.

The mortality after placental abruption was increased during the neonatal period (<28 days), at the age of 28 to 365 days, and at the age of over 1 year. The major causes of death in both groups are listed in Table 3. By cause, the strongest increase in neonatal mortality was observed for birth-related asphyxia (HR: 108; 95% CI 34–341). Mortality was also increased because of prematurity, intracerebral hemorrhage, RDS or other respiratory disorders, anomaly, infection, and other causes. A list of causes of deaths that counted as anomaly or other cause is presented in the Supplemental Information. At the age of 28 to 365 days, there were 27 deaths among the index children and 9 deaths among the referent children. The mortality difference in this age category was significant (HR: 10.3; 95% CI 4.83–21.8). Among index children, there were 4 deaths due to sudden infant death syndrome (SIDS) and none among the referent children. Beyond the age of 365 days, a total of 23 deaths occurred among the index children and 47 among the referent children. Mortality of the index children was increased (HR: 1.7; 95% CI 1.03–2.79) compared with the referent children. For the index children, the median was 12.86 years (range: 1.64–24.05), and for the referent children, the median was 15.13 years (range: 1.50–24.23). Nearly half of all deaths of the referent children were due to external causes, whereas for the index children, these causes covered only one-fourth of all deaths. Deaths due to cancer were also more common among referent children than among index children.

Selected baseline characteristics of the children of both groups are listed in Table 4. In very preterm and extremely preterm children, the groups did not differ in terms of mortality. However, both moderately preterm and term infants in the index group had increased overall mortality (HR: 2.77; 95% CI 1.54–4.98 and HR: 4.98; 95% CI 3.54–6.99, respectively). Among term infants in the index group, the mortality remained increased up to the age of 1 year compared with the referent children. For birth weight, significant difference in mortality was observed only in index infants weighing 2500 g or more (HR: 5.94; 95% CI 4.33–8.14). The mortality remained increased up to the age of 1 year. For infants in the index group who were SGA, the mortality was increased during the neonatal period, whereas for infants in the index group who were LGA, both the neonatal and overall mortality were increased compared with the referent children. After placental abruption, boys had an increased overall mortality relative to girls (HR: 1.39; 95% CI 1.09–1.77; data not shown).

In this population-based case-control study of over 1 million births, we showed that children surviving after placental abruption had increased overall mortality compared with the referent children. Mortality was higher not only during the neonatal period but throughout the follow-up, and it was further increased by maternal smoking. Birth-related asphyxia was the leading cause of death.

Our study has several strengths and limitations. The study was based on national data obtained from validated and comprehensive health registers. The index and reference groups were large. However, the actual number of deaths in both groups was relatively small, which is a limitation concerning the conclusions to be drawn. For children who had died as neonates or infants, the numbers of deaths were small for several cause-of-death categories; this is another limitation. Moreover, we did not have information on the severity of the abruption. Finally, although the data were population-based and of high quality, this is a single-country study, limiting the generalizability of the results.

We have previously demonstrated that maternal smoking is an independent risk factor for perinatal mortality among infants born after placental abruption.⁴ In the current study, maternal smoking increased offspring mortality in both abruption-related and non–abruption-related births. As expected, the highest overall offspring mortality was found after placental abruption and maternal smoking. Our data can be used to suggest an additive effect with the combination of placental abruption and maternal smoking. However, the underlying mechanism for this additive effect is difficult to explain. Children born after abruption are certainly vulnerable even beyond the immediate perinatal time period and thus remain more susceptible to neonatal morbidity and mortality compared with children without abruption.¹² Maternal smoking may further increase this and simply make things worse.

Placental abruption is strongly associated with preterm delivery.⁵ In the analysis of cause-specific mortality, we observed that children born after placental abruption had an increased mortality because of prematurity and also because of intracerebral hemorrhage, RDS or other respiratory disorders, and infections, all entities closely associated with prematurity. There were no statistically significant differences in mortality between the index and referent children born extremely preterm or very preterm. However, referent children probably had other health problems that led to prematurity. Thus, the causal effect of placental abruption on increased mortality among premature children should be interpreted with caution.¹⁹ 

We have shown earlier that ∼15% of newborns with abruption are born with asphyxia (pH 7.05 or less) compared with 1.5% of newborns without abruption.²⁰ In this study, the role of birth-related asphyxia as a cause of death among children born after placental abruption was remarkable. The largest number of asphyxia-related deaths of the index children was recorded in the neonatal period, with an over 100-fold increase in mortality relative to the referent children. With this observation, we highlight the fact that birth-related asphyxia is a strong negative marker even if the child primarily survives. In this study, birth-related asphyxia remained a cause of death also in the older age groups. In addition, it may play a role in deaths due to intracerebral hemorrhage, cerebral palsy, and SIDS. Placental abruption has been linked to SIDS.^21,22 This was also found in our study. Deaths due to SIDS were more common among children born after placental abruption compared with the expected number of cases in the general population.¹⁷ It might be useful to counsel parents of newborns with asphyxia to follow the recommended safe sleep practice guidelines to reduce the risk of sleep-related infant deaths.²³ 

Nine index children died of cerebral palsy after the neonatal period. There were no such deaths among the referent children. However, some causes of death, such as cancer and external causes, were more common among the referent children relative to the index children at the age of 1 year or more. This observation may merely be a chance finding or result from the small number of observed cases.

We have previously shown that major congenital anomalies are twice as common among infants born after placental abruption than among referent infants.³ In this study, the neonatal mortality due to anomalies was nearly fourfold among the index children relative to the referent children.

To the best of our knowledge, there are no earlier studies in which researchers focus on placental abruption and subsequent offspring mortality, although the impact of abruption on perinatal mortality is well known.^4,5 Moreover, in a recent US study,¹² abruption was associated with a more than sevenfold increased risk of neonatal mortality. In our study, the HR for neonatal mortality was nearly twice as high. This difference can probably be explained by our study period being longer and starting earlier (1987–2005) than in the American study, in which the children were born more recently (2002–2008). Our study also allowed evaluation of mortality trends; a decreasing trend was observed for infant mortality over time. This observation is in line with our previous study revealing a decreasing trend in perinatal mortality.⁴ 

Both overall and cause-specific mortality in Finnish children born after placental abruption are increased. The impact of placental abruption on offspring mortality clearly extends far beyond the perinatal period.

CI

confidence interval

HDR

hospital discharge register

HR

hazard ratio

LGA

large for gestational age

MBR

medical birth register

RDS

respiratory distress syndrome

SGA

small for gestational age

SIDS

sudden infant death syndrome