Parts of this chapter are reproduced with permission from Langham et al.73 This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) licence, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/. The text includes minor additions and formatting changes to the original text.
Background
Pregnant women with pre-existing mental illness have an increased risk of adverse birth outcomes, including preterm birth and a baby being small for gestational age (SGA). The risk appears greatest in women with a more severe mental illness.74
The evidence of an increased risk of stillbirth and neonatal death is less consistent, potentially due to the lack of statistical power of previous studies that examine these outcomes in women with less common complex of severe mental illnesses. A further issue is the lack of recording, or inconsistent recording, of the severity of symptoms in routinely collected mental healthcare data sets.75
Aim
The aim of this study was to compare the risks of adverse pregnancy outcomes between women with and without a history of mental illness. The outcomes we considered included fetal and neonatal death, preterm birth, and the baby born SGA as well as composite indicators for neonatal adverse outcomes and maternal morbidity. We compared women with and without a history of mental illness, defined as a care episode with pre-pregnancy specialist mental healthcare services.76,77
Methods
Data sources and linkage
This study used three consecutive versions of the national data set of specialist mental health care provided by the NHS in England, which together include mental healthcare episodes from April 2006, linked at the patient level to the Hospital Episode Statistics (HES), the administrative database of all care episodes in general NHS hospitals78,79 and the birth notifications of Personal Demographic Service (PDS).80 Records on mental healthcare episodes provided between December 2015 and March 2016 were not available for technical reasons.
Hospital Episode Statistics records include patient demographics, admission dates, diagnoses according to International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes81 and procedures coded according to OPCS-4 codes.82 HES records of maternity episodes include additional information in the ‘HES maternity tail’.83 HES and PDS records of women giving birth were linked to HES records of their babies. PDS birth records contained additional information on stillbirth, gestational age and birthweight that was used if HES data were missing.
Cohort selection, comparison groups and risk adjustment
Maternity episodes for women, aged 18 and above, with a recorded gestation of at least 24 completed weeks, who had a singleton birth between April 2014 and March 2018 in the English NHS were identified from HES and PDS birth notifications (see Report Supplementary Material 1). Women who gave birth before 24 weeks were not included because there is no legal requirement to register babies born before 24 weeks of gestation without signs of life in the UK.84 Women who had a multiple birth were not included because their risk of adverse pregnancy outcomes are inherently increased.85
We considered that a woman had a pre-existing mental illness if there was evidence in the mental healthcare data sets that she had had a contact with specialist mental health care within 7 years before pregnancy. These specialist care contacts typically need a referral from a GP or they are part of urgent or emergency care.86 To determine the onset of pregnancy, we subtracted the gestational age at birth minus 2 weeks (or 38 weeks if gestational age was not available) from the date of birth.
We categorised the specialist mental healthcare contacts according to their highest level of care:
- an admission to a psychiatric ward, including generic psychiatric wards, MBUs or secure wards (‘hospital admission’; highest level)
- involvement of a specialist mental health crisis resolution team providing intensive treatment at home (‘CRT’; next highest level)
- and other care contacts with specialist mental health care, including day care and outpatient or community-based care (‘community care’; lowest level).
We grouped the timing of the most recent mental healthcare contacts (of any type) as:
- more than 5 years
- between 1 and 5 years
- less than 1 year before of pregnancy.
Hospital Episode Statistics records provided data on maternal age, parity, and previous caesarean section, ethnicity, pre-existing hypertension, pre-existing diabetes mellitus, pre-eclampsia/eclampsia and gestational diabetes (see Report Supplementary Material 1). Socioeconomic deprivation was derived from quintiles of the national ranking of the IMD 2019 of the women’s area of residence.83
Outcomes
Study outcomes across analyses were:
- stillbirth or neonatal death within 7 days of birth
- preterm birth (birth before 37 completed weeks of gestation)
- birth of a baby born SGA (birthweight less than the 10th centile using the UK-WHO gestationally corrected growth charts)87
- and two composite adverse outcome indicators that capture neonatal and maternal morbidity.
- The English Neonatal Adverse Outcome Indicator (ENAOI) for liveborn babies is derived from the presence of 15 ICD-10 diagnoses and 7 OPCS-4 procedures present in the babies’ HES records before inpatient discharge after birth (see Report Supplementary Material 1).76
- The English Maternal Morbidity Outcome Indicator (EMMOI) is derived from 17 ICD-10 diagnoses (including acute psychosis, but no other mental illness diagnosis, suicide, or substance misuse) and 9 OPCS-5 procedures, in HES records of the maternity episode (see Report Supplementary Material 1).77
For the purpose of this study, acute psychosis was not included.
Data analysis
All obstetric and neonatal outcomes follow accepted definitions. The definition of the comparison groups according to severity and timing of the pre-pregnancy mental healthcare contacts was guided by creating clinically relevant groups as much as possible equal in size. ORs adjusted for all maternal characteristics described above (adjORs) and their 95% confidence intervals (CI) were estimated with logistic regression with robust standard errors to account for clustering of outcomes within hospitals. We also tested the differences according to the severity and the timing of these contacts using joint Wald tests.
For regression analyses, missing values for outcomes and risk factors were imputed using chained equations to generate 10 data sets.88 Stata 17 was used for all analyses. A p-value of < 0.05 was considered to represent statistical significance. Where we adjusted for related factors, the covariates were: maternal age, parity and previous caesarean section, maternal ethnicity, socioeconomic deprivation, pre-existing diabetes mellitus and pre-existing hypertension, gestational diabetes and pre-eclampsia/eclampsia.
Results
Of the 2,081,043 women who were identified in the administrative hospital or birth registrations data with a singleton birth, 151,770 (7.3%) had at least 1 pre-pregnancy face-to-face contact with specialist mental health care within 7 years before pregnancy: 7247 (0.3%) had at least 1 admission to a psychiatric hospital, 29,770 (1.4%) had at least 1 CRT contact (and no hospital admission) and 114,753 (5.5%) had at least 1 specialist community contact (and no hospital admission or CRT contact).10
Table 11 shows the characteristics of the women according to the level and timing of their pre-pregnancy mental healthcare contact.
| All women | Presence of contact | By type of contact | By timing of contact | ||||||
|---|---|---|---|---|---|---|---|---|---|
| No contact | At least one | Community care contact | CRT | Hospital admission | > 5 years | 1–5 years | < 1 year | ||
| 2,081,043 | 1,929,273 | 151,770 | 114,753 | 29,770 | 7247 | 22,007 | 85,097 | 44,666 | |
| Maternal age (missing n = 28, < 0.01%) | |||||||||
| 18–24 | 365,554 (17.6) | 323,329 (16.8) | 42,225 (27.8) | 30,640 (26.7) | 9715 (32.6) | 1870 (25.8) | 2722 (12.4) | 22,627 (26.6) | 16,876 (37.8) |
| 25–34 | 1,250,692 (60.1) | 1,167,627 (60.5) | 83,065 (54.7) | 63,574 (55.4) | 15,552 (52.2) | 3939 (54.4) | 14,542 (66.1) | 47,174 (55.4) | 21,349 (47.8) |
| 35–39 | 374,489 (18.0) | 353,674 (18.3) | 20,815 (13.7) | 16,225 (14.1) | 3477 (11.7) | 1113 (15.4) | 3703 (16.8) | 12,063 (14.2) | 5049 (11.3) |
| 40 ≤ | 90,280 (4.3) | 84,619 (4.4) | 5661 (3.7) | 4312 (3.8) | 1024 (3.4) | 325 (4.5) | 1039 (4.7) | 3232 (3.8) | 1390 (3.1) |
| Parity (missing n = 91,523, 4.4%) | |||||||||
| Nulliparous | 870,661 (43.8) | 816,530 (44.3) | 54,131 (37.3) | 40,284 (36.8) | 11,226 (39.4) | 2621 (38.0) | 7067 (33.7) | 29,081 (35.8) | 17,983 (42.1) |
| Multiparous, no previous cesearean section | 931,397 (46.8) | 85,6325 (46.4) | 75,072 (51.8) | 57,113 (52.1) | 14,396 (50.5) | 3563 (51.7) | 11,370 (54.2) | 42,743 (52.6) | 20,959 (49.0) |
| Multiparous, with previous caesarean section | 187,462 (9.4) | 171,720 (9.3) | 15,742 (10.9) | 12,137 (11.1) | 2895 (10.2) | 710 (10.3) | 2523 (12.0) | 9413 (11.6) | 3806 (8.9) |
| Ethnicity (missing n = 292,757, 14.1%) | |||||||||
| White | 1,388,951 (77.7) | 1,268,938 (76.8) | 120,013 (88.1) | 91,285 (88.8) | 23,510 (87.1) | 5218 (80.8) | 17,237 (87.8) | 67,270 (88.1) | 35,506 (88.3) |
| South Asian | 204,298 (11.4) | 197,302 (11.9) | 6996 (5.1) | 5199 (5.1) | 1403 (5.2) | 394 (6.1) | 1038 (5.3) | 4084 (5.3) | 1874 (4.7) |
| Black | 84,198 (4.7) | 80,238 (4.9) | 3960 (2.9) | 2607 (2.5) | 886 (3.3) | 467 (7.2) | 601 (3.1) | 2126 (2.8) | 1233 (3.1) |
| Mixed | 31,957 (1.8) | 29,213 (1.8) | 2744 (2.0) | 1930 (1.9) | 618 (2.3) | 196 (3.0) | 398 (2.0) | 1481 (1.9) | 865 (2.2) |
| Other stated | 78,882 (4.4) | 76,399 (4.6) | 2483 (1.8) | 1724 (1.7) | 573 (2.1) | 186 (2.9) | 353 (1.8) | 1407 (1.8) | 723 (1.8) |
| Socioeconomic deprivation (missing n = 35, < 0.01%) | |||||||||
| Least deprived = quintile 1 | 313,034 (15.0) | 297,314 (15.4) | 15,720 (10.4) | 12,662 (11.0) | 2527 (8.5) | 531 (7.3) | 2829 (12.9) | 9108 (10.7) | 3783 (8.5) |
| Quintile 2 | 352,242 (16.9) | 331,487 (17.2) | 20,755 (13.7) | 16,348 (14.2) | 3624 (12.2) | 783 (10.8) | 3416 (15.5) | 12,009 (14.1) | 5330 (11.9) |
| Quintile 3 | 402,454 (19.3) | 375,299 (19.5) | 27,155 (17.9) | 20,776 (18.1) | 5185 (17.4) | 1194 (16.5) | 4097 (18.6) | 15,436 (18.1) | 7622 (17.1) |
| Quintile 4 | 471,646 (22.7) | 435,640 (22.6) | 36,006 (23.7) | 26,834 (23.4) | 7231 (24.3) | 1941 (26.8) | 5134 (23.3) | 19,923 (23.4) | 10,949 (24.5) |
| Most deprived = quintile 5 | 541,632 (26.0) | 489,510 (25.4) | 52,122 (34.3) | 38,124 (33.2) | 11,201 (37.6) | 2797 (38.6) | 6531 (29.7) | 28,615 (33.6) | 16,976 (38.0) |
| Pregnancy risk factors (missing = 91,523, 4.4%) | |||||||||
| Pre-existing diabetes | 16,786 (0.8) | 14,839 (0.8) | 1947 (1.3) | 1395 (1.3) | 430 (1.5) | 122 (1.8) | 278 (1.3) | 1045 (1.3) | 624 (1.5) |
| Pre-existing hypertensive conditions | 12,467 (0.6) | 11,504 (0.6) | 963 (0.7) | 738 (0.7) | 172 (0.6) | 53 (0.8) | 158 (0.8) | 545 (0.7) | 260 (0.6) |
| Gestational diabetes | 112,043 (5.4) | 104,021 (5.6) | 8022 (5.5) | 5891 (5.4) | 1630 (5.7) | 501 (7.3) | 1208 (5.8) | 4415 (5.4) | 2399 (5.6) |
| Pre-/eclampsia | 42,680 (2.1) | 39,618 (2.1) | 3062 (2.1) | 2279 (2.1) | 628 (2.2) | 155 (2.2) | 417 (2.0) | 1666 (2.1) | 979 (2.3) |
Table 12 shows that the risk of stillbirth or neonatal death within 7 days was 0.45% in women without and 0.49% in women with a pre-pregnancy specialist mental healthcare contact, but this increase was not statistically significant (adjOR 1.11, 95% CI 0.99 to 1.24, p = 0.078). When we used a model that included both the level and timing of the pre-pregnancy contacts, we did not find statistically significant evidence that risks were increased in women who had a community care contact more than 5 years before the pregnancy compared to women without a pre-pregnancy contact (p = 0.229). Neither did we find that the risks differed according to the level (overall p = 0.145) or timing (overall p = 0.267) of the pre-pregnancy contact. There was no statistically significant interaction between the level and timing of the pre-pregnancy contacts (p = 0.203).
| Numerator/denominator | Rate (%) | Unadjusted OR (95% CI) | p-value | Adjusteda OR (95% CI) | p-value | |
|---|---|---|---|---|---|---|
| All women | 9121/2,021,289 | 0.45 | ||||
| According to any previous specialist mental healthcare contact | ||||||
| 8396/1,873,823 | 0.45 | Reference | Reference | |||
| 725/147,466 | 0.49 | 1.10 (0.98 to 1.24) | 0.119 | 1.11 (0.99 to 1.24) | 0.078 | |
| According to type and timing | ||||||
| No contact | Reference | Reference | ||||
| Community care contact > 5 years | 1.09 (0.92 to 1.30) | 0.301 | 1.12 (0.93 to 1.35) | 0.229 | ||
| TYPE (reference: community care contact > 5 years) | ||||||
| CRT | 1.11 (0.94 to 1.31) | 0.094b | 1.10 (0.93 to 1.30) | 0.145b | ||
| Hospital admission | 1.34 (1.02 to 1.77) | 1.32 (1.00 to 1.75) | ||||
| TIMING (reference: community care contact > 5 years) | ||||||
| 1–5 years | 0.92 (0.76 to 1.11) | 0.210b | 0.91 (0.76 to 1.09) | 0.267b | ||
| < 1 year | 1.04 (0.83 to 1.30) | 1.02 (0.83 to 1.25) | ||||
| Interaction TYPE × TIMING | 0.200 | 0.203 | ||||
Table 13 shows that the risk of preterm birth was 6.5% in women without and 9.8% in women with a pre-pregnancy specialist mental healthcare contact (adjOR 1.53, 95% CI 1.35 to 1.73, p < 0.001). This risk gradually increased according to the highest level of mental healthcare contact from 9.2% in women who had a community care contact only (adjOR 1.44, 95% CI 1.25 to 1.66), 11.0% in women who had a CRT contact (adjOR 1.73, 95% CI 1.57 to 1.90), to 13.4% in women with a psychiatric hospital admission (adjOR 2.11, 95% CI 1.96 to 2.28, overall p < 0.001). We also found a gradual increase according to the timing of the most recent contact from 8.5% in women with a contact more than 5 years before pregnancy (adjOR 1.33, 95% CI 1.15 to 1.54), 9.3% in women with a contact between 1 and 5 years before pregnancy (adjOR 1.46, 95% CI 1.28 to 1.66), to 11.3% in women with a contact less than 1 year before pregnancy (adjOR 1.77 95% CI 1.58 to 1.97; overall p = 0.001). This overall pattern of results was consistent with that seen when we analysed the risk of preterm birth using a model that included both the level and timing of the pre-pregnancy mental healthcare contacts. There was no evidence of a statistical interaction between the level and timing of the pre-pregnancy contacts (p = 0.554).
| Numerator/denominator | Rate (%) | Unadjusted OR (95% CI) | p-value | Adjusteda OR (95% CI) | p-value | |
|---|---|---|---|---|---|---|
| All women | 136,899/2,039,580 | 6.7 | ||||
| According to any previous specialist mental healthcare contact | ||||||
| No contact | 122,343/1,890,694 | 6.5 | Reference | Reference | ||
| At least one previous contact | 14,556/148,886 | 9.8 | 1.57 (1.36 to 1.81) | < 0.001 | 1.53 (1.35 to 1.73) | < 0.001 |
| According to type and timing | ||||||
| No contact | Reference | Reference | ||||
| Community care contact; > 5 years | 1.32 (1.11 to 1.56) | 0.001 | 1.31 (1.12 to 1.52) | 0.001 | ||
| TYPE (reference: community care contact; > 5 years) | ||||||
| CRT | 1.18 (1.11 to 1.26) | < 0.001b | 1.16 (1.09 to 1.24) | < 0.001b | ||
| Hospital admission | 1.41 (1.25 to 1.59) | 1.38 (1.22 to 1.55) | ||||
| TIMING (reference: community care contact; > 5 years) | ||||||
| 1–5 years | 1.08 (1.02 to 1.15) | < 0.001b | 1.07 (1.01 to 1.13) | < 0.001b | ||
| < 1 year | 1.28 (1.21 to 1.37) | 1.26 (1.18 to 1.33) | ||||
| Interaction TYPE × TIMING | 0.746 | 0.554 | ||||
Table 14 shows that the risk of a baby born SGA was 6.2% in women without and 7.5% in women with a pre-pregnancy specialist mental healthcare contact (adjOR 1.34, 95% CI 1.30 to 1.37, p < 0.001). We found a gradual increase in risk according to the highest level of the mental healthcare contact from 7.3% in women with a community care contact (adjOR 1.30, 95% CI 1.26 to 1.34), 8.0% in women with a CRT contact (adjOR 1.39, 95% CI 1.32 to 1.46), to 9.5% in women who had a hospital admission (adjOR 1.64, 95%CI 1.51 to 1.79, overall p < 0.001). We also found a gradual increase according to the timing of the most recent contact from 6.4% in women with a contact more than 5 years before pregnancy (adjOR 1.21, 95% CI 1.14 to 1.28), 7.3% in women who had a contact between 1 and 5 years before pregnancy (adjOR 1.31, 95% CI 1.27 to 1.35), to 8.5% in women with a contact less than 1 year before pregnancy (adjOR 1.45, 95% CI 1.39 to 1.51, overall p < 0.001). Again, this pattern of results was consistent with that seen when we modelled the risk of a baby born SGA, including both the level and timing of the pre-pregnancy contacts (see Langham et al. 2023, table 4).73 There was no evidence of a statistical interaction between the level and timing of the pre-pregnancy contacts (p = 0.179).
| Numerator/denominator | Rate (%) | Unadjusted OR (95% CI) | p-value | Adjusteda OR (95% CI) | p-value | |
|---|---|---|---|---|---|---|
| All women | 127,968/2,030,048 | 6.3 | ||||
| According to any previous specialist mental healthcare contact | ||||||
| No contact | 116,829/1,882,019 | 6.2 | Reference | Reference | ||
| At least one previous contact | 11,139/148,029 | 7.5 | 1.23 (1.19 to 1.27) | < 0.001 | 1.34 (1.30 to 1.37) | < 0.001 |
| According to type and timing | ||||||
| No contact | Reference | Reference | ||||
| Community care contact; > 5 years | 1.03 (0.97 to 1.09) | 0.349 | 1.20 (1.13 to 1.27) | < 0.001 | ||
| TYPE (reference: community care contact; > 5 years) | ||||||
| CRT | 1.08 (1.03 to 1.14) | < 0.001b | 1.05 (1.00 to 1.10) | < 0.001b | ||
| Hospital admission | 1.25 (1.15 to 1.37) | 1.22 (1.12 to 1.33) | ||||
| TIMING (reference: community care contact; > 5 years) | ||||||
| 1–5 years | 1.13 (1.06 to 1.20) | < 0.001b | 1.07 (1.01 to 1.14) | < 0.001b | ||
| < 1 year | 1.31 (1.22 to 1.41) | 1.17 (1.09 to 1.26) | ||||
| Interaction TYPE × TIMING | 0.065 | 0.179 | ||||
The risk of neonatal adverse outcomes, defined using the composite neonatal outcome indicator, was 6.4% without and 8.4% with a pre-pregnancy specialist mental healthcare contact (adjOR 1.37, 95% CI 1.21 to 1.55, p < 0.001). The pattern of results according to the level and timing of the mental healthcare contacts followed that observed for the risk of preterm birth and babies born SGA (see Report Supplementary Material 1).
The risk of maternal morbidity (Table 15), defined using the composite maternal outcome indicator, was 0.9% without and 1.0% with a pre-pregnancy specialist mental healthcare contact (adjOR 1.18, 95% CI 1.12 to 1.25, p < 0.001). We also found evidence of an increase in this risk according to the highest level of the mental healthcare contact (overall p < 0.001) or according to the timing of the most recent contact (overall p < 0.001). The pattern of results was less clear if we included both the level and timing of the pre-pregnancy contacts in the same model (see Report Supplementary Material 1).
| Numerator/denominator | Rate (%) | Unadjusted OR (95% CI) | p-value | Adjusteda OR (95% CI) | p-value | |
|---|---|---|---|---|---|---|
| All women | 17,968/1,989,520 | 0.9 | ||||
| According to any previous specialist mental healthcare contact | ||||||
| No contact | 16,528/1,844,575 | 0.9 | Reference | Reference | ||
| At least one previous contact | 1440/144,945 | 1.0 | 1.11 (1.05 to 1.17) | < 0.001 | 1.18 (1.12 to 1.25) | < 0.001 |
| According to type and timing | ||||||
| No contact | Reference | Reference | ||||
| Community care contact; > 5 years | 1.00 (0.87 to 1.15) | 0.976 | 1.07 (0.94 to 1.23) | 0.312 | ||
| TYPE (reference: community care contact; > 5 years) | ||||||
| CRT | 0.92 (0.81 to 1.06) | 0.374b | 0.90 (0.79 to 1.03) | 0.294b | ||
| Hospital admission | 1.09 (0.86 to 1.38) | 1.01 (0.79 to 1.28) | ||||
| TIMING (reference: community care contact; > 5 years) | ||||||
| 1–5 years | 1.11 (0.95 to 1.29) | 0.033b | 1.11 (0.96 to 1.29) | 0.046b | ||
| < 1 year | 1.21 (1.04 to 1.42) | 1.21 (1.02 to 1.41) | ||||
| Interaction TYPE × TIMING | 0.429 | 0.253 | ||||
Discussion
This study of about 2 million women who gave birth to a singleton baby in the English NHS between 2014 and 2018. Of these women, 7.3% had at least one face-to-face contact with specialist mental health care in the 7 years before their pregnancy.
We used the highest level of specialist mental health care as a proxy for the severity of the disorders and the timing of the most recent pre-pregnancy specialist mental healthcare contact to distinguish pregnant women’s mental healthcare histories.
We found that about 1 in 14 women had at least one pre-pregnancy specialist mental healthcare contact and that in these women; the risk of adverse obstetric and neonatal outcomes was increased, even if they only had specialist mental health care in a community setting more than 5 years before the pregnancy. Risks were higher in women with a pre-pregnancy specialist healthcare contact that reflects a more severe mental illness (a psychiatric hospital admission or involvement of a CRT), and in women who had a more recent pre-pregnancy specialist mental healthcare contact. The risk of maternal morbidity was only slightly increased in women who had a pre-pregnancy specialist mental healthcare contact.
Strengths and limitations
The major strength of this study is that we could use linked national data sets of care provided by the English NHS. We could trace face-to-face contacts with specialist mental health services for all women up to 7 years before their pregnancy, but this look-back period was shorter in younger women because we did not have access to adolescent mental health records.
A limitation of our study includes that because of technical issues experienced by the national data provider we did not have access to records of specialist mental healthcare episodes for a 4-month period between December 2015 and March 2016. Therefore, we will have misclassified a proportion of women whose highest level of specialist mental healthcare contact occurred during that period. It is unlikely that this period without access to mental healthcare records will have had a major impact, given that the number of women whose mental healthcare history in the year immediately before pregnancy is affected is relatively low, the duration for those whose mental health history between 1 and 5 years before the onset for the pregnancy is affected is relatively short, and most women with pre-pregnancy mental illness will have had multiple contacts.
Previous studies have described the prevalence of common and severe mental illness during pregnancy and the year after birth,9,74 but there is little systematic evidence on the prevalence of mental illness before the onset of pregnancy. Other new key findings are that the risks of preterm birth and a baby born SGA were increased for all women who had a pre-pregnancy contact with specialist mental health care, even if they had only experienced a community contact more than 5 years before the pregnancy.
A recent meta-analysis of 28 studies reported increases of 40–50% in the risk of stillbirth and neonatal mortality, irrespective of the type of mental disorder that was studied and the perinatal outcomes considered, but with considerable heterogeneity between studies.89 We did not find an overall increase in the risk of stillbirth and neonatal mortality considering all women with a pre-pregnancy specialist mental healthcare contact, but the overall increases in the risk of other adverse outcomes, including preterm and SGA birth, were found to be in a similar range as reported in the meta-analysis.
Implications for practice and research
Our results suggest adding questions about whether women who are planning a pregnancy or who have recently learnt they are pregnant have ever had contact with specialist mental healthcare services to the initial obstetric risk assessment, which already includes questions about current medication, smoking, alcohol, illicit drug use and other adverse conditions. Of note, questions about the highest level of the mental health care and the timing of the most recent contact provide an opportunity to identify women at the highest risk.
The results described in this chapter underpin the urgency of an impact of the specialist CPMHTs not only on mental outcomes during and the year after pregnancy but also on obstetric and neonatal outcomes. In many countries, new maternity care models, such as the midwifery caseload model, have been introduced that strengthen continuity of carer, especially for women with complex medical and social needs, which present opportunities for risk assessment and referral, for example through pregnancy smoking cessation services.90–92 However, to guide the prevention of adverse obstetric and neonatal outcomes, we need a detailed causal mediation analysis to gain a better understanding of the complex pathways that explain the increases in the risks in the women who had a specialist mental healthcare contact before the onset of pregnancy.9