Stillbirth rates in the UK have consistently been among the highest in Europe (Flenady et al, 2011). One contributory factor may be the heterogeneity of the UK population, including the raised perinatal mortality rate of migrant mothers, of whom the overall proportion of births in the UK is increasing. One in four (24%) babies born in 2011 was to non-UK-born women, and this trend looks likely to continue (Zumpe et al, 2012).
In a recent study of stillbirths (excluding congenital anomalies) in the West Midlands, mothers born outside the UK had a stillbirth rate of 5.2 per 1000 births, compared to mothers born in the UK who had a rate of 3.4 per 1000 births (Gardosi et al, 2013). This difference also applied within individual ethnic origin categories: migrant mothers from Pakistan had a higher stillbirth rate (6.9/1000) than Pakistani mothers born in the UK (4.1/1000).
In 2009, the West Midlands had the highest regional perinatal mortality rate in England, at 8.0 per 1000 total births (Centre for Maternal and Child Enquiries (CMACE), 2011). Birmingham Health and Wellbeing Partnership therefore commissioned the West Midlands Perinatal Institute to conduct a confidential enquiry into perinatal deaths of migrant mothers, to increase understanding of the high mortality rates.
Confidential enquiries (CEs) are multi-disciplinary case reviews, which have been used to assess cases of adverse perinatal outcome both nationally and regionally (Centre for Maternal and Child Enquiries, 2011; Lewis, 2011; West Midlands Perinatal Institute, 2010; 2011). They are an acknowledged method of examining the standard of care and avoidability of outcome, including system failures and external factors. The objectivity of the case reviews allows the identification of areas for improvement within maternity services for clinicians, health service providers and commissioners.
The purpose of our study was to investigate the factors that may have contributed to this high rate of adverse outcomes in pregnancies of migrant mothers in our population.
Methodology
Aim
The aim was to undertake a CE using peer review of perinatal deaths that occurred during 2008 in three large hospital Trusts in the West Midlands. All Trusts were urban-based and used to high numbers of migrant mothers (none employed a specialist midwife for vulnerable women at the time of the study, although all provided specialist support for safeguarding issues).
Inclusion criteria
Inclusion criteria were stillbirths from 24 weeks' gestation to term and early neonatal deaths (babies that died at less than 7 days old) from 34 weeks' gestation to term, to mothers who themselves were born outside of the UK but who received the majority of their maternity care and gave birth within the West Midlands. The gestation criterion was chosen to exclude babies that were less likely to die due to prematurity. Babies with congenital anomalies were included as long as they fulfilled the inclusion criteria.
Consent
Informed maternal consent was obtained as part of the routine booking process for maternity care within the three Trusts. All mothers agreed for anonymised details from their case notes to be used for audit and confidential case review. All mothers gave consent. Ethical approval was not required as the study did not constitute research but was part of a clinical audit and service evaluation. Each Trust agreed to provide individual case information and receive feedback following the reviews.
Sample
The first 36 deaths that occurred in 2008 and fulfilled the inclusion criteria were allocated for case review. (The sample size of 36 cases was anticipated to be the total number from the three Trusts for the 12-month period.) The cases comprised 30 stillbirths and six early neonatal deaths, and included 18 normally formed babies and 18 with congenital anomalies.
Process
Copies of relevant case notes were obtained, anonymised of all identifiable patient, hospital and staff information, and examined by CE panels using a standardised, purpose-designed proforma that had been used in other CEs in our region. The proforma was a tool designed to record the findings of the case note review after examining all aspects of the antenatal, social, intrapartum, postnatal and neonatal care.
Eleven multidisciplinary panels were constituted to review the 36 cases, which meant that on average three cases were discussed per panel. Each panel consisted of senior clinicians including two obstetricians and two midwives for stillbirths, and for neonatal deaths also two neonatologists and two neonatal nurses, as well as a health visitor and/or public health specialist. Each panel had an independent chair (JG) and was assisted by specialist midwives from the West Midlands Perinatal Institute (FCS and MW). Panellists only reviewed cases from outside their own Trust. On average, each case took an hour for review, discussion and comment on the standard of care and avoidability of the outcome. In all instances, findings were determined by consensus.
Risk assessment form
Prior to the panel meetings, a short risk assessment form was sent to the mothers' GPs and named community midwives (CMs) regarding general perception of social, psychological and medical risk factors for each individual case. It was sent as soon as possible after the death but in some cases, this was up to 6 months afterwards. Most midwives used the Trust case notes and GPs utilised their own case notes to remind themselves of the case details before completion of the form. This was in order to compare assessments and follow-up of identified risk from different health professionals compared to that of the Trust case notes as seen by the panel.
Findings
Cohort
The characteristics of the 36 cases are detailed in Table 1 and Table 2. Mothers were documented as being from a number of different countries, with the majority (n=23, 64%) from South Asia. Nearly half (n=17, 47%) of the deliveries were at term, while a quarter (n=9, 25%) were at very early gestation (<29/40 weeks). Eighteen (50%) deaths (12 stillbirths and six early neonatal deaths) were of babies with congenital anomalies including renal anomalies (n=5, 14%) neural tube defects (n=4, 11%), chromosomal (n=4, 11%) and miscellaneous (n=5, 14%) categories. A post-mortem was carried out on a third of the babies (n=12, 33%).
| Place of ethnic origin | Stillbirths n=30 | Early neonatal deaths n=6 | Total n=36 |
|---|---|---|---|
| South Asia total | 23 | ||
| Pakistan | 13 | 2 | 15 |
| India | 3 | 1 | 4 |
| Bangladesh | 3 | 1 | 4 |
| Africa total | 8 | ||
| Nigeria | 2 | 0 | 2 |
| Somalia | 2 | 0 | 2 |
| Ethiopia | 1 | 0 | 1 |
| Zimbabwe | 1 | 0 | 1 |
| Mali | 1 | 0 | 1 |
| Sudan | 0 | 1 | 1 |
| Europe total | 3 | ||
| Poland | 1 | 0 | 1 |
| Romania | 0 | 1 | 1 |
| Middle East total | 2 | ||
| Iraq | 2 | 0 | 2 |
| Australia | 1 | 0 | 1 |
| Details | Stillbirth n=30 | Early neonatal deaths n=6 | Total n=36 | |
|---|---|---|---|---|
| Maternal age | <20 | 0 | 1 | 1 |
| 20–29 | 13 | 4 | 17 | |
| 30+ | 17 | 1 | 18 | |
| Parity | Primigravida | 9 | 1 | 10 |
| Grand-multiparity (>4) | 6 | 1 | 7 | |
| Gestation at booking | <13 weeks | 27 | 6 | 27 |
| 13–20 weeks | 2 | 0 | 2 | |
| 21+ weeks | 1 | 0 | 1 | |
| Body mass index | <20 kg/m2 | 2 | 0 | 2 |
| >30 kg/m2 | 7 | 2 | 9 | |
| Smoking | Yes | 0 | 0 | 0 |
| No | 30 | 6 | 36 | |
| Gestational age | <29 weeks | 10 | 0 | 9 |
| 29–36 weeks | 7 | 2 | 9 | |
| 37+ weeks | 13 | 4 | 17 | |
| Birth weight | <10th customised centile | 19 | 2 | 21 |
| >90th customised centile | 2 | 3 | 5 | |
| Consanguinity* | Congenital anomaly | 3 | 1 | 4 |
| Normally formed | 6 | 0 | 6 | |
| Post-mortem done | Yes | 10 | 2 | 12 |
| No | 20 | 4 | 24 |
Risk factors
According to the panel reviews, 28/36 (78%) of mothers had at least one medical risk factor (Table 3) and 26/32 (81%) had at least one social risk factor identified (Table 4). However, it was noted that four cases had incomplete or no social assessment documented for the pregnancy. All cases, except for one, had at least one social and/or medical risk factor documented and 4/32 (13%) mothers had three or more of the social risk factors identified.
| Medical risk factors | Number (n=36) | % |
|---|---|---|
| Cases with one or more medical/obstetric risk factor | 28 | 78% |
| Poor obstetric history | 11 | 31% |
| Significant pre-existing medical condition | 8 | 22% |
| Obesity (body mass index >30 kg/m2) | 7 | 19% |
| Grand-multiparity (>4) | 7 | 19% |
| Previous intrauterine growth restriction | 4 | 11% |
| Previous stillbirth | 2 | 6% |
| Social risk factors | Number (n=32*) | % |
|---|---|---|
| Cases with one or more social risk factor | 26 | 81% |
| Non-English speaking | 23 | 72% |
| Unemployed household | 8 | 25% |
| Inappropriate housing e.g. overcrowded, temporary accommodation | 6 | 19% |
| Asylum seeker/new to UK (within 1 year of pregnancy) | 5 | 16% |
| Absent husband/difficult family circumstances | 5 | 16% |
| Significant psychological/emotional factors | 4 | 13% |
| Moved regions during pregnancy | 2 | 6% |
| Needed help claiming benefit | 2 | 6% |
| Not turning up for appointments (>1 Did Not Attends (DNAs)) | 2 | 6% |
In total, 63/72 risk assessment forms were completed: six named CMs and GPs stated that they did not know the woman and three declined to complete the form (Table 5). Of the forms returned, high social risk was only identified by CMs in six and by GPs in seven mothers, in comparison to 26 cases by the panel. Likewise, high medical risk was only identified in one woman by each of the primary care providers compared to 28 by the panel.
| Risk assessment form response | Identified risk (n=36) | ||
|---|---|---|---|
| Community midwife | GP | Panel | |
| No social or medical risk identified | 8 | 13 | 1 |
| Noted family/partner support | 16 | 11 | - |
| Noted high social risk | 6 | 7 | 26 |
| Noted high medical risk | 1 | 1 | 28 |
| Did not know woman/declined to complete | 5 | 4 | - |
Conversely, three risk assessments revealed significant information that was not documented elsewhere in the case notes (e.g. domestic abuse, asylum seeker). However, only one mother was referred for additional services or support.
Grading
Each case review examined all aspects of care, culminating in a grading of avoidability of outcome based on the standard Confidential Enquiry into Stillbirths and Deaths in Infancy (CESDI) criteria (Table 6). Of the 18 deaths of normally formed babies, nine (50%) were considered to have been possibly (grade 2) or probably avoidable (grade 3), with only four (11%) of all the cases given a grade 0 by the panel, i.e. to have had no suboptimal care. Most of the deaths with congenital anomalies were believed to be unavoidable; however, suboptimal care was present in 16 (89%) of the 18 cases. Additionally, eight (22%) mothers were considered by the panel to have not sought help when needed (e.g. when experiencing reduced fetal movements or ruptured membranes), although it is recognised that the mothers may not have understood that they needed to seek assistance.
| CESDI grade | Level of care | Congenital anomaly | |
|---|---|---|---|
| No | Yes | ||
| 0 | No suboptimal/substandard care | 2 (11%) | 2 (11%) |
| 1 | Suboptimal care, but different management would have made no difference to the outcome | 7 (39%) | 13 (72%) |
| 2 | Suboptimal care; different care might have made a difference (possibly avoidable death) | 5 (28%) | 3 (17%) |
| 3 | Suboptimal care; different care would reasonably be expected to have made a difference (probably avoidable death) | 4 (22%) | 0 (0%) |
| Total | 18 (100%) | 18 (100%) | |
CESDI–Confidential Enquiry into Stillbirths and Deaths in Infancy
Suboptimal care themes
Panels assessed the standard of care and identified examples of good practice (e.g. continuity of named carer in both antenatal and postnatal periods, extensive bereavement care, comprehensive postnatal follow-up) as well as examples of substandard care (Table 7).
| Key themes | Number of cases in which this was the primary issue | Examples of poor clinical practice (as primary or secondary issue) |
|---|---|---|
| Pregnancy care | 10* | Routine observations or investigations not done |
| Poor continuity in antenatal and/or postnatal period | ||
| Missed appointments and abnormal results not followed up | ||
| Risk assessment and/or management at booking | 5* | Incomplete booking information in pregnancy hand-held records |
| Inadequate/absent medical and social management plans of care | ||
| Appropriate management plans not being carried out | ||
| Social and/or psychological care | 5 | Lack of support after diagnosis of fetal abnormality |
| Significant delay confirming fetal death | ||
| Inadequate bereavement care or support | ||
| Interpretation and/or communication with mother | 3* | Lack of professional interpreter for key appointments |
| Partner, friend or family member being encouraged to interpret, particularly at difficult times e.g. at diagnosis of stillbirth | ||
| No documented discussion of importance of fetal movements | ||
| Multidisciplinary team communication | 3* | Inadequate neonatal preparation for the birth of babies with congenital anomalies e.g. resuscitating a baby with a known lethal abnormality |
| Poor information-sharing or referrals for mothers who birthed at a different hospital to where they received antenatal care or those with significant social issues e.g. domestic abuse | ||
| Identification and/or management of intrauterine growth restriction | 3* | Lack of serial scans for mothers with previous intrauterine growth restriction |
| Not acting on intrauterine growth restriction identified on serial scans | ||
| Postnatal follow-up | 2 | Absence of community care after discharge from hospital following birth |
| Lack of consultant follow-up appointments or plans for future pregnancies | ||
| Intrapartum care | 1 | Not acting on pathological cardiotocographs of babies with congenital anomalies |
| Mothers left alone for significant periods | ||
| Inadequate pain relief during labour | ||
| Placentas of stillborn babies not being sent for histology |
CESDI–Confidential Enquiry into Stillbirths and Deaths in Infancy
From each case that the panel considered to have received suboptimal care (n=32), a key theme was identified; if care had been appropriate for the 12 cases given a CESDI grade 2 or 3, there may also have been a different outcome. The most frequently occurring themes were inadequate pregnancy care (n=10); poor management and assessment of risk (n=5); inadequate and inconsistent psychological care (n=5); poor communication (n=6) either between the clinician and woman and her family, or between clinicians themselves; and poor identification or management of intrauterine growth restriction (n=3).
Lack of professional translation services at key times was evident in 19/23 (83%) cases where the mother did not speak English, and even when English was spoken, the panel noted that women appeared to be given little or no information about options for their care (e.g. pregnancy screening tests), management planning decisions, pregnancy advice (e.g. when to seek help for complications) and post-death screening resulting in a low post-mortem uptake (n=12, 33%).
Feedback to the Trusts
When the review panels had all been completed, a copy of each case review was requested from the Trusts so that the internal review could be compared to the independent multidisciplinary review (Table 8). In 21 (58%) of the cases, there had been no internal review. Of the cases that had been reviewed, only one (3%) had had a root cause analysis. The case summaries from the independent panels were then given to the relevant Trusts.
| Type of internal case review | Number (n=36) |
|---|---|
| Case review/summary | 3 |
| Root cause analysis | 1 |
| West Midlands Perinatal Institute proforma | 11 |
| Total reviewed | 15 |
| No review | 21 |
Discussion
The purpose of our study was to investigate the factors that may have contributed to the high perinatal mortality rate of migrant mothers, and the findings have highlighted a number of important care issues. This is, to our knowledge, the first study that links avoidable perinatal deaths to substandard care in pregnancies of migrant mothers.
Additionally, we found during the case feedback process that the internal review findings for these cases showed that Trusts were not routinely reviewing all perinatal deaths. While the Department of Health (DH, 2012) and the Royal College of Obstetricians and Gynaecologists (RCOG, 2010) have recommended that each perinatal death is reviewed by a multidisciplinary panel in a standardised way in order to evaluate care and make improvements where necessary, not all perinatal deaths appear to be reviewed, and those that are appear to be reviewed in a variety of ways. This supports findings of other regional CEs (West Midlands Perinatal Institute, 2010; 2011).
For the congenital anomaly cases, the panels identified that the majority were not preventable and instead highlighted issues relating to the poor care; 89% (n=16/18) of cases with congenital anomalies had had some degree of suboptimal care. As perinatal CEs usually exclude congenital anomaly cases, this was a new finding and demonstrates that, regardless of the avoidability of the outcome, the standard of care needs to improve.
Among normally formed babies, panels considered that as many as half of these deaths were potentially avoidable. Many of these instances were linked to inadequate risk assessment, poor social care and communication, and poor identification of intrauterine growth restriction, some of which were also described in CEs of UK-born mothers. (West Midlands Perinatal Institute, 2007; 2010; CMACE, 2011; Lewis, 2011). However, there was also a preponderance of medical and social risk factors which, in many instances, primary health-care providers failed to recognise. Furthermore, poor communication between the woman and her multidisciplinary team contributed to the lack of appropriate action for significant issues, highlighting potential barriers to appropriate maternity care for this vulnerable group.
Documented barriers for migrant women include lack of knowledge to access and navigate the maternity care system, communication, generalised poor care by health professionals who may have different cultural and religious practices, and poverty (Taylor and Newall, 2008; Phillimore et al, 2010; Raleigh et al, 2010; Henderson et al, 2013).
Communication is the key to effective care and women who do not speak English are recommended to have a professional interpreter (National Institute for Health and Care Excellence, 2010). However, as in other studies, this was not always provided, or information when given was poorly understood (Cross-Sudworth et al, 2011; Mumtaz et al, 2014). Minority ethnic mothers born in the UK and who are more culturally ‘British’, however, appear to receive better communication (Puthessey et al 2008; 2010). The panel considered that if the mothers who delayed seeking care for diminished fetal movements had clearly understood why it was important and acted promptly (some waited 3–4 days), some outcomes may have been different.
Jonkers et al (2011) suggest that even when migrant mothers are able to speak English and are highly educated, they are not always health-literate so may be unable to act when there is a cause for concern to themselves or the baby. For example, needing permission to attend appointments or for screening tests; not having the means to access care on their own; and an inability to communicate concerns to health care. Effective communication, regardless of ethnic background, requires time and sensitivity to religious and cultural issues as well as appropriate language.
In addition to communicating with the mothers, this study also found that communication within the multidisciplinary team was an issue. Important social as well as medical information was not passed between GPs, CMs and hospital despite recommendations supporting its importance (Lewis, 2011), for example, concerning asylum seekers and women experiencing domestic abuse or significant housing issues. As these social issues increase perinatal morbidity and mortality risks, this needs to be identified and relevant information appropriately shared with key care providers.
Health professionals in the study, however, appeared reluctant to both identify social risk and make referrals. This may be due to a lack of confidence regarding what to do or a fear of misunderstanding someone else's cultural or religious norms if the mother is from a different ethnic, religious or cultural background than that of the health professional. To address both of these issues, multidisciplinary social care and cultural competency training is needed, alongside clear interprofessional care pathways (Taylor and Newall, 2008; Lewis, 2011; Phillimore et al, 2010). Research is required to evaluate the outcomes of any training on both staff and mothers, as well as to understand why health professionals may be under-identifying social risk factors.
While CMs and GPs are the key health professionals caring for migrant mothers and are often the gatekeepers to other support and specialist input during the perinatal period, it is the combined effort of the multidisciplinary team that will improve migrant mothers' care within maternity services and may have an impact on the perinatal mortality rate. Basic measures that may make a difference in care for migrant and vulnerable mothers include continuity of carer and increasing the time for appointments, particularly when translation is needed. However, with time and resources increasingly squeezed, these basic factors—as well as meeting the training and support needs of the health professionals themselves—appear to be difficult to achieve.
Conclusion
The findings of this CE from one NHS region may not all be transferable to other settings. Nevertheless, the study raises important questions for all maternity service providers, particularly regarding the effect of communication on outcome. The findings of this study into perinatal deaths of migrant mothers support the need for better communication with mothers and other health professionals, as well as the need for internal case reviews following perinatal deaths.
This study adds to existing knowledge that migrant mothers experience many social and medical risk factors that are pertinent to perinatal outcome. In many instances these are missed, and the situation is aggravated by insufficient communication both with the mother and within the multidisciplinary health-care team. Prevention and enhanced quality of care requires a recognition of the special needs of migrant mothers during pregnancy, and sufficient resources, training and education to ensure that the health service can meet these needs.
Key Points
Conflict of interest: The authors have declared no conflict of interest.