Minimally invasive autopsy for fetuses and children based on a combination of post-mortem MRI and endoscopic examination: a feasibility study

Article history

The research reported in this issue of the journal was funded by the HTA programme as project number 14/168/02. The contractual start date was in May 2016. The draft report began editorial review in June 2018 and was accepted for publication in February 2019. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

none

Disclaimer

This report contains transcripts of interviews conducted in the course of the research and contains language that may offend some readers.

Copyright statement

© Queen’s Printer and Controller of HMSO 2019. This work was produced by Lewis et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2019 Queen’s Printer and Controller of HMSO

Introduction

In England and Wales, there are around 3000 terminations of pregnancy for fetal abnormality per annum and there are estimated to be at least 3000 miscarriages at 12–24 weeks’ gestation. 1 From approximately 700,000 births per annum, around 3000 result in stillbirth (SB), 2000 in neonatal death and 800 in infant death (of which 250 are unexplained). There are also around 50 unexpected child deaths (children aged 1–14 years) annually. 2–4 In those cases in which a coronial autopsy examination (otherwise known as a post-mortem) is not required, parents may be offered an autopsy examination to try to establish the cause of death and, when appropriate, to estimate the risk of problems recurring in future pregnancies. Standard autopsy examination results in clinically significant findings in 22–76% of cases, depending on the type of loss. 5,6 Without an autopsy, many parents are left with no understanding of the disease process related to their child’s death, and regret for some parents who decline autopsy has been identified. 7 Despite this, consent rates for prenatal, perinatal and paediatric autopsy have dropped significantly in the UK,7–9 Western Europe10,11 and the USA12,13 in the past 30 years. In the UK, less than half of parents of stillborn babies and only one-quarter of parents of neonates who die provide consent for standard autopsy examination. 9

Autopsy examination as a component of investigation of prenatal and paediatric death

The perinatal autopsy represents part of a specialist obstetric, fetal medicine or paediatric service and current guidelines recommend that the option of an autopsy, performed by specialist paediatric/perinatal pathologists, should be offered for all perinatal deaths, including miscarriages, SBs, termination of pregnancy for fetal abnormality (ToP) and neonatal deaths. 14 A sudden unexpected death in infancy (SUDI) or a sudden unexpected death in childhood (SUDC) require referral to the Her Majesty’s (HM) Coroner and, in most cases, a coronial autopsy is indicated.

Functions of perinatal and paediatric autopsy

The aim of investigation after death is to provide information for a number of functions, ranging from information relevant to the parents that may influence further management, through to providing data for clinical governance and service development.

Contribution of the perinatal and paediatric autopsy to clinical care

In both perinatal and paediatric practice, with reduced exposure of doctors in training to autopsy, there is a perception that with improvements in medical care the relevance of investigation after death is declining; however, published data suggest otherwise. A systematic review17 reported substantial discrepancies between clinical diagnoses and autopsy findings in many cases. ‘Major errors’ (previously unrecognised diagnosis of the cause of death) were reported in around one-quarter of cases, with 10% ‘class I errors’, defined as previously unknown conditions that may have affected patient outcome had they been diagnosed in life. 17 A further meta-analysis demonstrated a 15–40% discrepancy for major diagnoses and a 30–60% discrepancy for cause of death in adults, with 45–75% of autopsies demonstrating at least one additional finding. 18 Similar data are available for perinatal autopsies specifically, with a review of 27 studies reporting that perinatal autopsy resulted in a change in diagnosis or additional findings that might have influenced management or counselling in 22–76% of cases. 5 In this study, 35% of antenatal ultrasound diagnoses were modified by additional autopsy findings. If only studies reported in the last decade are included, significant additional information required to change the underlying diagnosis or counselling is reduced to around 10%; however, this still represents a significant proportion, with a detection rate far above the majority of investigations performed in routine critical practice.

The process of autopsy involves several procedures and the data presented in most studies do not allow information to be obtained regarding which component of the investigation provided the most useful additional information. A change in perspective from considering autopsy examination as one investigation, to a more targeted investigation after death, using an algorithmic approach, would allow optimal resource use to target those in whom the various components of the autopsy are likely to be most contributory. Furthermore, in some clinical circumstances, although an autopsy investigation may provide information which may be useful for the purposes listed above, such as governance and quality assurance, in these settings the underlying diagnosis and counselling of parents is unchanged.

Factors influencing the value of the perinatal and paediatric autopsy

Several factors influence the usefulness of autopsy, including the type of antenatal or antemortem care; the accuracy and availability of previous investigations; and the level of clinical expertise, which may all affect the likelihood of identifying additional findings. Autopsies performed by specialist perinatal or paediatric pathologists in tertiary centres are significantly more likely to provide useful additional information than those performed by general pathologists in this clinical setting. In addition, in highly complex cases, pathologists with specific expertise in, for example, complex congenital cardiac disease or neuropathology, may be required for optimal interpretation.

Various technical variables may limit the pathologist’s ability to identify abnormalities. The most important of these are the effects of secondary changes following fetal death, either in utero (maceration) or during the delivery to post-mortem interval. For example, 10–20% of antenatally detected brain abnormalities cannot be confirmed at autopsy due to the degree of maceration and/or post-mortem autolysis. 19 It is therefore important that if an autopsy is to be performed, the body be refrigerated and the procedure performed as soon as possible after death. The number and type of ancillary investigations performed (e.g. example microbiological investigations or genetic testing) will also affect yield.

From a practical perspective, rapid turnaround times and high-quality autopsy reports that include a comprehensive, inclusive and directed clinicopathological correlation, improve the usefulness of perinatal autopsy for contributing to patient management.

Indications of particular value and of limited value

Based on available data and empirical practice, the potential value of autopsy examination to contribute to future patient management is dependent on the clinical circumstances. For example, in termination of pregnancy for prenatally diagnosed chromosomal abnormalities, although autopsy may provide confirmation and allow detection of subtle features, which may contribute to detailed understanding, the underlying diagnosis and, hence, patient management, is unlikely to be altered. In contrast, in the case of a patient undergoing termination for a prenatally diagnosed abnormality which may have a wide range of underlying aetiologies, such as central nervous system malformations and cystic renal disease, autopsy investigation is highly likely to provide a specific diagnosis which cannot be achieved by other means. Other examples include prenatally diagnosed skeletal dysplasias, in whom the specific type of dysplasia, relevant for management of future pregnancies and prenatal testing, is often not determined until after death.

It is impossible to provide a definitive list of indications representing ‘high’ or ‘low’ autopsy yields, as the individual circumstances, clinical questions, and availability of specialist antenatal investigations and pathologists, will vary between centres. Nevertheless, it is suggested that this process is performed locally in order to maximise use of resources and to ensure that expectations of both parents and clinicians are realistic and evidence based.

Classification systems of perinatal and paediatric deaths and their issues

For the purposes of epidemiological studies and research, a range of classification systems have been used in order to categorise the type of perinatal and paediatric death and/or the underlying pathophysiological mechanism involved. No universally accepted classification system exists. Those in use range from simple systems with only a few generic categories (e.g. Wigglesworth and Singer20), to more recent and complex systems which attempt to take account of not only autopsy findings, but also any potentially relevant conditions from the clinical history or other investigations [e.g. Relevant Condition at Death (ReCoDe21)]. By using different classification systems it is possible to use the same underlying data sets but significantly alter the proportion of cases in a specific category, such as ‘unexplained’. However, the clinical and epidemiological usefulness of each system largely depend on the purpose for which the data will be used. It is therefore important to be aware of which system was used so that erroneous conclusions are not drawn when comparing data classified in different ways.

Consent process

Consent requirements for autopsy

Following issues associated with organ retention in UK, the Human Tissue Act came into force in 2006, overseen by the Human Tissue Authority (HTA), which has published Codes of Practice and licensed organisations. 22 The Human Tissue Act covers almost any activity related to human tissue, including autopsy examination, and requires consent for the removal, storage and use of human tissue. The Human Tissue Act applies to all SBs, neonatal, infant and child deaths, but it is recommended that consent be obtained for the examination of all fetal tissue and for its use for scheduled purposes, regardless of gestational age.

Giving consent must be a positive action following the provision of appropriate information and with adequate understanding. The absence of refusal is not adequate consent. Consent must be given voluntarily by an appropriately informed person who has the capacity to agree, which in the setting of perinatal autopsy examinations will usually be one or both parents. For SBs and neonatal deaths, it is recommended that, if possible, consent is obtained from the mother and, when appropriate, both parents are involved. Under the Human Tissue Act, consent from one parent is sufficient. However, if there is disagreement between the parents, it is recommended that this be sensitively discussed before proceeding. It is important that the consent process is not viewed as the single act of signing the consent form, but rather as a process in which parents can discuss the issues, ask questions and make an informed decision.

The responsibility to seek consent is usually the treating clinician’s who should be sufficiently senior and well informed, with an adequate knowledge of the post-mortem procedure. It is recommended that consenting clinicians are trained in the management of bereavement and should have witnessed an autopsy examination. As valid consent can only be given if appropriate communication has taken place, information leaflets and consent forms should be available in the main local community languages for patients whose first language is not English, and interpreters should be used.

Practical aspects of the standard autopsy examination

Limitations of autopsy examination

Despite the potential benefits, as outlined above, it is important that parents do not have unrealistic expectations. The examination may not answer their questions and in a significant number of cases may not establish a cause of death, especially for clinically unexpected third trimester SBs. Conversely, as outlined in Functions of perinatal and paediatric autopsy, although the autopsy examination may find ‘nothing new’, this too may be clinically helpful, providing reassurance to both clinicians and parents that nothing important had been overlooked during life.

The autopsy report

It is recommended that a final autopsy report, incorporating histological findings and results of further investigations, is provided 6 weeks after the autopsy examination. 23 The report should document the major findings and results of ancillary investigations, and also include an appropriate clinicopathological correlation and a summary. Parents are entitled to a copy of the report, but it is recommended that the contents be discussed with them by their clinician prior to receipt, preferably in person, as some parents may find the technical language used in such reports insensitive or distressing.

Prevalence of perinatal autopsy and reasons for consent refusal

Despite the potential benefits noted above, perinatal autopsy rates have decreased over recent years. The perinatal autopsy rate (including late fetal losses, SBs, and neonatal and postneonatal deaths) decreased from 48% of potential cases in 2000, to 39% in 2003, continuing to fall in recent years, with marked regional variation across the UK. Recent data have revealed a plateau for the proportion of neonatal deaths referred for consented autopsy examination, with 22% in 2003 and 21% in 2007. 15

In most cases, autopsy is offered by the clinical team but consent is not given by the parents. Although parents’ opinions regarding autopsy have been influenced by the organ retention issues in some hospitals in the UK,24 the main reason for refusal of standard autopsy is fear that the child will be mutilated or disfigured by the procedure. 25 It is of note that in a study of women’s reactions, 30% who refused autopsy subsequently regretted the decision, possibly as their questions remained unanswered. 26

Cultural and religious considerations also play a role, although most major religions do not explicitly prohibit an autopsy, especially if there is a perceived benefit to public health. The exception to this is in Islam and Judaism, in which cutting and disfigurement of the deceased, and removing internal organs, tissue and fluids, is forbidden. In addition, there is a religious requirement to bury the body as soon as possible. 27–30 Empirical evidence has highlighted uptake rates of fetal or neonatal autopsy among Muslim parents of 23%31 and 42%,32 compared with 77%31 and 75%32 among non-Muslim parents, respectively.

Clinicians find discussing the option of autopsy examination with bereaved parents difficult and distressing. This is further influenced by the complex and often lengthy consent forms now required. In conjunction with reduced exposure to autopsy, younger clinicians regard the autopsy as less useful, compared with senior colleagues. 33 Finally, perceptions may be influenced by the attitudes of pathologists, delays in issuing final autopsy reports, lack of clinicopathological correlation, and paucity of appropriate multidisciplinary team meetings contributing to the notion that the autopsy is of limited value in the immediate and subsequent management of the patient and/or parents.

New developments in autopsy investigation: less invasive autopsy approaches

To the best of our knowledge, traditional standard autopsy, based on prosection and dissection, has remained largely unchanged for hundreds of years. However, increasingly, parents are declining, such that at present the majority of parents do not agree to standard autopsy. In order to address these concerns and to improve uptake rates, the feasibility of less invasive autopsy (LIA) techniques has been developed and evaluated in recent years. 34 One promising approach is the use of cross-sectional imaging techniques, in particular magnetic resonance imaging (MRI). MRI provides excellent soft tissue imaging resolution in this patient population and MRI scanners are readily available in most hospitals.

A large Department of Health and Social Care-funded prospective, blinded trial comparing NIA, based on post-mortem MRI along with other ancillary investigations such as placental examination, genetic and metabolic tests, with standard autopsy, reported that there was around 95% concordance for major diagnoses for fetuses. Accuracy was less for infants and children (85% and 54%, respectively), as imaging alone is unable to detect cases of systemic disease with no anatomical features, such as sepsis. 35

In order to address this shortcoming, the combination of MRI plus targeted laparoscopic examination and biopsy of visceral organs (MIA) has been proposed as a possible alternative, as it combines the advantages of both imaging and tissue sampling; it is estimated that > 90% of significant histology findings from standard perinatal autopsies could be detected using a minimally invasive approach, although further evaluation in large numbers of cases would be required to provide accurate data for all pathologies. 36

This LIA approach was recommended as a realistic alternative to current invasive post-mortems in adults by the Department of Health and Social Care Post Mortem Forensic and Disaster Imaging Group in 2012, with the acknowledgement that there are important religious, cultural and humanitarian benefits offered by non-invasive post-mortems. 37

Less invasive autopsy has the potential to address sensitivities around organ retention, as well as accommodate the need to respect religious and cultural diversity. However, currently, very little is known about the acceptability of these alternative methods. A small study of 70 parents conducted in Belgium found a hypothetical increase in uptake of fetal and neonatal autopsy from 60% to almost 80% if MIA was offered. 32 Moreover, around 60% [95% confidence interval (CI) 35% to 82%] of Muslim participants in that study would hypothetically consent for MIA, although the sample size was small (n = 13). In a questionnaire study conducted in the UK, the authors found that parents preferred a full autopsy over a less invasive procedure if they were told it might provide more information. 38 However, owing to the small size of these studies, it is not possible to make generalisations about the acceptability of LIA and therefore further research is required, particularly to explore whether or not these alternative methods would be acceptable to those religious groups that traditionally decline autopsy. Among health professionals, MIA has been shown to be highly acceptable and its availability is considered beneficial for discussing autopsy with parents. 39

However, several important questions remain to be addressed, including likely acceptability in the wider population, how it might be implemented into existing care pathways, training requirements for staff offering LIA and which patient populations it is most appropriate for.

Importance of research on less invasive autopsy

Developing more acceptable alternatives could allow more parents to benefit from gaining information regarding prenatal and paediatric deaths, inform recurrence risk and have secondary benefits for researchers, policy-planners and society, by providing improved information regarding causes of fetal, infant and child deaths as a result of increased uptake. There are currently very few data regarding the acceptability of MIA and NIA compared with standard autopsy. If LIA represents an acceptable alternative to standard autopsy, this could radically change the future approach to investigating such deaths. Reducing perinatal mortality is a World Health Organization priority, ‘health of children and young people’ is a stated key priority for the European Commission and reducing premature mortality is one of the current NHS priority areas. 40,41 Improving the range of investigative strategies available for post-death investigation also aligns with NHS England’s goal of significantly improving patient choice by 2020. 42

Research aims

This research programme was submitted in response to the National Institute for Health Research commissioned call for its Health Technology Assessment programme.

The aims of this programme of research were as follows:

• Provide empirical data on the acceptability and likely uptake rates for different types of autopsy among various patient groups and other key stakeholders (health professionals, coroners, religious leaders), using a mixed-methods approach. In particular the study set out to address –

  • if MIA and/or NIA were more acceptable to parents than standard autopsy methods

  • how alternate methods of investigating death should fit into existing care pathways

  • which patient populations these methods are most appropriate for

  • how best to offer such a service to groups for whom standard autopsy is never acceptable, including specific ethnic and religious populations.

• Analyse existing autopsy data sources, including national data, retrospective autopsy series and existing LIA and MIA series, to provide estimates of the potential efficacy of MIA (incorporating a range of either no value or as good as standard autopsy) and its projected utility in clinical practice.

Components of the research programme

There were three components of the research programme: (1) a systematic review of factors having an impact on consent for autopsy; (2) empirical research with key stakeholders to determine the acceptability and likely uptake of LIA; and (3) a retrospective analysis of existing autopsy data.

Methods

We followed the method described by the Centre for Reviews and Dissemination45 and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklist46 to conduct this systematic review. We undertook a quality assessment of the studies before conducting a narrative synthesis47 of the results. The initial search was undertaken in December 2015, repeated in August 2016 and again in April 2018.

Study selection

Initial searches identified 1484 potential articles. Independent assessment by two researchers reduced these to 35 papers included for the quality assessment (Figure 1). A further five relevant papers were identified in April 2018 which were then included in the review.

FIGURE 1.

Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2009 flow diagram.

Results

Discussion

Conducting research into parental experience and attitudes towards autopsy examination has been recognised as being particularly challenging due to the sensitivity of the subject matter. 83,84 Despite these difficulties, these data indicate that researchers are attempting to meet these challenges, as a relatively large number of studies exploring the reasons parents accept and decline autopsy were identified. This systematic review highlights that current low autopsy uptake rates are a consequence of numerous factors encompassing not only procedural, psychological and cultural barriers from the parental perspective, but also a number of professional barriers, many of which were common across countries. This is not surprising, as parents are likely to have the same concerns around the invasiveness of the procedure and appearance of the body irrespective of where they live, their culture or their religious beliefs. This review also identified examples of best practice that may provide valuable evidence to health professionals and policy-makers around how these sensitive services should be delivered.

The concept of a less invasive approach to autopsy is relatively new85 and has thus far been considered in only four studies, all of which found it to be more acceptable than traditional autopsy for most participants, particularly those of Muslim faith who have low rates of acceptance of traditional autopsy. 32,39 This supports the view that less invasive methods may be more acceptable to those parents who currently decline autopsy. Professional concerns around the limitations of the technology were raised,6,38 highlighting that both advantages and limitations need to be made clear when discussing the various options for autopsy with bereaved families, which will require evidence-based data related to particular clinical circumstances. Although non-invasive imaging-only approaches may provide useful additional information in some circumstances, such as underlying structural malformations, they are unlikely in isolation to be useful for identification of many pathologies, such as metabolic diseases or infections. 86 However, when augmented with minimally invasive ancillary investigations (needle biopsy, placental examination, etc.), overall accuracy rates similar to that of conventional autopsy (> 90%) can be achieved in many circumstances. 35 Further work with key stakeholders, particularly parents and community leaders from those religious groups who traditionally decline autopsy, would be valuable to determine whether or not less invasive methods as part of routine clinical care would be religiously and culturally permissible and acceptable. In addition, as less invasive methods of autopsy become increasingly available, research to explore whether or not the psychological barriers identified in this review remain prominent factors would be useful. Finally, future studies reporting on autopsy yield should provide data regarding which specific aspects of the procedure contributed to the diagnosis or main findings, in order to allow appropriate counselling for parents considering more limited approaches.

Health professionals’ reluctance to raise the topic of autopsy was identified as a major barrier to uptake. Unless the death is being referred to the coroner, it is recommended that all parents should be offered the opportunity to discuss having an autopsy examination so that they can make an informed choice. 87 For this reason, creating environments that support health professionals to do this is critical. A number of studies identified examples of good practice when consenting parents for autopsy underscoring the importance of national guidelines. In the UK, the HTA have recently addressed this issue with the introduction of codes of practice for autopsy examination. 88 Many of the examples of best practice identified in this review are echoed in this document. The quality and appropriateness of the consent form and the need for national, standard information sheets and consent forms were also raised by health professionals. 6,70 In 2013, Sands launched the Sands perinatal post-mortem consent package, which was developed to provide information and guidance about post-mortems for health professionals seeking consent. 89 Research to determine the impact of this new consent package, as well as the HTA guidance, would be valuable to identify whether or not it has a significant impact on autopsy consent rates.

A notable facilitator of autopsy uptake was parental desire to contribute to research, as this created an opportunity for their child’s life to have meaning and value. Health professionals may be reluctant to ask bereaved parents about research as they are a potentially vulnerable group with high levels of distress, but it is clear that an opportunity for involvement in research is in fact valued by many families. Some research has specifically focused on including bereaved parents in research studies,83,90,91 with one study83 reporting that 73% of parents stated that taking part in research about autopsy decision-making had helped them to feel better about the decision, and they regarded such studies as valuable and important. These findings support the findings from this review and underscore the importance of health professionals explaining the potential for autopsy to contribute to medical knowledge when discussing the value of the procedure with parents.

The strengths of this review are the comprehensive search strategy used to identify papers and the use of a quality appraisal tool. Regarding limitations, older papers are not always well indexed, particularly if they are qualitative studies, and this is likely to be the reason why a significant number were identified through hand-searching of authors and reference lists. The quality appraisal identified variability in the methodological rigour of the included studies, including low response rates, lack of detail regarding data analysis (particularly in qualitative studies), omission of statistical significance testing or analysis of the influence of variables such as age, education, ethnicity, etc. Qualitative studies were predominantly limited to simple thematic analysis with very little interpretive content. Some papers included professional perceptions of parental views; these may not be consistent with the views of parents themselves as indicated by the diversity in patient and professional views, as reported in one study7 describing the impact of the organ retention publicity.

Conclusion

In summary, this review provides an insight into the parental and professional barriers around consent to autopsy that have resulted in suboptimal uptake rates. We have identified a number of important barriers, including system-level barriers within the NHS, as well as practical, psychological and religious parental barriers that impact consent uptake. We have also identified a number of facilitators that highlight the need for better health professional education and the fact that some of these concerns may be mitigated if less invasive methods of autopsy were routinely available to bereaved parents. Furthermore, new consent packages and HTA guidance may have a positive impact on the perception of examination after death in the UK. The landscape is changing; further research is necessary to assess its impact on autopsy uptake rates.

The research team and advisory group

The multidisciplinary research team included academics from the following disciplinary backgrounds: paediatric pathology (NS and CH), clinical radiology (OA), genetics and fetal medicine (LC), clinical psychology (MR), genetic counsellor (MH) and health psychology (CL). An advisory group was established at the start of the study in order to monitor the progress of the research and to include the perspectives of those with a stake in the research. The group included an academic clinical psychologist with expertise in bereavement research; a senior clinical lecturer in obstetrics with a special interest in SB research; and representatives from the patient groups Sands, The Lullaby Trust and Genetic Alliance UK. The advisory group had a bi-annual conference call and a written update was circulated every 3 months.

Ethics approval

NHS Research Ethics Committee approval for this empirical research was obtained in April 2016 (16/LO/0248 from London – Bloomsbury Ethics Committee).

Overall study design

This was a mixed-methods study comprising standard data collection approaches, namely a cross-sectional survey, semistructured interviews and focus groups. Mixed-methods research has been defined as ‘research in which the investigator collects and analyses data, integrates the findings, and draws inferences using both qualitative and quantitative approaches or methods in a single study or program of inquiry’. 92 By using a mixed-methods approach we aimed to ensure the validity of the research by cross-checking the findings from the different methods of data collection, increase the level of knowledge and capture different dimensions of the same phenomenon. 93 Data were collected by one of two researchers: a social scientist with experience in conducting qualitative studies on sensitive topics (CL) and a postdoctoral clinical psychologist experienced in conducting qualitative interviews (MR).

Substudy 1: mixed-methods study with bereaved parents to determine acceptability and likely uptake of less invasive autopsy

Substudy 2: interview study with health professionals and HM Coroners

Substudy 3: interview and focus group study with religious leaders and community members

Patient and public involvement

Patients were involved in the design of the Health Technology Assessment application. Two patient and public involvement meetings were held prior to writing the grant application – one involving the parent support groups (Sands, ARC and The Lullaby Trust) and one involving representatives from specific religious groups (Muslim, Jewish, Christian and Hindu). One of the patient representatives (Jane Fisher from ARC) is a co-applicant on this grant and was involved in writing the grant and the design of the study.

Patient representatives were involved in the management of the research on both the Study Steering Committee, in which we had representatives from The Lullaby Trust, Sands and Genetic Alliance UK, and on the advisory team (Jane Fisher from ARC). There were bi-annual meetings (as well as written updates every 3 months) with the Study Steering Committee, who oversaw the research.

Patients and patient representatives were involved in developing participant information resources, the interview topic guide and the bereaved parent survey. The survey was initially designed by Celine Lewis and revised by members of the advisory team. It was then circulated to three bereaved parents (from The Lullaby Trust and Sands) who reviewed the survey, including the descriptions of standard autopsy, NIA and MIA. The three bereaved parents also checked the survey for comprehension and their comments were incorporated into the final version. Participants for the survey and for the parent interviews were recruited by the patient groups Sands, The Lullaby Trust, ARC and Child Bereavement UK.

Key findings from the research were presented to the advisory team and the Study Steering Committee, who helped to translate the findings into implications for practice. The findings from this body of research were also presented to the advisory team, Study Steering Committee, as well as health professionals and religious/community leaders, at a dissemination meeting which took place at the University College London Great Osmond Street Hospital Institute of Child Health on 3 May 2018. Key findings from the empirical research will be summarised and disseminated through the support groups who were recruited into the study, following publication in academic journals (either through a newsletter or on the website), to ensure that those parents who took part have access to the results.

Survey participants

In total, 938 questionnaires were returned: 870 through retrospective recruitment (655 via Sands, 108 via ARC, 81 via The Lullaby Trust and 26 via Child Bereavement UK) and 68 through prospective recruitment. Seventy-nine questionnaires were excluded because of missing data [whereby questions on likely uptake, acceptability and preferences for the different methods were omitted (questions 1–7)], leaving 859 questionnaires for inclusion in the analysis (79 from the retrospective recruitment and 0 from the prospective recruitment). Owing to the nature of the retrospective recruitment, it is not possible to calculate a recruitment rate. For the prospectively recruited participants, 23 participants actively declined to take part at the time of discussing the study and 160 participants passively declined (study packs were taken home and not returned) (30% response rate). Over 99% of respondents had experienced a perinatal or child loss (0.7% did not answer this question) and 48.6% of parents consented to a standard autopsy. Survey responder characteristics are presented in Table 1.

Quantitative results

Interview participants

For the interviews, 36 participants were contacted following their initial voluntary expression of interest and 20 consented and took part (56% response rate). Interview participant characteristics are presented in Table 5.

Qualitative findings

The findings below summarise the overall themes that emerged from the qualitative analysis around acceptability and likely uptake of standard autopsy, MIA and NIA.

Discussion

This is the largest UK study examining the views and preferences of bereaved parents towards standard and less invasive forms of autopsy. The findings suggest a likely large overall increase in uptake, which could be as high as 90% if less invasive methods were available, although in practice this will depend on availability and effectiveness of different methods of autopsy in specific circumstances. Both MIA and NIA were clearly the more acceptable approaches (86% and 87%, respectively, compared with 77% for standard autopsy), a finding supported by the qualitative work, in which less invasive approaches were preferable, as they were considered ‘kinder’ to both parent and child. If asked to choose between the different approaches, nearly half of participants (46%) would choose MIA, as it enables tissue sampling without requiring large incisions to the body and is therefore perceived as a ‘good compromise’.

Although 74–77% of participants reported that they would consent to MIA or NIA, around 20% of responders remained undecided (‘not sure’). This uncertainty might indicate that further information about the procedure is required, including details of under which circumstances it is likely to result in a diagnosis. If the full range of approaches for examination after death were available, around 15% of responders indicated that they would still chose standard autopsy, highlighting the need for provision of choice. This finding suggests that some parents will continue to opt for the method which can theoretically provide the most information.

Notably, there was a marked difference (24%) in the proportion of responders in our study who thought that standard autopsy was acceptable (77%), compared with the number who would consent to it themselves (54%). This indicates that there is a difference between a procedure that is considered theoretically acceptable for ‘others to have’ and what parents would choose for themselves in practice.

Small, early studies with bereaved parents suggest that less invasive approaches are preferable to standard autopsy for some parents;31 however, other studies have found that the method likely to provide the most information is preferred. 38 Consistent with our study, Henderson et al. 79 found that fewer well-educated women were significantly less likely to consent to a standard autopsy. Our finding that survey responders who were less well-educated and would decline standard autopsy were likely to consent to LIA highlights that this would open up opportunities for this group that would otherwise be unavailable.

Although our findings support those of Cannie et al. 31 in that, overall, there was a preference for less invasive methods, 15% of participants still indicated a preference for standard autopsy as it was the most complete test available. Around half of parents (54%) stated that they would consent to a standard autopsy, findings which align with recent UK national data9 and an international survey of parents following SB,105 although a recent study in the UK found that around 80% of parents who experienced SB consented to a standard autopsy. 79 The primary reasons why bereaved parents consented to autopsy in this study, namely to understand why it happened and if it would happen again, have also been identified as the main motivators in other studies,7,38,54,68 supporting the importance of information to inform recurrence risk for many parents.

In the current survey, religion was not found to be an important factor in decision-making around autopsy. However, this is likely to be because of the small number of Muslim and Jewish responders (12 in total) and therefore should not be considered generalisable. Our qualitative research with parents and religious leaders from these two religious groups has shown that the invasiveness of standard autopsy is a key reason why Muslim and Jewish families decline, along with the requirement to bury the body as soon as possible, and that less invasive methods, in particular non-invasive approaches, would be considered religiously permissible and much preferred.

A major strength of the present study is the large number of participants and the use of a mixed-methods approach, which enabled corroboration of findings and a deeper understanding of the topic. Moreover, responses from prospectively recruited parents suggest that the majority of survey responders did not experience any adverse effect from taking part in the study. Although there was a significantly larger number of participants recruited retrospectively than prospectively, there was no overall difference in preferences for autopsy between the two cohorts. Moreover, we did not identify any differences in autopsy preferences across participant characteristics or loss experience (late miscarriage, SB, etc.), despite the majority of survey responders self-selecting their ethnicity as white and their education level as well educated. There was a greater proportion of responders who had experienced SB than other forms of perinatal loss, but this mirrors the proportions from recent national data on childhood, infant and perinatal mortality106 and, similarly, 49% of parents in our sample had previously consented to a standard autopsy, which is comparable to the national average of around 45%. 9 Not unexpectedly given the sensitive subject matter, there was a low response rate from parents recruited prospectively, which may have biased findings (30%). Women with poorer emotional health may have been less likely to respond. However, this figure is comparable with a recent survey of parental experience of autopsy following SB, in which the response rate was also 30%. 79 In addition, there was a low response rate for the interview study (56%). Thus, the qualitative findings may not be representative of the broader sample. Online recruitment enabled us to recruit a much larger sample than was likely to have been possible prospectively, yet it is subject to non-response bias, as participants may have been more interested in or enthusiastic about the topic. Finally, the hypothetically high uptake of MIA and NIA seen in our findings may not concur with actual uptake in a real-life scenario.

Conclusion

Further clinical work is required to assess concordance between MIA, NIA and standard autopsy so that appropriate counselling can be provided across a range of specific clinical scenarios (e.g. SB vs. infant death). Furthermore, economic and implementation evaluations are required to assess cost implications, as well as how LIA could be offered routinely as a clinical service within the NHS.

Sample characteristics

Forty health professionals were contacted and 25 from 11 hospitals consented and took part in the study (63% recruitment rate). Ten coroners were approached: one actively declined, five did not respond and four participated (40% recruitment rate). At the time of the interview, nine health professionals were already offering NIA to parents as a clinical service and all four coroners explained that it was available in their jurisdictions at a cost, although none recalled using it for infant or childhood deaths. Six health professionals were offering MIA either as part of a study or as a clinical service. None of the coroners were offering MIA (Table 6).

Four overarching themes were established during the coding process: (1) benefits, (2) concerns, (3) implementation issues and (4) predicted uptake.

Benefits of minimally invasive autopsy and non-invasive autopsy

Health professionals and coroners acknowledged that availability of less invasive methods of autopsy was a positive development which would be of benefit to bereaved parents and potentially increase uptake:

Who in their right mind wants interference with a body of a loved one, if you didn’t have to? No one would. So I think I would encourage research into ways of replacing, augmenting and providing an alternative to traditional autopsy.

Coroner 1

They both sound really good, and I think you’d get a lot more uptake of post-mortem by having them available.

Obstetrician 1

Concerns

Implementation

Predicted uptake of minimally invasive autopsy and non-invasive autopsy

All participants felt that the availability of NIA and MIA would increase uptake, although this varied from a ‘much, much bigger uptake’ (bereavement midwife) to ‘you might adjust the margins somewhat’ (consultant ICU). For health professionals who were already offering NIA or MIA, there had already been an increase in uptake:

I can certainly think of quite a number of people that it was like ‘no, no, no’ to the full post-mortem and then you would say ‘well, actually there is this that we can offer’ and that they would agree to that.

Bereavement midwife 1

It was acknowledged that some parents may prefer a traditional autopsy if it was likely to yield the most information.

Regarding how NIA and MIA should be offered, the vast majority advocated an approach whereby parents’ consent to the most invasive option they would consider acceptable, but if a diagnosis could be made less invasively, nothing further would be done.

Discussion

This study provides a unique insight into the views of health professionals and coroners working in perinatal and paediatric settings towards LIA. Examining the views of health professionals who discuss and consent bereaved parents is important, as paediatric and perinatal autopsy examination represent the largest group of consented autopsies. 108 Overall, participants viewed NIA and MIA as positive developments, likely to increase uptake of post-mortem, as they anticipated that parents would find it more acceptable, particularly those for whom current options are morally or religiously objectionable. These findings are consistent with a questionnaire study,39 in which LIA was found to be an acceptable alternative to traditional autopsy for most health professionals.

Participants in this study acknowledged a number of practical challenges that require attention before widespread implementation of less invasive perinatal and paediatric autopsy into clinical practice, many of which were identified when considering the use of autopsy imaging in adults. 109,110 Recommendations from that work included standards of practice and training programmes for pathologists, radiologists and APTs to be developed to deliver a trained workforce (e.g. APTs trained to operate scanners and undertake some minimally invasive procedures); imaging to be performed in any hospital equipped with scanning equipment, which could be used out of hours, with images then sent to a centre of expertise for reporting; and LIA to be conducted within already established centres of pathology based regionally across England, addressing concerns around equity of access. 109 Such strategies require investment and support from clinicians, as well as hospital decision-makers, royal colleges and local authorities. A detailed cost–benefit economic analysis to determine the true cost of implementing the service is therefore required.

Health professionals identified numerous situations in which imaging could be as reliable as, or even superior to, a full autopsy; comments that are supported by the current evidence. 111–113 Nevertheless, there were some circumstances in which it was unclear if imaging would be useful; for example, imaging of a child who had already had scans prior to death. This highlights the importance of ongoing research to understand which circumstances are most suitable for which method of autopsy, and development of formal guidance is required, both to ensure application in appropriate settings and also to inform clinicians’ consultations with bereaved parents regarding the likely yield of imaging or other investigations.

Further work is also required to determine the effects on reporting times for NIA and MIA. Despite NHS England recommending that 60% of autopsy reports should be issued within 42 days of examination,114 lengthy reporting times with current autopsy practice have been identified as a critical issue for parents, many of whom feel that they cannot move on until they receive a result. 79 The need for approaches that reduce the time required for laboratory processing and sample analysis may help to create a viable clinical service. 115 To the best of our knowledge, there is no published evidence to date demonstrating that MIA reduces the time from examination to final autopsy report, although NIA should result in much shorter turnaround times as there is no requirement for histological examination of internal organs.

The main strengths are that a range of health professionals were included with diverse views and experiences, and recruitment continued until saturation was reached. Participants were self-selecting, hence there may be responder bias towards people who have strong views. We tried to minimise this by purposely inviting people with varied views towards the technology. Snowball sampling is associated with problems of representativeness. There was a low response rate from coroners (40%). Finally, this research was conducted with health professionals and coroners in the UK only; opinions may differ outside the UK.

Conclusion

This work provides useful insights into the acceptability and the potential challenges of implementing LIA in clinical practice. Further research with bereaved parents, as well as research with religious groups, is required and forms aspects of the larger study of which this is part.

Study participants

Sixteen of 19 (84%) religious or community leaders participated (Table 7). Eight focus groups comprising 60 Muslim participants and two with 16 Jewish participants were conducted (Table 8).

Three Jewish questionnaire responders were contacted and agreed to take part in an interview (age range 50–64 years; all educated to A Level or above; two were ‘quite’ religious, one was ‘not very’ religious). One responder had experienced a neonatal/infant death and had a coronial autopsy, one responder had a SB and one responder a ToP; the latter two responders had consented to a full autopsy. None of the questionnaire responders from the Muslim community (n = 6) agreed to be contacted further.

Key themes

Participants acknowledged that religious devotion involved a negotiation between religious observance and the complexities of daily life. For some participants, it was not always possible to ‘live our lives as strictly as we would like’ (FG9, Pakistani Muslim, women), or that sometimes, for example when a tragedy occurs, one’s emotions over-rode one’s observance of religious beliefs and practices:

Everyone’s faith goes up and down, so when the tragedy does happen to you, it depends what emotional state you’re at, a lot of things comes into it, sometimes instead of religion coming into it, it’s more your personal feelings.

FG7, mixed ethnicity, Muslim, women

Others discussed how their religion gave them comfort in times of tragedy and that the rituals around death ‘were just incredibly supportive’ (FG8 Jewish Orthodox, mixed gender). Moreover, there was an understanding that everyone interprets and practices their religion differently and that religion is not always ‘black and white, cast in stone’ (FG5 Bangladeshi Muslim, women). These factors influenced discussions around the different types of autopsy and accounted for the variation in viewpoints that were observed.

Discussion

There are 2.7 million Muslims and 263,000 Jewish people currently living in England and Wales, and both populations are growing. 100 Moreover, in the UK, two-thirds of Muslims are of Asian origin (primarily Pakistan and Bangladesh) and they have the second highest risk for SB (6.3/1000 births vs. 4.16/1000 births for UK population), neonatal death (2.3,1000 births vs. 1.77/1000 for UK population) and extended perinatal death (8.7/1000 births vs. 5.92/1000 for UK population). 9 To our knowledge, this is the first study to explore in depth the acceptability of LIA among members of the Muslim and Jewish communities. Even though the study was conducted in the UK, the findings may be of relevance to other countries with significant Muslim and/or Jewish populations.

In this unique exploration of the acceptability of LIA, we have shown that there are no religious objections to NIA and that it is far more acceptable to Muslim and Jewish parents than a traditional autopsy, which, as other studies have also demonstrated,31,32,39,117,118 most would decline. Although MIA is still viewed as invasive and is therefore not strictly religiously permissible, it is less objectionable than a traditional autopsy, particularly in circumstances when investigation is required (such as a coronial procedure), and it would be acceptable to a significant number of Muslim and Jewish parents if the information could help prevent losses in future pregnancies.

In most focus groups and parent interviews, a diverse range of views and attitudes were expressed relating to the acceptability of the three methods of autopsy, irrespective of ethnic group (for the Muslim participants), gender or previous experience of loss. The one group in which this differed was the Jewish Haredi focus group, in which all the participants identified themselves as being ‘very religious’ and attitudes towards the different types of autopsy were homogeneous. Only NIA was seen as acceptable in specific circumstances and with consent from a rabbi, with all other types of autopsy considered unacceptable when not required by HM Coroner. This is likely to reflect the strict observance to Jewish law within this community and the common viewpoint that the loss of a baby or a child is the will of God and, as such, does not require further explanation (a view also expressed by a number of Muslim participants). Seeking advice from a rabbi when faced with important life decisions is common among this religious group and it is rooted in the principle of da’as Torah (Torah knowledge), which mandates that one must seek Torah-based guidance from a rabbinic authority on all matters of life. 119

Most religious leaders and community leaders who took part in this study were aware of the use of NIA for adult coronial cases and valued this option in those circumstances in which an autopsy was legally required. The use of NIA for adult coronial cases was established at the request of the Jewish community in Manchester in 1997120 and has gained increasing awareness among the Muslim and Jewish communities in recent years. 121–123 In most of the adult cases, however, the cost of the service is covered by the community themselves124 at a cost of around £500–900. 125 The new Coroners and Justice Act 2009 recognised the importance of religious requirements relating to autopsy examination and permitted expeditions release of bodies where appropriate, as well as permitting less invasive post-mortem examinations. 126 Moreover, in 2015 a high court judge backed the religious right of Muslim and Jewish people to ask for NIA and that it must be considered by the coroner if there was a ‘reasonable possibility’ that it could establish the cause of death. 123 Given that NIA has been allowed within both communities for coronial cases, it is unsurprising that extending its use for non-coronial cases in childhood was perceived to be acceptable.

The findings from this study suggest that the availability of LIA may result in an increase in uptake from members of the Muslim and Jewish communities, although this is likely to require a quick turnaround to enable burial, preferably within 24 hours. Implementing such a service raises a number of practical and ethical challenges. Providing a 24 hours a day, 7 days a week service would require capital investment, training and support of health-care management to provide dedicated autopsy imaging facilities and personnel, including pathologists, radiographers and APTs. These implementation challenges have been identified in previous research with health professionals and HM Coroner. 107 Moreover, important questions around prioritising access for Muslim and Jewish parents at the expense of other parents having to wait longer, who may also prefer a less invasive approach, need addressing. Further thought will need to be given as to how LIA might be implemented into clinical practice in a way that is sensitive to the preferences and needs of all parents. In addition, as part of any possible future implementation study, the impact of timing of release and burial in relation to examination type should be assessed.

Participants identified a number of practical recommendations for the successful implementation of LIA among the Muslim and Jewish communities, largely related to raising awareness among the community as well as religious leaders and health professionals. Pathologists, radiologists and other professionals with experience of LIA will need to play a key role in terms of training and community outreach. One recommendation related to the importance of training health professionals about Muslim and Jewish laws and customs concerning autopsy, particularly as many health professionals lack confidence in communicating across cultural groups different from their own. 127 However, as participants had varying attitudes towards the acceptability of different types of autopsy, health professionals must avoid cultural assumptions or stereotyping and focus on facilitating individualised care.

Less invasive autopsy was preferable to full autopsy, not just for religious reasons but also because it was perceived to be ‘kinder’ and less traumatic to both parent and child. These viewpoints align with those that have been observed among the general parent population, for whom dislike of the invasiveness of the procedure and the desire to protect one’s baby or child from harm have been identified as key barriers to autopsy. 44 As such, it is highly likely that many non-Muslim and non-Jewish parents will find LIA preferable to a traditional autopsy. Further research with bereaved parents across a wider range of cultural, ethnic and religious groups is therefore required.

Key strengths of the study lie in the high response rate from the religious leaders and community leaders (thus reducing the risk of non-response bias), the range of religious leaders and community leaders that were included, the inclusion of community members from different religious and ethnic backgrounds and locations, and the range of experiences of loss, thus ensuring a wide range of viewpoints. Furthermore, we validated the findings through member checking with key informants to increase rigour. Although focus groups offer advantages associated with insights into attitudes and opinions that group interactions can enable, a challenge is the possibility of group dynamics promoting uniformity of views. We tried to address this by emphasising our interest in different perspectives at the start of each focus group. A further limitation is that in some cases community facilitators and key informants supporting recruitment into focus groups approached community members drawn from an established group of contacts, and in some cases focus group participants knew each other. This may have had an impact on the diversity of opinions expressed in the group. However, given the range of opinions expressed in most group discussions, this is unlikely to have influenced our conclusions. We provided a generalised description of MIA to study participants, involving a small incision to the stomach. It is possible that additional or alternative approaches may be developed that involve more puncture sites (e.g. up the nasal cavity or at the base of the brain), which will require comparative evaluation with parents to assess acceptability. Finally, the discussion around acceptability and likely uptake of NIA and MIA with the parents was hypothetical and may not reflect decisions people make in real life. Further research to assess actual uptake is required if LIA becomes routinely offered in clinical practice.

Conclusion

Less invasive autopsy is more acceptable to the Muslim and Jewish communities in the UK and has the potential to increase uptake in these religious groups, particularly if turnaround times can be minimised and awareness raised among community members. Our findings are likely to be useful for health professionals and decision-makers who direct future clinical practice in this area and may be of relevance to other countries with significant Muslim and/or Jewish populations. Further work with bereaved parents from other cultural, religious and ethnic groups, as well as quantitative data to provide more accurate estimates of potential uptake, is required.

Methods

Great Ormond Street Hospital is a tertiary referral centre for perinatal and paediatric autopsies, with several hundred cases performed per annum by specialist pathologists. Detailed data from all such autopsies since 1995 are recorded in a dedicated autopsy database, including > 400 variables/fields, with associated objective criteria described in a database handbook. The database allows identification of cases for research use, has Research Ethics Committee approval and has been used for several previous research studies. 128,129

Specifically, for the purposes of this study, each case was classified as SB/intrauterine fetal death (IUFD), ToP, SUDI or SUDC, according to clinical presentation. For each case, all organs examined were recorded as being normal, abnormal but not relevant to cause of death or main diagnosis, abnormal and possibly relevant to cause of death or main diagnosis, or abnormal and definitely relevant to cause of death or main diagnosis, based on both macroscopic appearance and histological (microscopic) examination.

As part of data entry, judgements made by the reporting pathologist regarding the abnormalities present at internal examination and on histological examination were recorded according to predefined categories (Figure 2). These categories were independent of each other, so they could be used in any combination for macroscopic and microscopic examination, and could be explored further within free-text boxes within the database. These categories were also applied to placental examination and placental histology, when appropriate.

FIGURE 2.

Categories of abnormalities.

Following completion of data entry, data were extracted for all completed cases between 2005 and 2016, according to referral category (SUDI, SUDC, IUFD or ToP) and were analysed using Microsoft Excel^® (Microsoft Corporation, Redmond, WA, USA) and StatsDirect (StatsDirect, Altrincham, UK), using descriptive statistics, chi-squared and comparison of proportions tests when appropriate.

Results

The relative yield (displayed as a percentage of the total number of cases in the relevant referral category in which data on that specific organ were available) of macroscopic examination and histological examination of organs by referral category are presented in Table 9 and Figure 3.

FIGURE 3.

Percentage of cases of SUDC for which cause of death was determined by macroscopic examination of organs.

Sudden unexpected death in childhood

There were 824 cases of SUDC recorded in the database. Analysis of the cases revealed enough information to code the organs appropriately for both macroscopic and histological examination in most cases (range 65–95%); reasons that data for an organ may have been unavailable for analysis include lack of sampling, autolysis, or failure of the target tissue to survive histological processing.

Figure 4 demonstrates the percentage of SUDC cases in which a microscopic abnormality was discovered on histological investigation of an organ when the macroscopic evaluation of that organ was described as normal.

FIGURE 4.

Yield of histological examination in cases of SUDC (% of cases with a microscopic abnormality representing the cause of death when macroscopy was normal).

Sudden unexpected death in infancy

There were 1739 SUDI cases within the database. Analysis of the cases revealed enough information to code both the macroscopic and microscopic (Table 10 and Figures 5–10).

TABLE 10 - Sudden unexpected death in infancy cases with adequate macroscopic and histological comment by organ

Organ	Number of SUDI cases with adequate macroscopic and histological comment	Percentage of SUDI cases with adequate macroscopic and histological comment
Heart	1638	94
Lungs	1680	97
Brain	1503	86
Adrenals	1586	91
Thyroid	1210	70
Liver	1647	95
Kidney	1631	94
Pancreas	1447	83

FIGURE 5.

Percentage of cases in SUDI for which cause of death was determined by macroscopic examination of organs.

FIGURE 6.

Yield (% of cases) of histological examination in SUDI (microscopic abnormality representing cause of death when macroscopy is normal).

FIGURE 7.

Percentage of cases in SB/IUFD for which cause of death was determined by macroscopic examination of organs.

FIGURE 8.

Yield (% of cases) of histological examination in SB/IUFD (microscopic abnormality representing cause of death when macroscopy is normal).

FIGURE 9.

Percentage of cases in ToP for which cause of death was determined by macroscopic examination of organs.

FIGURE 10.

Yield (% of cases) of histological examination in ToP (microscopic abnormality representing cause of death when macroscopy is normal).

Discussion

These data demonstrate that in 5–10% of SUDC and SUDI cases (see Figure 3) the final cause of death is determined by routine histological sampling of macroscopically normal organs, predominantly the heart and lungs, with a few cases contributed by brain, liver and kidney examination. Routine histological sampling therefore remains an important aspect of investigation even if post-mortem imaging appears normal.

In contrast, routine sampling of any macroscopically normal organs only very rarely [< 0.2% (4/2749), 95% CI 0.06% to 0.4%] provides the cause of death in fetal cases (including SB, IUFD and ToP; the approximately 1% of cases in the SB and IUFD group represented detection of ascending infection on lung sampling which would have been detected on placental examination and sampling had this been available). Therefore, in fetal cases in which macroscopic examination for structural abnormalities and post-mortem imaging is normal, there is little indication or yield for invasive organ sampling and histological examination to determine the cause of death. Targeted sampling of abnormal organs and lung may be sufficient to identify abnormal and contributory cases.

Routine histological sampling of macroscopically normal organs provides significant contribution to the cause of death in around 5–10% of SUDI and SUDC cases and is therefore recommended. Histological sampling of macroscopically normal organs in fetal cases, including ToP, SB and IUFD, only very rarely provides useful information for determining the cause of death or the main diagnosis and for many organs the yield is close to zero. Therefore, routine sampling in such cases is of limited value. In cases with contributory histological findings, almost all cases relate to heart, lungs, liver and kidney.

Discussion

To the best of our knowledge, this is the largest study conducted in the UK examining the views and preferences of bereaved parents towards standard and less invasive forms of autopsy. The major finding from this research is that there is likely to be a significant increase in uptake of post-death investigation if personalised and more acceptable approaches were routinely available, with parental preference for less invasive procedures. Our study suggests uptake could be as high as 90%. Dislike of the invasiveness of the standard procedure was found to be one of the main reasons that parents declined an autopsy, as has been highlighted through both our qualitative and quantitative research with parents as well as the systematic review of the literature. The availability of less invasive options would therefore make autopsy available to parents who currently decline, including those from the Muslim and Jewish communities for whom autopsy is unacceptable in the majority of cases for religious reasons.

Strengths and limitations

The empirical research conducted had a number of notable strengths. First, we used a mixed-methods approach employing both qualitative and quantitative research methods. Mixed-methods approaches are likely to yield the most comprehensive picture, as their integration facilitates a deeper and broader understanding of the phenomenon. Another benefit is that each method can help to validate the other (triangulation). Second, we had a large sample size in our survey which yielded statistically significant results. Third, the free-text survey comments along with qualitative interviews provided a rich data source from which to conduct qualitative analysis. Finally, a further strength was that patient and public involvement was embedded from the outset; patient and public input into the survey design was particularly important in ensuring that the questions were clearly phrased and included the benefits and limitations of the different approaches.

A limitation of the survey was that there was a significantly larger number of participants recruited retrospectively rather than prospectively, although there was no overall difference in preferences for autopsy between the two cohorts. In addition, there was a low response rate from parents recruited prospectively, which may lead to biased findings. Another potential limitation is that we did not collect data on how long ago the loss occurred; time since loss may have influenced their views and preferences. Although the qualitative data have given us a rich in-depth insight into the views of key stakeholders towards less invasive forms of autopsy, a potential limitation frequently levied at this form of research concerns the validity and generalisability of the work. However, many of the findings from this research have been identified in similar studies and key findings, and recommendations were supported and validated through participant feedback at a dissemination meeting attended by key stakeholders (parent advocates, health professionals and religious leaders). Finally, although we have developed recommendations in respect of how LIA might potentially be implemented into clinical practice, there is currently a lack of health economic data to understand how affordable such proposals are.

The data presented reflect the relative proportions of fetal and paediatric autopsies performed, specifically that the majority represent intrauterine deaths, terminations of pregnancies or SBs, with a significant proportion of coronial infant deaths but very few paediatric deaths. Therefore, caution should be exercised in extrapolating results to this latter group, as limited data are available.

Implications for practice

The findings from this study have significant implications for future clinical practice. Given the current low uptake rates for standard autopsy, an alternative is required, both to provide best care for parents and families and for public health reasons, including evaluating the quality of obstetric and neonatal care. 115 The broadening range of investigative strategies now available aligns with NHS England’s goal of significantly improving patient choice by 2020. 42 Several requirements must be put in place to make LIA a viable alternative for parents. The findings from our interviews with health professionals and coroners have highlighted the importance of training radiologists to interpret imaging results and pathologists to conduct image-guided biopsies, the availability of scanning equipment, training for health professionals to offer LIA appropriately, and adapted consent procedures and consent forms. 107 This would require significant investment from hospital managers and thus a detailed cost–benefit analysis would be required to determine the cost of implementing such a service. Furthermore, although all post-mortem MRI can be performed in standard NHS scanners (as was the case in the current study), such that additional dedicated MRI machines are not required, there are logistic issues that require attention in order to perform scans out of hours. These particularly relate to staffing by radiographers who are both prepared to work out of normal hours and with deceased patients, and ensuring that appropriate cleaning facilities are available for routine scanner use the following day.

Maintaining a sufficient level of health professional education and awareness with regards to autopsy options could prove to be a major challenge, as numerous barriers to an effective autopsy consent process exist, even prior to the addition of less invasive options such as MIA and NIA. 7,132 These barriers could be compounded by the increasing complexity of the available autopsy options, patchy availability of MIA and NIA, and the relatively high turnover environment of clinical staff. Consideration will need to be given as to how autopsy options might be offered. For example, parents supported a stepwise approach to consenting to autopsy, beginning with the least invasive option, but moving to more invasive options if a diagnosis was not found. A more personalised form of counselling around autopsy whereby type of death, antenatal findings, clinical history, etc., are taken into account when presenting the various options to investigate cause of death is also important. Further clinical work is also required to assess concordance between MIA, NIA and standard autopsy in order that appropriate counselling can be provided across a range of specific clinical scenarios.

Further research

Although this study has demonstrated the likely acceptability of LIA by bereaved parents, further research is now required to evaluate the implications of delivering this approach as a routine service. For example, at present, as expertise is located in only one or two centres, there will need to be both training and education of obstetric and perinatal pathology staff. In addition, resource implications, both for radiology and for pathology departments, must be considered.

Conclusion

The findings of this study have determined that LIA approaches are highly acceptable to the majority of bereaved parents and that the widespread implementation of such LIA options would likely result in a significant increase (more than doubling) of the number of perinatal deaths undergoing formal investigations, compared with current standard practice.

Many parents, including those from religious groups for whom invasive post-mortem examination is unacceptable, would opt for non-invasive investigations, which would require a large increase in the availability and expertise of post-mortem imaging investigations.

As concordance with standard autopsy is already very high, with a limited role of tissue sampling, especially for fetal cases, further clinical trials to determine the additional impact of less invasive tissue sampling protocols would require prohibitively large numbers of cases across many centres and are likely to be infeasible. Further studies should focus on evaluation of logistics and cost-effectiveness of widespread implementation of the LIA approach and are likely to require the involvement of policy-makers, commissioners, providers, health professionals, health economists, patient group representatives and, most importantly, parents and families.

Contributions of authors

Celine Lewis (Senior Social Scientist) designed the study, conducted qualitative and quantitative data collection and analysis, drafted and approved the final report.

John C Hutchinson (Specialty Trainee in Paediatric and Perinatal Pathology) performed MIAs, performed data entry and analysis for the Great Ormond Street Hospital autopsy database and co-authored Chapter 7 of this report. He has approved the final report.

Megan Riddington (Clinical Psychologist) conducted data collection and analysis of the qualitative empirical data, drafted and approved the final report.

Melissa Hill (Senior Social Scientist) conducted data collection and analysis of the qualitative empirical data with key stakeholders, critically revised and approved the final report.

Owen J Arthurs (Consultant Radiologist) designed the study, critically revised and approved the final report.

Jane Fisher (Director, ARC) designed the study, contributed to data collection, critically revised and approved the final report.

Angie Wade (Professor of Medical Statistics) designed the study and statistical analysis, critically revised and approved the final report.

Caroline J Doré (Professor of Clinical Trials and Statistics) provided input related to study design, critically revised and approved the final report.

Lyn S Chitty (Professor Genetics and Fetal Medicine) designed the study, critically revised and approved the final report.

Neil J Sebire (Professor of Paediatric Pathology) conceived the study, interpreted the data, drafted and approved the final report.

Publications

Lewis C, Hill M, Arthurs OJ, Hutchinson C, Chitty LS, Sebire NJ. Factors affecting uptake of postmortem examination in the prenatal, perinatal and paediatric setting. BJOG 2018;125:172–81.

Lewis C, Hill M, Arthurs OJ, Hutchinson JC, Chitty LS, Sebire N. Health professionals’ and coroners’ views on less invasive perinatal and paediatric autopsy: a qualitative study. Arch Dis Child 2018;103:572–8.

Lewis C, Latif Z, Hill M, Riddington M, Lakhanpaul M, Arthurs OJ, et al. ‘We might get a lot more families who will agree’: Muslim and Jewish perspectives on less invasive perinatal and paediatric autopsy. PLOS ONE 2018;13:e0202023.

Lewis C, Riddington M, Hill M, Arthurs OJ, Hutchinson JC, Chitty LS, et al. Availability of less invasive prenatal, perinatal and paediatric autopsy will improve uptake rates: a mixed-methods study with bereaved parents. BJOG 2019;126:745–53.

Data-sharing statement

All available qualitative data arising from the study is contained in the report and nothing further can be shared. All data requests should be submitted to the corresponding author for consideration. Access to available anonymised data may be granted following review.

Patient data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health and Social Care.