Perinatal and infant autopsy rate in Wales over a ten year period
Sarah J Kotecha, Cardiff University; Kim Rolfe, Cardiff University; W John Watkins, Cardiff University; Sailesh Kotecha, Cardiff University; Roshan Adappa UHW.
A full autopsy performed by a perinatal pathologist is an optimal method to investigate perinatal and infant deaths1. A study of deaths in a tertiary neonatal referral centre for the south east of Scotland over a ten year period from 1990- 1999 noted that a neonatal autopsy was performed in 67% of cases and new information was obtained in 50% of autopsies. However, they reported that over the 10 year study period, there was a decline in the autopsy rate2. In Wales, perinatal and infant autopsy rates were studied over a ten year period 1994-2003. Overall, parental consent for an autopsy was requested in 89% of cases and permission for an autopsy was obtained in 68% of cases. The study also reported a decline in autopsy rates over the 10 years studied with parental refusal being noted as the main cause of the decrease3. There is concern that this decline may be due to parental concerns about organ retention.1 In addition, clinicians may also be reluctant to seek consent for an autopsy1.
The aim of this commentary is to review the perinatal and infant autopsy rates in Wales over the last ten year period and to investigate whether rates varied depending on year of birth and type of perinatal and infant death e.g. stillbirth or neonatal death etc.
All stillbirths, spontaneous abortions, therapeutic abortions and infant deaths between 1st January 2003 and 31 December 2012 in Wales were identified using data contained in the AWPS database. Late terminations were included. Data was collected on whether or not the autopsy was performed and reasons for not performing the autopsy were noted as “not requested” or “not permitted”. The rates of autopsy performed per year and by type of case were studied. We investigated for 2008-2012 whether or not an autopsy changed the Clinico-Pathological (CP) classification by year and type of case and whether for 2009-2012 the placenta was sent for histology or not. The differences in the rates of placentas being sent for histology for the period 2009-2012 was investigated using a chi-square test.
In the period 2003-2012 we identified 4618 deaths. No data regarding autopsy were available for 286/4618 (6.2%) of cases, these were excluded from further analysis leaving 4332 cases. In 573/4332 (13.2%) of cases permission for an autopsy was not requested. Consent for an autopsy was requested in the remainder of cases 3759/4332 (86.8%) and parental consent was given in 1866/3759 (49.6%) of those approached. Hence there was an overall autopsy rate of 1866/4332 (43.1%).
The rates of request for permission and permission granted for an autopsy according to each study year are shown in table 1. It seems that permission for an autopsy is being requested more in the later part of the decade but consent granted for an autopsy is between 38.2% to 60.2% during the decade.
The rates of request for permission and permission granted for an autopsy according to type of case are shown in table 2. Permission for an autopsy is being requested less when an early or late or post neonatal death occurs and rates of consent given for an autopsy are lower when an early or late neonatal death occurs. The post-neonatal deaths may appear to have a high consent rate as many of these cases are referred to the local coroner for further investigation and thus the parents may have little choice.
We studied whether or not an autopsy changed the CP classification by year. The results are shown in table 3. For example it is possible that a code could have changed from unclassifiable to a congenital anomaly. There is significant missing data hence the number of deaths are different to table 1. It appears that an autopsy changed the CP classification in approximately 35% of cases where information was available.
We examined whether or not an autopsy changed the CP classification by type of case. The results are shown in table 4. There is a significant amount of missing data hence the number of deaths are different to table 1. Overall an autopsy changed the CP classification in 35.3% of cases where information is available but highest percentage was seen in neonatal deaths where interestingly overall rates of consent for an autopsy are the lowest.
Information on whether or not the placenta was sent for histology was available from 2009 onwards and is shown in table 5. There is some missing data hence number of deaths are different to table 1. In the later years placentas were sent for histology more often than in the earlier years. The rate of placentas being sent for histology in 2009 was significantly different (p<0.05) to the rates in the other three years as was the rate of placentas being sent for histology in 2010. The rate of placentas being sent for histology in 2011 was not significantly different to 2012.
It appears that in Wales over the 10 year period between 2003-2012 rates of consent being requested for an autopsy have increased but rates of permission being granted remains at just below 50%. Of particular note is the fact that autopsy is being requested less when an early or late or post neonatal death occurs and rates of consent given for an autopsy are lower when an early or late neonatal death occurs. The post neonatal deaths may have a high consent rate as in many cases the autopsy is requested by the coroners. Autopsies are providing additional and/or new information in approximately 30% of cases as CP classification codes were changed after the autopsy. Placentas have been sent for histology more frequently in the last few years.
The results from AWPS data are in line with the previous study using Welsh data which noted that lack of parental consent for an autopsy is the main cause of the low autopsy rate3. This is despite another publication from Wales which showed how invaluable an autopsy is and how much clinically relevant information can be obtained4. Wales’ low autopsy rate is not unique: an Australian report published in 2006 noted that perinatal autopsy rates had fallen over the previous ten years. They concluded that fewer perinatal autopsies were being carried out following the organ retention controversy5. A regional audit of perinatal and infant autopsies in Northern Ireland reported a similar autopsy rate of 47.4%11. In the West Indies, autopsies were requested more often when a stillbirth had occurred than when an early neonatal death had occurred and the average stillbirth autopsy rate was higher than the average early neonatal death autopsy rate6. Again this is in line with this Welsh data and we have been able to demonstrate a similar pattern where an autopsy is being requested less when an early or late or post neonatal death occurs and rates of consent given for an autopsy are lower when an early or late neonatal death occurs6.
Studies have investigated parental and professional attitudes towards perinatal autopsy. Much of this work has focused on stillbirths. Holste et al. researched the mothers’ attitudes towards autopsy of their stillborn babies7. They reported that mothers’ did not regret their decision consenting or not consenting for an autopsy but they reported mothers’ do not always receive timely and thorough information concerning the results7. In another study, nearly twice as many parents who declined an autopsy after a stillbirth regretted their decision when compared to those who had an autopsy8. The most common reason for the autopsy from parents point of view was to discover the underlying reason for their baby’s death; but many professionals and parents ranked a prolonged waiting period for results and emotional distress as barriers to giving consent for an autopsy8. A recent Cochrane Review, which investigated interventions to support parents’ decisions regarding autopsy after a stillbirth occurred, concluded: “support for parents making decisions about autopsy or other post-mortem examinations after stillbirth must rely on the ad hoc knowledge and experience of those involved at the time.” 9
The Welsh Initiative for stillbirth reduction recognised the importance of autopsies and placental examinations after a stillbirth. They formed a Perinatal Pathology Sub-group which aimed to explore issues around improving consent to autopsies. This included improving/standardising consent training, improved information for families, identifying gaps in the current consent package, improved transfer of the baby for an autopsy, prompt turn-around of reports and regular meetings between pathology staff and health boards. The improved transfer involves minimising the time the baby is away from the parents.
Other work undertaken by the initiative included work around
- Exemplar policy and guidance for management of reduced fetal movements
- Implementation of the GROW assessment tool for the identification of growth restricted and retarded fetuses across all Welsh health Boards.
Placentas have been sent for histology more frequently in recent years. This is important as it has been reported that in nearly half of the cases where a placenta was pathologically examined after a stillbirth, the findings influenced the subsequent classification of the stillbirth and in 16% of the cases the classification was derived mainly from the histological examination of the placenta10. Thus, it is important that placental histology is requested even if consent for a full autopsy is not available.
We have shown that the CP classification codes are still being modified after an autopsy highlighting that additional and/or new information can be obtained frequently after an autopsy similar to the previously described study in Wales where after an autopsy, the CP classification was changed in 13% of cases; new information noted in 26% of cases; and in 18% of cases, the autopsy revealed the cause of death4. The regional audit of perinatal and infant autopsies in Northern Ireland also reported that the Wigglesworth classification was changed in 20.7% of cases after an autopsy11.
In conclusion there is little doubt that an autopsy after a perinatal or infant death can provide valuable additional information. Rates of autopsy remain low in Wales despite consent for an autopsy being sought more often in the later part of the decade studied. Rates of autopsies are greater for some types of perinatal deaths than others. Histological examination of the placenta may provide valuable additional information and should be requested even when consent for full autopsy examination has not been given.
1. Lyon A. Perinatal autopsy remains the “gold standard”. Arch Dis Child Fetal Neonatal Ed 2004;89:F284.
2. Brodlie M, Laing IA, Keeling JW, McKenzie KJ. Ten years of neonatal autopsies in tertiary referral centre: retrospective study. BMJ;324:761-763.
3. Adappa R, Paranjothy S, Roberts Z, Cartlidge PHT. Perinatal and infant autopsy. Arch Dis Child Fetal and Neonatal Ed 2007;92:F49-F50.
4. Cartlidge PHT, Dawson AJ, Stewart JH, Vujanic GM. Value and quality of perinatal and infant post-mortem examinations: cohort analysis of 400 consecutive deaths. BMJ 1995;310:155.
5. Khong TY, Tanner AR. Foetal and neonatal autopsy rates and use of tissue for research: The influence of ‘organ retention’ controversy and new consent process. Journal of Paediatrics and Child Health. 2006;42:366-369.
6. Bishop KL, Dupuis C, Nanton P, Clarke K, Bolt C, Chin-See C. Perinatal Autopsy Rates at the University Hospital of the West Indies: 2002-2008. West Indian Med J 2013;62(1):35.
7. Holste C, Pilo C, Pettersson K, Radestad I, Papadogiannakis N. Mothers’ attitudes towards perinatal autopsy after stillbirth. Acta Obstetricia et Gynecologica Scandinavica 2011;90:1287-1290.
8. Heazell AEP, Mclaughlin M-J, Schmidt EB, Cox P, Flenady V, Khong TY, Downe S. A difficult conversation? The views and experiences of parents and professionals on the consent process for perinatal post-mortem after still birth. BJOG 2012;119(8):987-97.
9. Horey D, Flenady V, Heazell AE, khong TY. Interventions for supporting parents’ decisions about autopsy after stillbirth. Cochrane Database Syst Rev. 2013 Feb28;2:CD009932.
10. Heazell AEP, Martindale A. Can post-mortem examination of the placenta help determine the cause of stillbirth ? Journal of Obstetrics and Gynaecology 2009;29(3);225-228.
11. Thornton CM, O’Hara MD. A regional audit of perinatal and infant autopsies in Northern Ireland. British Journal of Obstetrics and Gynaecology 1998;105:18-23.