METHODS

The register

All SUDS in the Republic of Ireland are notified to the coroner and then registered with the National Sudden Infant Death Register (NSIDR). Each quarter, the Central Statistics Office (CSO) issues to the register details of deaths in children from birth to 2 years. For each death, a medical certificate of the cause of death and CSO form 102 are provided. The NSIDR compiles these data in accordance with the Data Protection Act. In addition, a system of notification ensures that the appropriate professionals report all suspected cases of SUDS to the NSIDR within 48 hours and the register then obtains information regarding birth details, necropsy report, medical certificate as to cause of death, CSO form 102, medical history, and medical report from the general practitioner. Parents are also invited to participate in home interviews at which a comprehensive questionnaire is completed. The information collected by the NSIDR provides comprehensive data on all sudden, unexpected deaths among the infant population in the Republic of Ireland.

Subjects

Cases were selected from the NSIDR where a full necropsy report was available. The necropsy report is obtained from the coroner in whose jurisdiction the child has died. In this study, reports were collected from 48 coroner districts in the Republic of Ireland.

Data analysis

Necropsy reports from the years 1994–1996 and 1998–2000 were evaluated. The quality of the reports were scored by a pathologist using a modification of the system described by Rushton,⁶ which includes factors that are identified by the Royal College of Pathologists as being an essential part of the necropsy in an infant who dies suddenly and unexpectedly.^9,10 The parameters that were scored included: (1) body weights and measurements, (2) organ weights, (3) quoting normal values, (4) histology of main organs (heart, liver, lungs, kidneys, and brain), (5) histology of other organs (thymus, spleen, and adrenals), (6) radiology, (7) microbiology, (8) biochemistry, (9) toxicology, and (10) special tests (such as metabolic test and virology). Table 1 provides a detailed breakdown of parameters and scores assigned. The total maximum obtainable score was 500 and the minimum accepted score (MAS) was set at 205, based on previous similar studies.^5–7

The study also set out to assess the value of performing the necropsy and to determine whether any additional information was obtained from detailed postmortem investigations and whether they were contributory to the cause of death. This was assessed by examining the final clinicopathological conclusion in the reports and categorising the data as: (1) new diagnostic information obtained that contributed to the cause of death, (2) additional information obtained but non-contributory to cause of death, and (3) no new information obtained. The data were analysed using the Student’s t test.

RESULTS

According to the CSO, there were 307 993 live births registered during the years 1994–1996 and 1998–2000. During this period, 274 cases of sudden and unexpected infant death were registered with the NSIDR, and the necropsy reports were available for assessment in 245 cases. The mean age of the child at death was 18.2 weeks. The male to female ratio was 1.4 : 1. The necropsies were performed in 36 different hospitals in 48 coroner’s districts around the Republic of Ireland.

Total scores for the quality of the necropsy reports ranged from 0 to 500. The median score was 184. The overall quality of the necropsy report was below the MAS of 205 in 55.5% (136 of 245) of all cases (fig 1). Twenty four reports scored 0 and one report scored 500. One hundred and fifteen of the necropsies (47%) were performed in a regional paediatric pathology centre by specialist consultant paediatric pathologists. Four necropsies were performed by consultant forensic pathologists and the remainder by general consultant histopathologists. The quality of the necropsies performed in regional centres was significantly higher than those performed in non-regional centres (p < 0.001). The mean score for necropsies performed in a regional centre was 320.7 compared with 111.4 for necropsies performed in non-regional centres. The quality score for the necropsy performed in a regional centre failed to achieve a MAS in 21% of cases compared with 84% for those done in non-regional centres. There was no relation identified between the quality of the necropsy performed and the age of the infant at death.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Overall quality of necropsy scores, 1994–2000. Total scores for the quality of the necropsy reports ranged from 0 to 500. The median score was 184. The overall quality of the necropsy report was below the minimum accepted standard (MAS) of 205 in 55.5% (136 of 245) of all cases. Twenty four reports scored 0 and one report scored 500.

The total quality score was then analysed according to the individual parameters that were assessed. This revealed that histology, body weights and measurements, and organ weights had the highest scores (fig 2), whereas radiology, microbiology, and biochemistry had the lowest scores. Although toxicology and metabolic tests are difficult to perform and not widely available, they had higher scores than microbiology and biochemistry. Performing histology yielded more information than any other test performed, whether in cases where a cause of death was found or as additional information in SIDS.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Summary scoring of individual necropsy parameters. The data are expressed as the mean score as a percentage of the maximum obtainable score for each parameter.

The necropsy scores were then analysed in the six individual years evaluated. The mean quality score in the year 1994 was 147, compared with a mean score of 260 in 2000, with a gradual improvement in quality with time (fig 3). When comparing the years before (1994–1996) and after (1998–2000) the issue of the guidelines in Ireland, there was a significant improvement in the quality of the necropsies performed (mean score 166 v 245; p < 0.01).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Mean quality of necropsy score by year. In comparing the years before (1994–1996) and after (1998–2000) the issue of the guidelines in Ireland, there was a significant improvement in the quality of the necropsy performed (mean score, 166 v 245; p < 0.01).

The second aim of the study was to determine the value of the necropsy and to assess the relevance of the additional pathological information obtained. In 13.4% (n = 33) of cases, a cause of death was found after necropsy. The remaining 86% (n = 212) of cases were classified as SIDS, in which no cause of death was found. In 18% (n = 45) of the SIDS group, additional information was found at necropsy, which was non-contributory to the cause of death. There was no significant difference in the quality of the necropsy performed if a cause of death was identified at necropsy. Assessment of the overall quality scores showed a mean score of 205 where a cause of death was found and a score of 210 where no cause of death was found (p = 0.7). However, in the SIDS group, the finding of additional pathological information was significantly related to the extent of the necropsy. The mean quality score was 341.6 when additional information was obtained and 175.2 in the absence of additional findings (p < 0.001).

DISCUSSION

The overall infant mortality rate in the Republic of Ireland has declined from a rate of 8/1000 live births in 1991 to 6/1000 live births in 2000. This mortality rate remains one of the lowest among developed countries, and results, in part, from the decline in neonatal mortality and SIDS. Whereas the rate of SIDS has been declining, the proportion of SIDS as a percentage of SUDS remains constant.¹² In contrast to previous studies,^4,5 we only evaluated those infants who died suddenly and unexpectedly, so that perinatal deaths, spontaneous abortions, stillbirths, and expected infant deaths were excluded. Our study was made possible by the fact that all SUDS in Ireland require a coroner’s necropsy, so that necropsy reports were made available to the NSIDR.

Fifty five per cent of the SUDS necropsy reports were below the accepted standard. In two comparable reviews of SUDS necropsy reports in the UK, substandard rates of 68% and 91% have been reported.^4,5 When all infant deaths are included, failure rates are approximately 46%.^4–6 When analysing the individual parameters assessed, in keeping with previous studies,⁶ body weights and measurements were well recorded, with special techniques such as radiology and microbiology scoring < 25% of their potential maximum score. However, although most of the reports included body weights and measurements, < 12% compared them with the normal expected values, questioning the purpose of performing this procedure. Moreover, five reports did not include the age of the child at death, so that in those cases, body weights and measurements were meaningless. It must be emphasised that radiology was not included in the MAS, although ideally it should be part of a standard necropsy because, in its absence, non-accidental injury cannot be excluded. Similarly, metabolic tests were not considered part of the MAS, yet if not included, there is the risk, albeit small, of an inherited metabolic disorder being overlooked.

“We feel there should be no minimum accepted standard”

The fact that necropsies performed in regional centres produced a higher quality score than non-regional necropsies supports the concept that infant necropsies should be performed in a regional paediatric centre. This finding was also reported by Vujanic et al,⁵ who found that there was an increase in paediatric necropsy referrals to specialist centres after the revision of guidelines.⁷ In a follow up study, where 46% of all cases were below the accepted standard, only 4.9% of those were performed by paediatric pathologists.⁴ In Ireland, in recent years, there also has been a tendency to move towards this practice; however, with only a handful of paediatric pathologists, this is not always feasible. In addition, if a child dies in a remote district, it is not always geographically possible to perform the necropsy in a specialist centre, particularly when trying to achieve a balanced sensitivity towards family predicaments.

There are limitations to our study. The scoring system involved an arbitrary score given to each parameter, which is clearly easy to apply to each case, but it does not give us insight into the expertise of the pathologist performing the necropsy, the facilities available, and the ability to interpret the information gathered. For example, in one case, a diagnosis of aspiration pneumonia was given in the absence of pulmonary histology. A recurring problem in evaluating reports was in reviewing histology, it was often impossible to ascertain which organs were examined, because a blank statement “histology was normal” was given. Furthermore, in our study, a “minimum accepted standard” was outlined which, in practice, could be easily achieved. In reality, this standard would not be accepted in a specialist centre. Nevertheless, any hospital that has the facilities for a pathologist to do a necropsy must have a weighing scales, a measuring tape, and techniques for histology—a few simple items which alone would achieve the MAS score. However, it must be stressed that we feel there should be no minimum accepted standard. Indeed, the professional guidelines do not include or support a “minimum practice”.

Interestingly, although toxicology and metabolic tests are somewhat difficult to perform and are not widely available, they had higher scores than microbiology and biochemistry, which are available in every hospital. The reason for this is unclear, but we speculate that this may be related to the ease of taking the samples for these tests and the time available: toxicology involves taking blood/urine only, and general histopathologists would be very familiar with performing this test for adult necropsies. In contrast, microbiology requires having sterile swabs and collection utensils close to hand, and these would be infrequently used for necropsies. In addition, biochemistry is best performed on vitreous humour, and most general pathologists would not routinely perform this test. Metabolic tests may have earned higher mean scores because any one metabolic test was required to achieve the maximum score for that parameter (table 1).

It is interesting to note that where a cause of death was found, the quality of the necropsy did not appear different to that when no cause was found. This is possibly because if a cause of death was found at necropsy, the pathologist then limited the number of additional ancillary investigations (such as metabolic tests and virology) because they were no longer indicated. In the SIDS group, the finding of additional pathological information was significantly related to a high quality necropsy score. Although these findings were non-contributory to death, they have potential value. Parents may feel guilty if a necropsy finds “nothing”. The finding of a minor pathological abnormality indicates that a detailed examination was performed.³ Furthermore, it may explain some of the non-specific illnesses that these children often have before death—for example, oesophagitis.

It is conceivable that the organ retention controversy may have had an impact on the quality of sampling by pathologists in Ireland. However, the controversy came to light in 2000, and would therefore not have had a significant impact on our study, which evaluated reports from 1994–96 and 1998–2000. In addition, all of the cases in our study were coroner’s necropsies, so that the issue of consent for organ retention did not occur. Nevertheless, a policy was endorsed at the end of our study whereby parents and next of kin were actively informed if an organ was retained.

One of the positive findings of our study was the improvement in necropsy scores after the issue of the Faculty of Pathology guidelines in Ireland in 1996. However, during the years before and after the issue of the guidelines, there was not a significant increase in referrals to specialist centres: 48% v 46%, respectively. This is encouraging and suggests that the non-regional centres are improving their practice in an attempt to achieve the high standard already present in regional centres. Although the improvement was predominantly in the regional centres (score 307 to 432), the non-regional centres also had better scores (70 to 138).

In summary, the overall quality of SUDS necropsies in Ireland is less than adequate. Although the standard of necropsy reporting has improved in recent years, this study emphasises the greater need for pathologists to continue to adhere to the published guidelines. In addition, it highlights the importance of auditing the effect of introducing practice guidelines on clinical practice to complete the audit loop.