In living patients, the identity of a microorganism in blood culture (a normally sterile site; NSS) is an independent predictor of primary significance. For example, in cases of true bacteraemia/fungaemia the finding of Staphylococcus aureus, Streptococcus pneumoniae, Streptococcus pyogenes, Streptococcus agalactiae, coliforms (Escherichia coli and other Enterobacteriaceae), Pseudomonas aeruginosa, Haemophilus influenzae, Neisseria meningitidis, Neisseria gonorrhoeae, Listeria monocytogenes, Bacteroides spp, Candida spp and Cryptococcus neoformans nearly always (>90%) represent true infection. By contrast, there has been a reluctance to accept that the same is true in patients after death. Arguments abound that such organisms have translocated from mucosal or intestinal sites or are the result of contamination during the autopsy. All may be true but it is apparent that these events are relatively rare in cases where autopsies have been performed and bodies have been kept under appropriate conditions.1 The finding of a pathogen in a NSS such as the blood, spleen or cerebrospinal fluid (CSF) has often been overlooked as a cause of death.

This paper addresses the issue of sterile site bacteriology in SIDS and uses for comparison a group of babies who have died suddenly and unexpectedly by accident or infectious cause. From the data, a guideline is derived to assist in increasing the accuracy of diagnosis in sudden unexpected deaths in infancy (SUDI).

METHODOLOGY

Case definition/criteria

Autopsies were conducted by experienced paediatric pathologists. The definition of SIDS used for the purpose of this study was that developed in 2004.2 Non-SIDS cases comprise the age-matched comparison group (“controls”) who died from a variety of natural (non-SIDS) causes as detailed in table 1.

Cases of SIDS and comparison deaths (sudden accidental deaths and deaths caused by proven infection) were those that occurred in South Australia (total population ∼1.2 million) between the years 1978­ and 2004. Autopsy reports from 130 SIDS cases (mean age 4.08; range 0.75–12 months), 32 age-matched cases of death due to infection (mean age 5.16; range 0.75–12 months) and 33 age-matched cases deemed to be non-infectious sudden deaths (mean age 4.02; range 1.2–10 months) were examined and bacteriological findings were recorded. A number of the non-infectious sudden deaths were unable to be deemed as SIDS because of underlying disease such as congenital heart defects or cerebral palsy. The autopsies had been conducted by experienced paediatric pathologists, and bacteriological specimens were collected using aseptic technique involving searing of the organ surface prior to aspiration with a sterile needle and syringe or insertion of a sterile swab into the tissue or obtaining a piece of tissue under aseptic conditions. Culture results from the sterile sites of heart blood, spleen and CSF were recorded. In addition, bacteriological findings from lung cultures were recorded for comparison.

Statistical analysis

Fisher’s exact test and χ² were used to calculate the significance of sterile site data using the Epistat program.

RESULTS

Isolation of significant bacterial pathogens from normally sterile sites

Sterile site infection (table 2) was usually not found in cases of sudden accidental death (eg, motor vehicle accident or drowning); however, the finding of true pathogens such as S aureus in sterile sites in SIDS and deaths associated with infection was relatively common. Fourteen out of 130 (10.76%) SIDS cases had S aureus in a sterile site (heart blood 7, spleen 4, CSF 3) compared with 6/32 (18.75%) in cases of infection-related deaths (heart blood 5, spleen 1). S aureus was not found in sterile sites in sudden accidental deaths (p = 0.036, Fisher’s exact test); however, S aureus was found in the lung in three of the 33 (9.1%) cases of sudden accidental deaths (p = 0.078). The incidence of coliform bacteria in sterile sites in SIDS (23/130 (17.7%)) was similar to that seen in 5/33 (15.2%) sudden accidental deaths, but, as was expected, was higher in cases of infection-related deaths (10/32 (31.2%)). Sterile site culture yielded S pneumoniae in 2/130 (1.5%) SIDS, no cases of sudden accidental death and 5/32 (15.6%) of infection-related deaths. Similar figures were seen with H influenzae (5/130 (3.1%), 0% and 4/32 (9.4%) respectively). Sterile-site bacteriology yielded no growth in 15/33 (45.5%) sudden accidental deaths, 56/130 (43%) SIDS and only 9/32 (28.1%) infectious causes of death (all patients had received antibiotic treatment but had died of complications arising from the primary infection diagnosed during life: N meningiditis sepsis, H influenzae type b meningitis, staphylococcal bronchopneumonia and the presumptively infectious entity Kawasaki disease).

Isolation of significant pathogens from lung tissue

As expected, S aureus was cultured more frequently from the lung of infection-related deaths 11/32 (34.4%) compared with SIDS (28/130 (21.5%); p = 0.19; table 3). Coliforms were found in the lungs of 7/33 (21.2%) of accidental deaths, in 8/32 (25%) of infection-related deaths and 21/130 (16.2%) of SIDS.

Isolation of other micro-organisms

Organisms regarded as post-mortem translocation of organisms (eg, non-fermentative Gram-negative bacilli, Clostridium spp) were uncommon for all three groups of autopsies and contaminants (coagulase-negative staphylococci, diphtheroids, etc.) were found more commonly in SIDS (40.8%) than the other two groups of autopsies (table 4). Non-fermentative Gram-negative bacilli were not isolated from lungs in any autopsy group. Mixed upper respiratory flora were commonly isolated from the lungs of all three groups.

Effect of post-mortem interval on bacteriological results

Analysis of mean post-mortem intervals (time from death to time of autopsy) showed no effect on bacteriological results. Mean (SD) post-mortem intervals for SIDS cases with no growth or positive significant bacterial culture from sterile sites were 27.5 (14.9) h and 21.9 (21.3) h, respectively. The mean post-mortem interval of SIDS cases from whom S aureus was isolated from sterile sites was 25.3 (21) h and for SIDS cases from whom S aureus was not isolated was 23.06 (25.8) h. Similarly, insignificant differences in post-mortem intervals were seen for other pathogens found in the three groups of cases.

DISCUSSION

The value of unexpected positive microbiological findings in sudden unexpected infant death was demonstrated by Sadler et al3 who showed that 16% of apparent cot deaths were explained on the sole basis of these findings, that is, most were pneumococcal or meningococcal meningitis and/or septicaemia. These authors went so far as to recommend mandatory routine early bacteriological culture of the CSF (by cisternal puncture) and blood before necropsy in the investigation of all sudden infant deaths.3 Similarly Sonnabend et al4 whose careful bacteriological study of 70 babies found eight (11.4%) babies had evidence of overwhelming bacterial infection (based on post-mortem cultures) as the cause of death. These authors commented that “the post-mortem cultures were of diagnostic value, providing the sole means of identifying the cause of death”.

Morris et al1 reviewed published articles on post-mortem bacteriology and examined 468 cases of SUDI. They showed that in studies in which careful precautions had been undertaken to reduce contamination, approximately two thirds of blood cultures were negative; the present study revealed heart blood was indeed sterile in 53.8% of cases. The review concluded that, for adults and neonates, a pure growth of a known pathogen carries a >50% likelihood of being found in association with genuine infection. Although the main post-mortem artefact is contamination, with careful technique this can be almost eliminated. The present study showed no effect of post-mortem interval on isolation rate and supports the similar findings of Morris et al. Thus it is considered that “agonal” spread (if it exists) is less common than originally thought. Morris et al showed that post-mortem translocation was not a problem if the bodies are appropriately stored. These authors concluded that a pure growth of a pathogen in blood or CSF should be regarded as a possible contributing factor to death at all ages. Indeed the study by Eisenfeld et al5 showed that in all 43 infants with both bacteraemia and meningitis, bacteria isolated from both respective sites were identical. In addition, these post-mortem cultures were of diagnostic value, providing the sole means for definitive bacteriologic diagnosis in 82 (65%) of the 126 infected infants. The findings of the present study would support the contention that a pure growth of a pathogen in a sterile site should be regarded as a possible contributing factor to death. Moreover, in this context, the normally sterile site of the spleen should, in future, accompany blood and CSF.

The Avon UK SUDI study,6 conducted between 1987 and 1989, included post-mortem bacteriological examinations in 95 babies affected by SUDI aged from 1 week to 2 years. These included blood and CSF obtained within a median interval of 3.5 h from the discovery of death (range 0.25–46 h). Further tissue samples were obtained when the necropsy was conducted (median time, 25 h; range 2.8–73 h). The blood cultures yielded bacterial pathogens in eight out of 95 cases. Spleen cultures were positive for pathogens in four out of 94. The CSF was positive for pathogens in one out of 95. The report did not document the number of cultures containing organisms considered to be contaminants.

The CESDI-SUDI study was a case control study of SUDI carried out in England between 1993 and 1996.7 There were 456 cases of SUDI and four controls for each case. All cases of SUDI had a detailed necropsy, which included the examination of specimens for bacteriology. Blood cultures were recorded in 287 cases, and 144 (50%) were sterile. Mixed growth was obtained in 130 cases (45%). A major pathogen with corresponding histological changes was found in seven cases; in eight cases there was a single pathogen without corresponding histology. In the latter group the organisms were S aureus in three cases, group B beta-haemolytic streptococcus in three cases, and α-haemolytic streptococcus in two cases. Of the 279 CSF cultures, 211 (76%) were sterile. There were seven positive cultures with corresponding histological changes in the brain. The organisms were N meningitides (three cases), H influenzae type B, E coli, group A beta-haemolytic streptococcus, and S epidermidis. In a further 10 patients, a single pathogenic organism was grown from the CSF without corresponding histological changes in the brain. The organisms listed were N meningitidis, S pneumoniae (two cases), haemolytic streptococcus (three cases) and haemophilus (two cases). In most of the last group, the organisms were grown from other sites and were associated with respiratory tract infection. In 51 other cases there was mixed growth or growth of single organisms considered to be contaminants. Thus, blood cultures yielded 15 possibly significant results (5.2%) and CSF cultures yielded 17 positives (6%).

As mentioned above, Sadler et al3 stressed the value of bacteriological investigations in cases of SUDI. He presented data on 95 necropsies of patients aged from 3 days to 29 months, 63 of whom presented as “cot deaths”. The protocol included obtaining, at necropsy, CSF by cisternal puncture, blood from the subclavian vein and lung tissue and spleen. Nine diagnoses depended on positive bacteriological findings. Adelson and Kinney8 studied 126 cases of SUDI (age range 10 days to 2 years) in which detailed post-mortem examinations were conducted and specimens obtained for bacteriology. Cultures of heart blood, spinal fluid, pharynx, ileum and lungs were made in each case. Of 120 heart blood cultures, 95 (79%) were sterile, and the other 25 yielded a total of 32 organisms (19 significant pathogens).

Although Sonnabend and colleagues4 failed to clarify how many cultures had contaminants, their otherwise careful bacteriological study of SUDI allowed them to diagnose the cause of death on the basis of the microbiological results alone. Their assessment depended on the nature of the organism grown and the number of positive sites. Eight of 70 patients were judged to have disseminated bacterial infection. Sadler et al3 were also prepared to diagnose the cause of death in SUDI on the basis of a significant pathogen isolated at necropsy in the absence of histological confirmation. The present study would also support such a position based on finding known pathogens in one or more normally sterile sites (heart blood, spleen and CSF).

The study of SUDI by Gilbert et al7 repays careful analysis. The authors found increased throat carriage of S aureus (OR 5), coliforms (OR 29) and group B streptococci (OR 11) in a comparison with normal healthy age-matched infants. In epidemiological studies with the requisite statistical power, ORs of this level would normally be judged as possible causative factors. However, in this case, the authors were cautious because of the possibility of a post-mortem artefact. In simple terms, they did not trust the microbiological results obtained after death because of the possibility that the bacteria had in some way been introduced during the interval between death and obtaining the specimen. A mechanism by which this could occur, however, is not clear. The present study was unable to compare upper respiratory tract (URT) results with the study of Gilbert et al7 because specimens were not routinely obtained from the URT from the South Australian SUDI cases and so discussion on this aspect of post-mortem bacteriology is omitted.

One potential criticism of the present study is that cases were drawn from autopsies conducted over a long time-span; this was unavoidable because of the relative rarity of non-SIDS SUDI age-matched cases. The possibility of seasonal variation in bacterial pathogen carriage remains an unavoidable weakness of this study.

The finding of toxigenic bacteria in various sterile and or non-sterile sites (eg, lung) in the absence of significant inflammatory responses in cases of SUDI could explain these deaths on the basis of a toxaemic event. This is biologically plausible because some of the species isolated can produce potent toxins (eg, pyrogenic toxins of S aureus and soluble toxins of E coli) that could contribute to death by means similar to those observed in toxic shock syndrome.9^–12 Indeed, staphylococcal toxins have been demonstrated in the tissues of a large proportion of babies deemed to have died of SIDS.12

It has been hypothesised that pro-inflammatory cytokine responses to bacterial infection and/or toxins play a role in triggering a “cytokine storm” resulting in SIDS. Support for this is found in the relationship between pro-inflammatory gene polymorphism frequencies (eg, IL-6 and TNF-α) in SIDS compared with controls. Comparison of isolation rates of key bacteria (eg, S aureus) that could play a role in this support this idea. The presence of a genetic polymorphism denoting a low IL-10 production in combination with additional factors such as exposure to cigarette smoke would contribute to an increased possibility of the occurrence of a SIDS event.12

In conclusion, NSS cultures can provide important information regarding possible causes of SUDI. It is noteworthy that the Kennedy Report, in its autopsy protocol, specified collection of cultures from blood and CSF but not the spleen.13 The recent study by Weber et al14 provided data on sterile sites (including the spleen) and showed similar findings to the present study with significantly more cultures from unexplained SUDI (ie SIDS) having S aureus and E coli than other causes of infant death. Methods of analysis of data in these two studies differed and each study provided interesting angles on and interpretation of infection in SUDI. Both these studies showed that isolation of a recognised pathogen in pure or mixed culture from a normally sterile site should alert the pathologist to the possibility of the pathogen having a causal role in an infant's death. Future SUDI autopsy guidelines should be modelled on both studies. On the basis of this and the cited studies, isolation of S aureus, in particular, should raise the possibility of causal involvement, especially if the isolate is shown to carry enterotoxin or TSST-1 genes. When organisms are isolated from non-sterile sites (such as the upper respiratory tract) interpretation is more difficult; but we should allow our minds to remain open to the possibility of toxigenic organisms playing a causal role in SUDI. Nevertheless, the finding of significant pathogens in SUDI should always be considered in the context of the many factors (eg, neglect, smoking and sleep position) involved in the complex investigation of SUDI.