Potential of Minimally Invasive Tissue Sampling for Attributing Specific Causes of Childhood Deaths in South Africa: A Pilot, Epidemiological Study

Abstract Background Current estimates for causes of childhood deaths are mainly premised on modeling of vital registration and limited verbal autopsy data and generally only characterize the underlying cause of death (CoD). We investigated the potential of minimally invasive tissue sampling (MITS) for ascertaining the underlying and immediate CoD in children 1 month to 14 years of age. Methods MITS included postmortem tissue biopsies of brain, liver, and lung for histopathology examination; microbial culture of blood, cerebrospinal fluid (CSF), liver, and lung samples; and molecular microbial testing on blood, CSF, lung, and rectal swabs. Each case was individually adjudicated for underlying, antecedent, and immediate CoD by an international multidisciplinary team of medical experts and coded using the International Classification of Diseases, Tenth Revision (ICD-10). Results An underlying CoD was determined for 99% of 127 cases, leading causes being congenital malformations (18.9%), complications of prematurity (14.2%), human immunodeficiency virus/AIDS (12.6%), diarrheal disease (8.7%), acute respiratory infections (7.9%), injuries (7.9%), and malignancies (7.1%). The main immediate CoD was pneumonia, sepsis, and diarrhea in 33.9%, 19.7%, and 10.2% of cases, respectively. Infection-related deaths were either an underlying or immediate CoD in 78.0% of cases. Community-acquired pneumonia deaths (n = 32) were attributed to respiratory syncytial virus (21.9%), Pneumocystis jirovecii (18.8%), cytomegalovirus (15.6%), Klebsiella pneumoniae (15.6%), and Streptococcus pneumoniae (12.5%). Seventy-one percent of 24 sepsis deaths were hospital-acquired, mainly due to Acinetobacter baumannii (47.1%) and K. pneumoniae (35.3%). Sixty-two percent of cases were malnourished. Conclusions MITS, coupled with antemortem clinical information, provides detailed insight into causes of childhood deaths that could be informative for prioritization of strategies aimed at reducing under-5 mortality.

on acceptability [5,6]. Minimally invasive tissue sampling (MITS), also known as minimally invasive autopsy, involves postmortem collection of fluid and solid tissue samples using biopsy needles. Community engagement studies have suggested that MITS would be theoretically culturally acceptable in African and South Asian countries [6]. Furthermore, a validation study undertaken in Mozambique demonstrated moderate concordance (κ = 0.75) in CoD attribution between complete diagnostic autopsy and MITS in children >1 month of age [7].
We aimed to evaluate the acceptability and utility of MITS, as proof of concept, in ascertaining the causal pathway of death in children >1 month of age in an LMIC setting (Soweto, South Africa).

Study Site and Population
The study was conducted from 29 June 2015 to 1 August 2016 at Chris Hani Baragwanath Academic Hospital (CHBAH). Further details of the study site, including healthcare setting and human immunodeficiency virus (HIV) prevalence, are given in the companion manuscript on neonatal deaths, also published in this supplement [8]. Additionally, in South Africa, the Expanded Programme on Immunization includes pneumococcal conjugate vaccine (PCV) and rotavirus vaccine that have been introduced since 2009, in addition to 8 other routinely administered childhood vaccines [9].

Study Design and Procedures
In this prospective, observational study, deaths of children <14 years of age occurring in the medical and emergency department of CHBAH, and those certified as dead upon arrival at the hospital, were identified through daily screening of death registries, excluding from 18 December 2015 to 3 January 2016 (vacation period). We initiated the study by initially only screening and enrolling deaths occurring in the general medical wards. The surveillance and enrollment of cases was expanded to include deaths occurring in the surgical burn unit (31 August 2015), hematology/oncology ward (16 September 2015), surgical wards (30 September 2015), and the nursery ward for care of stable very low-birth-weight infants (kangaroo-care facility; 7 October 2015). Details on the counseling and consenting processes are as detailed elsewhere in the companion paper on neonatal deaths [8].
We aimed to undertake MITS within 24 hours of death and excluded those cases in which the procedure could not be performed within 72 hours. Corpses were kept in the hospital mortuary at 4°C until retrieval for burial.

Minimally Invasive Tissue Sampling
Trained study staff (medical doctor or professional nurse assisted by research assistants) undertook MITS as detailed elsewhere in this supplement [8]. The testing algorithm for the collected samples was similar to that done in neonatal deaths, with the exception that molecular testing of blood samples using the FastTrack Diagnostics (FTD; Sliema, Malta) neonatal sepsis kit was not undertaken in the age group 1 month to 14 years; and only the FTD Respiratory-33 panel kit was used on lung samples. Methods used for tissue histology processing are also described elsewhere in this supplement [8].

Determination of Cause of Death
The CoD attribution was based on consensus of an international panel convened in South Africa from 26 March to 5 April 2017 composed of pathologists, pediatricians, epidemiologists, microbiologists, obstetricians, infectious disease specialists, and international coding and certification experts (the Determination of Cause of Death [DeCoDe] panel group is listed in the Acknowledgements). The panel, chaired by either Chris Wilson or Scott Dowell, deliberated on CoD attribution using the International Classification of Diseases, Tenth Revision (ICD-10) [10]. This included attributing a single underlying medical condition that likely initiated the sequence of events culminating in death, and the most proximal event (ie, immediate cause) of death, if applicable. Also evaluated were any antecedent conditions in the casual pathway leading to death, and medical conditions considered to have a contributory role, but not directly involved in the causal pathway. Details on the working methods of the DeCoDe panel are described in the companion manuscript on causes of neonatal deaths [8]. The DeCoDe panel allowed for inclusion of two "immediate" CoD, where concurrent diseases could not be prioritized as the main condition that caused the death; for example, histologically confirmed Pneumocytis jirovecii and respiratory syncytial virus (RSV) pneumonitis being concurrently present (and attributed as the immediate CoD) in a child with underlying HIV infection (attributed as the underlying CoD).
The CoD was generally based on consensus, in the absence of which a majority viewpoint was used, and recorded according to World Health Organization (WHO) guidelines for death certification. In addition, CoD was stratified into first-level United Nations Inter-agency Group on Child Mortality Estimation (UN-IGME) categories [11]: group I (an aggregation of communicable, maternal, perinatal, and nutritional conditions), group II (noncommunicable diseases), and group III (injuries).

Statistical Analysis
We stratified analyses to infants (1-11 months of age), children (12-59 months of age), and older children (60 months to 14 years of age). Descriptive statistics were calculated providing medians with interquartile ranges for continuous variables, and proportions for categorical variables. For select variables, differences between the infants, children, and older children were tested using Kruskal-Wallis or Fisher exact test. Weight-for-age was calculated in Stata software using the UK WHO Preterm Growth Charts (version UKWHOpreterm) and UK WHO Term Growth Charts (version UKWHOterm) for the preterm babies and children aged 0-20 years, respectively. Statistical analysis was done using Stata software version 15 (StataCorp, College Station, Texas).

Ethical Considerations
The University of the Witwatersrand Human Research Ethics Committee (number 150215) and the CHBAH Protocol Review Committee approved the study. Parents/legal guardians provided consent prior to any MITS study procedure.

RESULTS
There were 412 deaths among children aged 1 month to 14 years at CHBAH during the study period. Sixty percent (n = 247) of the deaths were screened for study participation, while 165 (40%) cases were not screened mainly due to the timing of phasing in of enrollment in the nonmedical wards ( Figure 1). Seven of the 247 (2.8%) screened cases were ineligible, and in 19.2% cases the parents could not be contacted within 72 hours of death. Of the remaining 194 screened cases, 127 (65.5%) consented to study participation, including 14 of 127 (11.0%) who were certified dead upon hospital arrival.
The median age of enrolled cases was 11 months; 52.8% were infants, 29.1% children, and 18.1% older children. The median time from hospital admission to death was 3.0 days. MITS sampling was done at a median of 20.2 (11.3-27.6) hours after death. Overall, 36.4% of cases were born to women living with HIV; however, only 12.8% of cases were confirmed to be HIV infected on postmortem HIV polymerase chain reaction testing. Overall, 62.4% of cases, including 71.7% of infants, were malnourished. Sixteen percent of cases had underlying congenital abnormalities. Approximately 20% of the cases had been admitted for mechanical ventilation support in the intensive care unit (Table 1).

Cause of Death Attribution
Using the findings from the molecular testing, microbiology, and histopathology in addition to the antemortem clinical findings, the DeCoDe panel assigned a CoD in 99% of cases. Details of the certainty of diagnoses for the different underlying CoD and immediate CoD categories are reported in Supplementary Tables 1 and 2. The DeCoDe panel was confident (level 1) for the majority of underlying (81.1%) and immediate CoD (77.2%) diagnoses. Overall, the UN-IGME categories for underlying CoD were 55.1% due to communicable diseases, maternal, perinatal, and nutritional conditions (group 1), 36% due to noncommunicable diseases (group 2), and 8% due to injuries (group 3) (Tables 2 and 3; Figure 2). In group I, these included deaths attributed to underlying HIV/AIDS (22.9%), diarrheal disease (15.7%), acute respiratory infections (14.3%), and complications of prematurity (14.2%).
The majority of deaths classified as UN-IGME group III (n = 10) occurred in children (60%) and older children (30%).  (Table 3). Among the group III deaths, hospitalacquired infection was the immediate CoD in 40.0% of cases, and community-acquired infection in 10.0% ( Figure 3; Table 2).

Overall Contribution of Infections Either as an Immediate or Underlying Cause of Death
Overall, infections were attributed as an underlying or immediate cause in 78% (n = 99) of deaths. This included 48 of 127 (37.8%) cases where infection was an underlying CoD, and 51 cases where the underlying CoD was a noninfectious medical condition (64.6% of 79).
Of the 6 deaths due to meningitis as the immediate cause, 4 were community-acquired (S. pneumoniae; Neisseria meningitidis and S. pneumoniae coinfection; E. coli; and Mycobacterium tuberculosis); and 2 hospital-acquired cases were due to A. baumannii and Candida albicans coinfection (n = 1) and E. coli (n = 1).

DISCUSSION
This pilot study demonstrates the value of MITS, interpreted with antemortem clinical data, in attributing highly specific causes of death in an LMIC setting where under-5 mortality rate (per 1000 live births) was estimated to be 53 in 2013 (unpublished data); compared to the national estimate of 42 per 1000 in 2015 [2]. The DeCoDe panel, with generally high level of confidence, attributed an underlying and immediate CoD in nearly all (99%) cases. Furthermore, for deaths associated with    infection, a specific pathogen was identified in 96% of cases. Elucidating such granular information on the causes of death could be extremely useful for planning and prioritizing future interventions aimed to reduce childhood mortality, even when the underlying cause itself might not be preventable (eg, premature birth).
f Group III category includes all conditions whose ICD-10 code is included in V01-Y89 [12].
g Ill-defined refers to a cause of death which could not be determined using the available evidence.  Consequently, differences in the percentage of deaths attributed to prematurity and HIV/AIDS as the underlying CoD in our study, compared with the national modeled estimates, could be due to biased case enrollment in our study. Alternately, it could be reflective of inaccuracies in the current modeling approaches of CoD attribution. Interestingly, however, the proportion of deaths attributed to diarrhea and acute respiratory infections as an underlying cause in our study mirrored those of the national estimates [1,15]. Also, the percentage of deaths attributed to meningitis (1.6%) as an underlying cause in our study was similar to national estimates (1.5%), whereas the respective figure for sepsis (2.4% in our study) as an underlying cause was higher than the national estimate (0.2%) [15].
Despite the above study limitation, our findings provide detailed insight into the specific causes of death, which would otherwise mainly have been analyzed at the syndromic level. This included the dominant role of infections as either the immediate or underlying CoD in children (78%), with specifically identified preventable or treatable organisms contributing to 94% of HIV-related deaths and 90% of pneumonia deaths. The role of infections reflected here excludes those contributing to death as antecedent CoD, and their overall contribution is likely to be even greater when antecedent morbid conditions are considered. Also, notably, only a single death was attributed to severe malnutrition as the underlying cause, although 62% of the children who died were categorized as malnourished.
In our setting where PCV was introduced into the public immunization program in 2009, RSV was the commonest pathogen implicated in respiratory deaths, albeit a limited number of pneumonia cases. This included deaths occurring in prematurely born infants, and children with underlying neoplasms and congenital malformations. Without postmortem investigation of these cases, the role of RSV and other pathogens implicated in pneumonia-associated childhood deaths in our study would be underappreciated.
Significant progress has occurred in reducing mother-to-child HIV transmission in South Africa, as well as providing antiretroviral treatment to all HIV-infected children upon their diagnosis [16,17]. Nonetheless, we observed pneumonia from P. jirovecii (n = 6) and CMV (n = 5) as important causes of death in these children, similar to observations before antiretroviral treatment was standard of care in settings such as ours [18]. This observation suggests deficiencies in the current HIV treatment program in South Africa, where death from preventable causes such as P. jirovecii pneumonia is still prevalent in HIV-infected children.
The findings indicate the utility of MITS in prompting a review of recommendations and practices to avoid these preventable deaths.
Analysis of the immediate CoD in our study also highlights the contribution of sepsis, which would largely be missed if focusing only on the underlying CoD [19]. Considering that death from sepsis is treatable, knowledge of the commonly implicated pathogens as revealed by MITS could inform empiric antibiotic therapy and sensitize physicians to their important role in childhood death, including the role of hospital-acquired infections. In developing countries, Klebsiella species and S. aureus have been attributed as important pathogens of hospital-acquired sepsis in infants [20][21][22]. Our study identified multidrug-resistant (data not shown) A. baumannii as the dominant (47.1%) pathogen causing hospital-acquired sepsis. The emerging dominance of multidrug-resistant Klebsiella species and A. baumannii infections necessitates a review of empiric treatment of hospital-acquired infections in settings such as ours [23][24][25][26][27].
Congenital malformations (22.1%) were the leading underlying CoD among children 1-59 months of age (group II); the proportion is 11-fold higher than national estimates (2%) [15]. The difference might reflect biases in enrollment in our study or, conversely, underascertainment of underlying congenital abnormalities in current CoD modeling exercises. Nevertheless, MITS provided further insight that the majority of deaths in these children were also from treatable or preventable infections (73.9%), with only 21.7% of the deaths in this group attributed directly to the underlying malformation.
Although MITS provided insight into the causal pathway of death in most cases, it might be of limited utility in attributing the underlying CoD in noncommunicable disease such as occult congenital malformations, which we mainly diagnosed based on antemortem clinical information. This highlights the need for a holistic approach in using all available information for fully characterizing the CoD in children. Another limitation of MITS, which systematically samples predefined anatomical regions, is that focal abnormalities could be missed. Also, although we used molecular assays for identifying some organisms, the interpretation thereof could be controversial in the absence of validating the significance of their presence. This was partly addressed in our study by interpreting the finding from the molecular and microbial culture tests, together with the clinical and histopathological results.
In conclusion, our pilot demonstrated that almost twothirds of screened parents approached for performing MITS on their deceased children consented to study participation. Furthermore, despite the study limitations, including the generalizability of the study findings, we demonstrate the utility of MITS in contributing to a granular understanding of the causal and diarrhea) includes other circulatory, hepatic failure, pulmonary tuberculosis, encephalitis, aspiration pneumonia, gastrointestinal hemorrhage, poisoning, hepatitis, pulmonary mucormycosis, hyperosmolality, intracranial abscess, status epilepticus, pulmonary embolism, acute respiratory distress syndrome, acute interstitial pneumonitis, kidney failure, necrotizing enterocolitis, tetralogy of Fallot, hemorrhage from respiratory passage, sudden infant death syndrome, asphyxiation, drowning, and complications during surgery. Abbreviations: acq, acquired; HIV, human immunodeficiency virus.  pathway of death in children. Future studies, such as CHAMPS [13], which are designed to be more generalizable to the studied populations, would help inform the prioritization of interventions and research that is required toward achieving the UN Sustainable Development Goal 3.2 target of reducing under-5 childhood death rates (per 1000 live births) from 39.1 in 2017 to 25 by 2030.

Supplementary Data
Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes
Diagnosis and Pathogen Overall (N = 127) <12 mo (n = 67) 12-59 mo (n = 37) ≥60 mo (n = 23) f The total is less than the column sum because of the following coinfections: S. pneumoniae meningitis and N. meningitidis meningitis. g The total is less than the column sum because of 1 case with A. baumannii and C. albicans meningitis. h Although the 2 cases of disseminated CMV were community-acquired, they had coinfections with nosocomial A. baumannii sepsis and nosocomial K. pneumoniae sepsis.