Neonatal and postneonatal tetanus at a referral hospital in Kamsar, Guinea: a retrospective audit of paediatric records (2014–2018)

Abstract Background Tetanus is a vaccine-preventable disease caused by the bacterium Clostridium tetani. In 2018, all of Guinea was considered to be at risk of the disease and the country is currently in the elimination phase. Methods A 5-y audit (1 January 2014–31 December 2018) of all admissions to the neonatal and general paediatric units of Kamsar Hospital (Western Guinea) was undertaken to identify cases of neonatal tetanus (NNT) and postneonatal tetanus (PNNT). Results There were 5670 admissions during the study period, of which 39 (0.7%) were due to tetanus (22 NNT and 17 PNNT). Among NNT patients, the bacterial entry site was the umbilical cord (n=20) or wound following circumcision (n=2). For PNNT, the entry site was surface wound (n=12), limb fracture (n=1) or could not be established (n=4). A majority of the patients (36/39, 92.3%) were born to unvaccinated mothers or those who received suboptimal vaccination during pregnancy. Overall, 21 (53.8%) children died within 7 d of admission with a higher mortality observed among neonates (16/22, 72.7%) compared with postneonates (5/17, 29.4%). Conclusions Tetanus was a rare cause of admission at Kamsar Hospital with a very high case fatality rate. The disease primarily occurred among children born to mothers who were unvaccinated/inadequately vaccinated during pregnancy.


Introduction
Tetanus is a vaccine-preventable disease caused by a neurotoxin produced by the bacterium Clostridium tetani. The global burden of tetanus has gradually declined from an estimated 800 000 to 1 million deaths in the 1980s and 1990s to 56 743 (95% confidence interval (CI): 48 199 to 80 042) in 2015. 1,2 Despite this sharp decline, tetanus remains one of the top infectious causes of under-five mortality in West Africa. 3 In particular, neonatal tetanus (NNT), defined as tetanus occurring within 28 d after birth, accounts for the majority of tetanus-related deaths. 4 Newborns from unvaccinated mothers and those born at home are at an increased risk of exposure to the spores of the bacterium due to traditional umbilical cord practices such as the application of Karité nut butter, herbs or cow dung. [5][6][7][8][9] Postneonatal tetanus (PNNT) occurring outside the neonatal period is less common and is associated with substantially lower mortality compared with NNT. 2 The maternal and neonatal tetanus elimination (MNTE) initiative of the WHO launched in 1989 aimed to reduce the incidence of maternal and neonatal tetanus by 2015 to less than one case per thousand live births at district levels. 10   also participated in global initiatives to meet the elimination target by implementing supplementary immunisation activities in high-risk regions. 11 The raft of measures adopted has led to a steady reduction in the burden of neonatal tetanus. 12,13 In 2017, the annual NNT incidence was 0.227 per 1000 live births (a total of 103 reported cases). Similarly, there has been a decline in overall incidence of tetanus from 35.59 per 100 000 people in 1990 to 3.96 per 100 000 in 2017. 14 However, these rates are still higher than neighbouring countries: the corresponding incidence per 100 000 was 1.32 in Sierra Leone, 1.54 in Cote d'Ivoire, 3.18 in Guinea-Bissau and 3.05 in Mali. 14 In 2019, the Guinean government launched three additional rounds of mass vaccination campaigns in all 38 health districts. The campaigns mobilised a workforce of 11 127 health workers and volunteers with a target to vaccinate over 3 million women of childbearing age. 9 Despite these efforts to combat maternal and neonatal tetanus (MNT), the entire country is still considered to be at risk of MNT. At least one case of NNT has been reported in each of the 38 health districts during the preceding 5 y. 9 Overall, a total of 193 tetanus cases were reported to the WHO in 2017, 326 in 2018 and 107 in 2019 ( Figure 1).
The authors (IC, MEH, MD, ASD and FCon) have treated several cases of tetanus in the past few years while working in the paediatric ward of Kamsar Hospital, a referral hospital located in the Boké region in Western Guinea, which has below-national tetanus vaccine coverage during pregnancy. 12 This work presents the results of an audit of all admissions to the neonatal and paediatric wards of the Kamsar Hospital in order to guide local interventions to eliminate the disease within the region.

Study design
This is a retrospective audit of all admissions to the neonatal and general paediatric units of Kamsar Hospital from 1 January 2014 up to and including 31 December 2018. The aim of this study was to determine the incidence of NNT and PNNT among those ad-mitted to the neonatal and general paediatric units of Kamsar Hospital.

Study location
This study was conducted at Kamsar Hospital, which is located in the Boke region of Western Guinea. The port city of Kamsar lies ∼250 km north of the capital Conakry. In 2014, according to the population census, Kamsar had a population of 133 350, of whom 32% were children aged <10 y. 15 Bauxite was discovered in the region in 1963 and this led to the development of Kamsar as a major mining city. The city has permanent access to electricity, running water and health services, including the well-equipped Kamsar Hospital. Kamsar Hospital is a major referral hospital in the region and admits patients from the entire Boké region, which has a population of just over a million. The hospital has an outpatient ward dedicated to consultation and an inpatient ward for hospitalisation along with intensive care units (ICUs). The hospital provides a dedicated paediatric service along with a general paediatric unit and a specialised neonatal unit. The paediatric ward has a capacity of 50 beds, including 12 beds for neonatal care. As of 2019, there were five paediatricians and several general physicians, as well as nurses and supporting staff providing care.

Diagnosis of tetanus and patient management
The case definitions of NNT and PNNT were based on the WHO protocol. 16 The Dakar prognostic scoring system was used to grade the severity of cases: mild (Dakar scores from 0 to 2), moderate (scores equal to 3) and severe (scores from 4 to 6). 1 On admission, metronidazole and diazepam were administered to all patients. Tetanus immunoglobulin antitoxin was administered intramuscularly, and the following drugs were administered based on the judgement of treating clinicians: phenobarbital, electrolyte supplements and penicillin G. For severe cases, intensive care support was provided and the patients were kept on a ventilator with an oxygen extractor and a nasogastric tube for hydration, feeding and administration of oral medications.

Data collection
The following patient characteristics were extracted: age, gender, body temperature, place of delivery of the newborn (home or hospital), suspected port of entry of the bacterium, incubation and onset period, complications encountered, requirement of nasogastric intubation support upon admission and the treatment regimen administered. Maternal characteristics such as education level, place of residence (rural/urban), number of antenatal care visits and vaccination against tetanus, including the number of doses received, were extracted. Information on clinical outcomes after treatment was extracted, including the length of hospital stay and mortality outcome.

Statistical analyses
The descriptive characteristics of the included patients are presented. Survival probability was calculated using the Kaplan-Meier method. Univariable and multivariable Cox regression models were used to estimate the relative hazards of mortality within 7 d of admission. The results were reported following the RECORD (Reporting of studies Conducted using Observational Routinely collected Data) statement. 17

Discussion
Our audit identified 22 neonatal and 17 postneonatal patients admitted to the paediatric unit at Kamsar Hospital from 2014 to International Health  Table 2). The reasons for the three NNT cases developing in those born to vaccinated mothers remain unclear. The quality of the vaccine, potential degradation during transportation and poor storage facilities may be possible explanations. 19 It could also possibly suggest interference with transplacental antibody transfer or could simply reflect natural variation in response as only 84% of neonates are estimated to be protected from tetanus by maternal vaccination. 4  The umbilical cord was the bacterial entry site for most (20/22) of the NNT cases. Just over a third of the neonates were born at home (information on whether the birth was attended by a trained assistant was not known) and the remaining were born at a hospital. While homebirth might be associated with an increased risk of tetanus through contamination compared with facility-based deliveries, it was not possible to elucidate the reasons for 13 neonatal cases among those with hospital birth. Among those who were born at a hospital, the median age at admission to our hospital was 7 d (IQR: 6-7 d). It is possible that traditional practices believed to facilitate rapid healing of the umbilical stump were applied to these babies after they were discharged. 20 For example, in 1996, Roisin et al. reported practices of applying Karité nut butter in Conakry, Guinea, and also remarked on the wide usage of similar products throughout West Africa. 6 Chlorhexidine has already been adopted or is being adopted in over 25 countries globally as a part of umbilical cord care. 21 It is currently not implemented in Guinea and when it is approved by the government, it will strengthen the cord care and reduce the likelihood of tetanus acquisition through the umbilical cord. There were also two cases of NNT that occurred following circumcision (both were from the rural Kamsar region). Information regarding the qualifications of the personnel who performed the circumcision was unavailable, but it is not uncommon for it to be performed by traditional practitioners, as reported in a study from the capital Conakry. 22 Reports on NNT following circumcision remain rare in the literature, and the observation of two cases in our study suggests that the current elimination activities should also consider promoting safe circumcision practices.
The overall mortality within 7 d of admission was very high (53.8%), with a greater than 6-fold risk of death among those with severe disease relative to mild/moderate cases. Although the risk of mortality was also 4-fold higher in the NNT group than in the PNNT group in the univariable analysis, the difference was no longer apparent in a multivariable model that adjusted for severity status (Table 4). This strongly suggests that the higher mortality observed is likely attributable to disease severity at presentation rather than other patient characteristics. Just over half of the patients in our study had severe disease at presentation, which indicates an urgent need for the adoption of a standardised protocol for optimal patient management of severe cases to prevent case fatality. This further suggests that despite adequate hospital care, this disease remains difficult to treat. Taken together, our study suggests a strong need for raising awareness regarding maternal immunisation, encouraging optimal antenatal care and raising awareness of hygienic birth and circumcision practices. Finally, our study also points towards the possibility of underreporting of tetanus in the country despite it being a notifiable disease. For example, there were no officially reported cases of tetanus to the WHO in 2015, despite recording 10 cases at Kamsar Hospital (Figure 1, Table 1). This discrepancy could have arisen because Kamsar Hospital officially falls within the administration of the Ministry of Mining rather than the Ministry of Health. This indicates an urgent need for the integration of databases from different health service units to successfully eliminate the disease.

Conclusions
In our setting, NNT and PNNT were rare causes of hospital admission, but when presented, the cases were often severe with a high mortality rate. The cases occurred primarily in children born at home to mothers who were either unvaccinated or inadequately vaccinated during pregnancy. Elimination efforts should focus on encouraging optimal antenatal care and raising awareness of maternal immunisation. These findings can help to design programmes to strengthen data surveillance systems and devise targeted interventions to eliminate NNT and PNNT in Guinea.

Availability of data and materials:
The datasets used for current study are available from the Paediatrics unit at Kamsar Hospital University upon reasonable request. A request can be made to the corresponding author.