• Afghanistan
• asphyxia
• cranial ultrasonography
• IVH
• low-income countries
• newborn
• NICU
• palliative care

Introduction: Afghanistan is one of the poorest countries in the world. Neonatal mortality rate is 35:1000 live births, more than 15 times that of Europe. In 2015 the main causes of neonatal death were prematurity (31,2%), birth asphyxia and birth trauma (26,8%) and sepsis (18,3%). Emergency Maternity Hospital of Anabah was opened in 2003 and has improved through the years aiming to offer and define the gold standard of a comprehensive cost-effective neonatal care. Cranial ultrasonography (CUS) is the preferred initial method for imaging the neonatal brain, and there are no studies in literature on the implementation of a point-of-care ultrasonography (POCUS) program in afghan NICUs. In absence of access to brain MRI, with a cranial POCUS it is possible to evaluate the presence of the most common pathological findings in preterm newborns, newborns with perinatal asphyxia or traumatic birth.
Aim of the study: This study seeks to better predict the outcome of high-risk babies born in Emergency hospital to improve their management. More specifically, the aim is to understand which are the major risk factors for pathological neurological outcomes in the facility, to predict pathological neurological outcomes, to evaluate the impact of a CUS screening, to help in the decision of undertaking a palliative care path. Secondary aims were to compare the incidence of diseases and pathological findings in the hospital, like perinatal asphyxia and intraventricular haemorrhage (IVH), with international epidemiology, to understand where there is space for improvement.
Methods: The study was done on the population of newborns born in Emergency Maternity Hospital, Afghanistan, between January and May 2024. The newborns included in the study were the ones with a higher risk of having/developing a pathological CUS: all newborns under 1500 g of birth weight (BW), all newborns with higher risk of birth trauma or distress because of vacuum delivery, breech extraction or caesarean section for fetal indication, all newborns included in the perinatal asphyxia cohort. At least one CUS was performed on all babies included in the study. A CUS was performed in the first day of life then, if pathological or in babies considered more at risk of complications, it was repeated at 3, 7, 14, 21, 28 days of life and then every other week until discharge. The examination was defined as pathological if there were at least 1 among these findings: IVH, cerebral edema, pathological RI, basal ganglia and thalami injury, white matter injury, congenital malformations, a few particular cases. Also a blood gas analysis (BG) at birth was performed to understand if it could improve the initial diagnostic assessment and help predicting the outcome. Both Sarnat and Thompson scores were used to give a neurological scoring. For the follow-up of the babies an “Outcome questionnaire” was created; the nurses asked the questions through a phone call made between 3 and 7 months of life of the baby.
Results and Discussion: During the period of the study 2805 newborns were born in Anabah hospital; 715 were admitted in neonatology ward because they fitted the admission criteria. 113 of the admitted patients were included in the study; of these 32 died (28,3%).
53 patients underwent BG within 1 hour from birth (48%). Other 12 patients had a BG done in the first day of life. 30 BGs were defined as indicative of perinatal asphyxia. All of these patients had an Apgar ≤5. Since none of the patients with a pathological BG had an Apgar >5, BG does not give supplementary information for the management of these patients, hence it should not be implemented in this setting. 38 neonates had a pathological US; here is a description of pathological CUS findings: 10 neonates had an IVH, of which 2 also had a PHI, 8 had cerebral edema, 1 had diffuse post-anoxic damage, 2 had a significant ventriculomegaly of which one evolved in a tetraventricular hydrocephalus, 1 had third ventricle dilatation, 1 had caudate nucleus damage, 1 had persistent periventricular hyperechogenicity, 1 had a transient small occipital hyperechogenicity, 17 had pathological RI at 24 hours, 6 had pathological RI at 72 h, 1 had bilateral subdural haemorrhage, 1 had a subgaleal haemorrhage. Other CUS findings considered as physiological: 6 had periventricular hyperechogenicity, 11 had hyperechogenicity of the thalami due to prematurity[48], 2 had choroid plexus cyst. Of the 32 patients in the study that died for all causes, 19 had a pathological CUS (sensitivity of 60%). Other CUS performed during the period of the study but outside the population of the study showed that also a point-of-care CUS done on a patient with an acute event without prior risk factors can help deciding the future management of the patient: 5 patients that had done a CUS outside the study died, and they all had a pathological US. In these cases CUS efficiently predicted a poor outcome, helping the decision to stop the supportive therapy.
Patients were defined as “asphyxia5” if they had an Apgar ≤5 at 5 minutes; the sensitivity of CUS to predict death in this cohort was 80%, while for “asphyxia7” (Apgar ≤7 at 5 minutes) sensitivity was 69%; the ability of CUS to predict death decreases for Apgar 6 and 7. Excluding preterm newborns deaths, the mortality rate of patients with Apgar 6 and 7 at 5 minutes is 0, hence the risk of death for these patients is not higher than that of patients without asphyxia, unless they are also premature.
8 patients had cerebral edema at CUS; the main association with cerebral edema is perinatal asphyxia. 2 patients with cerebral edema did not have perinatal asphyxia but had a traumatic operative birth; in these cases, edema was more transient and it was possible to quickly discharge the patients in good conditions. Of the other 6, 4 died and 2 had a bad outcome, the finding of cerebral edema in a patient with perinatal asphyxia is a strong predictor of poor outcome (either death or a bad outcome). Nevertheless, while it is important to know if there are any pathological signs at CUS for a baby with a traumatic birth and a pathological neurological examination, the outcome of babies with mild cerebral edema caused by traumatic birth is usually benign.
10 of the patients included in the study were found to have an IVH, and also other 4 patients not included the study with a BW between 1500 and 1700 g. These findings suggest that CUS screening should be extended at least to babies <1800 g, so not to miss eventual IVHs in LBW patients.
The patients that had a pathological RI within 72 hours from birth were 22. While the sensitivity of a pathological RI to detect a poor outcome in patients with asphyxia is quite high (64%), sensitivity for patients without asphyxia is low. Hence, RI is a value that should be used in the evaluation of babies with perinatal asphyxia, while in case of pathological values in other babies other factors should be considered.
16 patients were treated as palliative care, based on clinical criteria (BW, Apgar, severe asphyxia, polymalformations) decided outside of the study: 12 died, 1 had a bad outcome at follow-up, 2 had a good outcome, 1 was a fall-out. In total 10 of the 16 palliative care babies also had a pathological US. The 2 patients with a good outcome had a normal CUS. There is a good correlation between clinical and ultrasonographic criteria to include a patient within the palliative care cohort.
The patients that were discharged and included in the follow-up were 88, of which 78 answered to the follow-up calls, so there was a 12% of fall-outs. The patients with a bad outcome (defined as with at least 1 answer among seizures, not eating everything through breastfeeding or bottle, not growing or not crying) were 30.9%. The total of patients that had a bad outcome or died was 54 (32 dead+ 22 bad outcome), of which 24 had a pathological CUS, 44%. The data show that most of ELBW (91%) babies born in the hospital will have a poor outcome, which accounts for the choice of including them in the palliative care babies.
Conclusions: In the future Emergency aims to train the staff to perform a complete POCUS of clinically unstable neonates, this study was a first step towards it. This was the first study where a follow-up was attempted in the project, we hope to continue. CUS is a very helpful and inexpensive tool in managing high-risk babies in NICU; it helps to predict poor outcome, so it helps in the decision of undertaking a palliative care path. It may be useful to include CUS in the criteria for palliative care, considering as criteria IVH grade III for newborns <1000 and very low RI or undetectable RI for all newborns with perinatal asphyxia, especially if term. A very high Thompson score could also be a criterion to include patients in palliative care treatment, but a cut-off needs to be set.