The aim of the study is to examine if a pragmatic, evidenced based and generalisable intrapartum care bundle involving birth companions and empowering mothers will reduce birth injury-related epilepsy at 18 months of age in India.

The care bundle will have four key elements (interventions): (1) birth companion providing constant 1:1 care during labour and early perinatal period; (2) fetal surveillance during active labour by a nurse or midwife using a graphic display Doppler; (3) labour management by an electronic partogram with an 'alert' and 'nag' feature based on the current WHO guidelines; (4) brain oriented early newborn care with resuscitation where indicated.

The care bundle will be evaluated using a prospective interrupted time series design, recruiting 80,000 women delivering in one of the three participating centres in south India, over two years. Accurate baseline data will be collected during the first year and the optimised care bundle will be introduced during the second year. All full term newborn infants admitted to the neonatal unit with perinatal brain injury during both periods, will have detailed assessments including video electroencephalography, and magnetic resonance imaging, and will be followed up until 18 months of age.

Primary outcome is the number of infants with epilepsy (categorised per current ILAE guidelines) at 18 months of age expressed as per 1000 term livebirths. The investigators will use a segmented logistic regression to divide the time series into pre- and post-intervention segments, with the intervention date as the intersection between segments. The difference in the two segments will be quantified using the level (step change) and slope (trend change). The total duration of the study is four years including 24 months of recruitment and 18 months of follow-up.

Epilepsy is a condition where individuals are prone to recurrent epileptic seizures; which means a change in the electrical activity of the brain resulting in a change in behaviour or movement. Epilepsy is a symptom of the condition of which there are many different causes, including brain injury occurring around the time of birth.

Worldwide approximately 50-70 million people have epilepsy, and 4.6 million develop epilepsy each year. The incidence of epilepsy in low and middle-income countries (LMICs) (1.3 per 1000 people) is 2 to 3 times higher than that in high-income countries (0.49 per 1000 people). The Prevention Task Force of the International League Against Epilepsy [1] estimated that perinatal brain insults accounted for the largest attributable fraction of paediatric and adult epilepsy in LMICs with median (95% confidence intervals) estimated fractions of 17.4% (14.7 to 18.9) and 11.4% (7.8 to 15.4). The contribution (population attributable risk) of perinatal brain insults towards preventable epilepsy is 3 times higher than that of central nervous system infections with parasites, virus and bacteria (5.3%), and traumatic brain injury (6.6%). The task force concluded that health programmes to improve prenatal and intrapartum care in LMIC may prevent a substantial proportion of epilepsy in LMICs[1].

12 million people were estimated to be living with epilepsy in India in 2015, accounting for 1/6th of the global epilepsy burden [2]. The annual economic burden of epilepsy to the Indian economy was estimated to be £1.3 billion ($1.74 USD) with a cost of £260 ($344 USD) per patient per year in 2001[3], based on the burden of 5 million people with epilepsy in India at that time. Hence the current costs, based on 12 million people with epilepsy in 2015 [2], may be much higher. Approximately 500,000 new epilepsy cases occur in India every year[3], of which 87,000 (17.4%) are likely to be related to a birth related brain injury. The vast majority of these cases will have additional neurodisabilities including cerebral palsy, deafness and blindness. Hence, the social and economic burden of epilepsy related to perinatal brain injury is likely to be much higher than isolated epilepsy.

Hypothesis. Epilepsy from perinatal brain injury can be reduced by the use of a pragmatic, evidenced based and generalisable intrapartum care bundle that involves birth companions, intelligent fetal heart rate monitoring, an e-partogram and brain oriented neonatal resuscitation, in Indian public sector hospitals.

5. Aims and objectives

Primary aim

1. To examine if epilepsy caused by birth related brain injury in Indian public sector hospitals can be prevented by pragmatic care bundle for improving the intra-partum care.

Secondary aims

1. To examine if the intra-partum care bundle reduces the incidence of neonatal encephalopathy.

2. To examine if the intra-partum care bundle reduces the incidence of neonatal seizures.

3. To examine if the intra-partum care bundle reduces caesarean rates.

4. To examine the acceptability of the care bundle by families and frontline labour room staff.

5. To examine the cost-benefits of the intrapartum care bundle

6. Methods The study will be conducted over a four-year period. Baseline data will be collected during year 1, and the care bundle will be introduced during year 2. The primary outcome will be assessed at 18 months of age.

   Element I: Constant birth companion providing 1:1 care during labour and early perinatal period

   Following necessary training, the companion will contribute towards maternal empowerment through support described below during the entire first and second stage:

   (i) Hydration/nutrition: provide drinking water and light refreshments to women as desired, during labour.

   (ii) Mobilisation/massage/reassurance: encourage and assist women to be mobile during labour, massage back and body as required, and provide constant reassurance.

   (iii) Prompt healthcare staff to perform regular observations as described under fetal surveillance.

   (iv) Prompt staff to adhere to WHO's infection and control guidelines and hand hygiene.

   (v) Encouragement during 2nd stage: support women to push at appropriate times. (vi) Assist in immediate skin to skin care after delivery and early breast feeding.

   Element 2: Fetal surveillance during active labour by a nurse or midwife (i) Clinical assessment of the woman with appropriate risk categorisation (ii) Intelligent auscultation on admission using graphic display Doppler, and subsequent fetal heart recording every 30-45 minutes in 1st stage, and every 5-10 minutes in 2nd stage, and after each contraction.

   (iii) Escalation to medical staff when appropriate

   Element 3. Labour management by an electronic medical record with an 'alert' and 'nag' feature (e-partogram) (i) Prompt, regular observation of maternal, and fetal wellbeing and progress of labour with the 'alert feature' and 'nag' feature of the software on a Tablet computer.

   (ii) Rationalise induction and augmentation i.e. artificial rupture of membranes and use of oxytocin only with infusion pumps.

   Element 4. Brain oriented early newborn care with resuscitation where indicated (i) Early skin to skin care after birth, promoting early feeding (ii) Newborn resuscitation (NLS) by nurses/midwives and doctors as indicated (iii) Prevention of hyperthermia in babies with suspected brain injury

   Although, the primary focus of the care bundle is prevention, rather than treatment of perinatal brain injury, inadequate resuscitation and hyperthermia can worsen the brain injury, and hence may increase the risk of epilepsy. All recruiting centres in the study currently have good neonatal resuscitation facilities managed by neonatal doctors, and this will be maintained in both phases of the study. In most cases, the inadvertent hyperthermia can be prevented by not using radiant warmers, or by securely attaching the servo control temperature probes over the abdomen when such warmers are used. This element of the care bundle will be further modified based on the baseline data on the exact incidence of neonatal meningitis and isolated symptomatic hypoglycaemia with occipital brain injury in these settings.

   Assessment of Neonatal brain injury The following data will be collected on all babies admitted to the neonatal unit or presenting with brain injury (e.g. seizures, encephalopathy, meningitis, intracranial bleeds) before 72 hours of age, following informed parental consent.

   1. Clinical examination including accurate staging of neonatal encephalopathy using the NICHD Neonatal Research Network modified Sarnat staging. The NICHD examination will be performed between 1 to 6 hours age and repeated at 24 hours and 72 hours of age. The babies will be classified as having mild, moderate or severe hypoxic ischemic encephalopathy based on the neurological examination between 1 to 6 hours of birth, if in addition these babies have evidence of intra partum hypoxia as defined by - an acute perinatal event (e.g. cord prolapse, cord rupture, uterine rupture, maternal trauma, haemorrhage, or acute cardiorespiratory arrest) and either a 10-minute Apgar score of 5 or less at 5 minutes or assisted ventilation initiated at birth and continued for at least 5 minutes. In cases where a blood gas is available, a pH of 7.0 or less or a base deficit of 16 mmol/L or more in a sample of umbilical cord blood or any blood during the first hour after birth will be used an evidence of an acute intrapartum event [27].

   2. Infection screening including automated blood cultures and cerebrospinal fluid examination in all babies with suspected early onset (age <72 hours) sepsis. Part of this blood will be collected and stored for transcriptomic analysis of infection and encephalopathy.

   3. Twelve lead video EEG and amplitude integrated EEG for at least 4 hours using a dedicated research EEG monitor in all babies with encephalopathy or suspected seizures. Approximately 2 to 3 EEG will be performed daily at each centre. The EEG will be uploaded on to a secure cloud-based server for central reporting, while the aEEG will be reported locally in real time for clinical decision making.

      Many neonatal seizures manifest with subtle clinical signs or may remain entirely subclinical despite the presence of clear electrographic seizure activity on EEG[28]. The investigators will classify neonatal seizures as per the levels of diagnostic certainty proposed by the recent Brighton Collaboration Neonatal Seizures Working Group led by Dr Ronit Pressler[29]. A neonatal seizure will be defined as a transient electrographic change in the brain due to an abnormal, excessive or synchronous neuronal activity either with the occurrence of clinical signs (electro-clinical) or without them (electrographic-only).

      Level 1 (Definite seizures) - Seizures confirmed on conventional EEG with (electro-clinical seizure) or without (electro-graphic only) clinical manifestations.

      Level 2a (Probable seizures) - Seizures confirmed on aEEG with (electro-clinical seizure) or without clinical manifestations (electro-graphic only) Level 2b (Probable seizures) - Clinically assessed focal clonic or focal tonic seizure directly witnessed or reviewed on video by experienced medical personnel when EEG or aEEG was not available.

      Level 3: (Possible seizure) - Clinical events suggestive of epileptic seizures other than focal clonic or focal tonic seizures, directly witnessed or reviewed on video by experienced medical personnel Level 4: (Not seizure) - Reported seizure event (as previously defined) but insufficient evidence to meet the case definition Level 5: (Not seizure) - Reported seizure event (as previously defined), documented or witnessed by experienced medical personnel and evaluated by simultaneous conventional EEG or aEEG and determined NOT to be a case of neonatal seizure.

      The investigators will use an optimised neonatal EEG with predefined positions for easy application of the electrodes by research nurses. These leads will provide both a 12 montage EEG and aEEG, which will be uploaded on the cloud server for reporting (Figure 1B and C).

   4. 3 Tesla magnetic resonance imaging, diffusion tensor imaging and single voxel thalamic proton spectroscopy using the HELIX trial sequences. Approximately 3 to 4 neonatal MR scans per week will be performed at each centre, during year 1 and 2. In addition, repeat MR scans will be performed at around 18 months of age in all children with epilepsy. Thus, approximately 3 to 5 repeat MRI scans per month will be performed during years 3 and 4, at each centre. The anonymised MR data will be uploaded on to a central server at Imperial College. The conventional MR images will be reported locally in real time, while DTI and MRS will be analysed centrally at Imperial College London following the study completion.

   Neurodevelopmental follow up and Epilepsy

   1. All babies admitted to the neonatal unit with a birth related brain injury within 72 hours of birth will be followed up until 18 months. This will include all babies with HIE, seizures, proven early onset neonatal meningitis, symptomatic hypoglycaemia associated with brain injury, or any other birth related injury on the MRI scan (e.g. intracranial bleed, arterial stroke) associated with increased risk of epilepsy. The same criteria for follow up will be strictly followed during both phases of the study to avoid a selection bias in the follow up.
   2. The research nurses will provide information and training (including video demonstration of seizures) to parents on recognising seizures at home and will advise on the subsequent actions to be taken. A 24/7 mobile phone contact number will be provided to the parents for reporting any seizures. In addition, the research nurses will maintain a monthly telephone/WhatsApp message contact with the parents after discharge for a clinical update. If seizures are present, information on the seizure type, exact age at onset of seizures, frequency and medication use will be collected. The research nurses will also encourage parents to video record the seizure episodes on mobile phones, whenever possible.
   3. A dedicated monthly multi-disciplinary research clinic will be established at each of the three recruiting centres at Chennai, Bangalore and Calicut. The clinic will be jointly conducted by a neonatologist, paediatric neurologist, speech and language therapist and physiotherapist, and care pathway will be provided to each infant based on a detailed neurological assessment including Prechtl's at 3 months. Depending on these babies will have 3 to 7 further clinic appointment until 18 months of age. The speech and language and the physiotherapy teams at the recruiting sites will be trained and supervised by the study occupational therapy consultant.
   4. Any infant developing seizure will be asked to attend the next monthly multidisciplinary research clinic for further evaluation (please see more details under the evaluation phase). It is very likely and expected that the acute management of the seizure would be at the nearest health care facility, either in a private or public sector, rather than the recruiting centre. These data will be captured by the research nurses. The clinical data will be entered into the study database using an electronic case report form. The diagnosis of epilepsy will be based on the standard ILAE criteria of two or more unprovoked seizures 24 hours apart beyond the neonatal period[62]. A repeat video EEG may be performed any time in selected cases, if clinically required or if there is any diagnostic uncertainty.
   5. At 18 months of age, all children with brain injury (with or without epilepsy) will have a detailed neurological examination for cerebral palsy, hearing and vision assessment and a Bayley Scales of Infant developmental examination (Version III). The children with epilepsy may have a repeat MRI scan under light oral sedation, full montage sleep and awake video EEG if clinically appropriate. All neurodevelopmental assessments will be performed by a Bayley III certified neurodevelopmental paediatrician or occupational/physiotherapist.

Inclusion Criteria:

• All pregnant women delivering at 36 weeks or later

Exclusion Criteria:

• Premature deliveries

Magnetic resonance biomarkers will be quantified masked to the interventional group