Introduction Birth asphyxia is one of leading causes of neonatal mortality in Uganda. It is associated with long term neuro-developmental complications among the babies that survive. Preventive measures for birth asphyxia intrauterine are not clearly understood and thus the need for this study.

The aim of the study is to assess the effect of intrapartum oxygen administration on fetal and early neonatal outcomes.

Methods A double-blind randomized clinical trial which will be conducted in Gulu regional referral and Kawempe National referral hospitals in Uganda. A total sample size of 1108 women in labour will be enrolled with 554 participants per group. The intervention will include administration of 10 L/min of 100% oxygen for 15 minutes to women in established labor who have signs of fetal distress with fetal heart rate of less than 120 or above 160 beats per minute. The control group will receive medical air (21% oxygen) using the same criteria. Women and babies will be followed up until 7 days after birth to document the outcomes. Statistical analysis to identify difference in outcomes between the control and intervention groups will be performed.

Ethical considerations Ethical approval and permission was received from relevant research and ethics committees. Informed consent will be sought from the participants. A data and safety monitoring board will be set up to review periodically the progress of the clinical trial study. Participants will be monitored for adverse events and severe adverse events; reporting will be done according to the research and ethics committee guidelines.

Background:

Foetal distress which refers to fetal heart rate of less than 120 or more than 160 beats per minute(Pildner von Steinburg et al., 2013) during labour is one of the contributing factors to poor newborn outcomes. A proportion of these newborns will require interventions after birth such as resuscitation, oxygen therapy, antibiotics and prolonged hospital admissions. According to a retrospective study in Nigeria, 23% of foetuses diagnosed with fetal distress were born with asphyxia which is a major cause of newborn deaths(Adanikin & Awoleke, 2017). Several interventions to optimise fetal wellbeing as preparations for birth are being instituted have been suggested. These include amnio-infusion, maternal positioning, tocolysis, oxygen administration and intravenous fluid bolus(Bullens, van Runnard Heimel, van der Hout-van, & Oei, 2015; Simpson, 2007). Reviews of intrauterine resuscitation techniques have reported sufficient evidence for use of tocolysis and maternal repositioning to improve fetal pH and Apgar scores(Bullens et al., 2015; Velayudhareddy & Kirankumar, 2010). Although oxygen is used in several settings for intrauterine resuscitation(Dawood & Al-Arnous, 2002; Siriussawakul et al., 2014), its effect when used in labour for management of foetal distress remains debatable. In a Cochrane review to evaluate efficacy of oxygen administration in preventing fetal distress, only two clinical trials were found. There was evidence that the fetal arterial partial pressures of oxygen were improved by oxygen administration although there was a negative effect on pH levels. There is insufficient evidence to recommend or disprove use of oxygen for intrauterine resuscitation(Fawole & Hofmeyr, 2012). This therefore creates sufficient equipoise to undertake this study. In low resource setting where pregnant women are most likely to have anaemia, suffer from malaria and other infections, the role of oxygen for intrauterine resuscitation needs to be explored.

Aim: This study will assess the effect of administration of oxygen to women with fetal distress in the first stage of labour on fetal and early neonatal outcomes in a low resource setting.

Hypothesis: Administering 100% oxygen for 15 minutes to women with fetal distress will reduce the rate of fetal distress by 30% in the first stage of labour.

Study design: A randomised clinical trial with parallel assignment will be conducted.

Setting: The trial will take place in Gulu regional and Kawempe National referral hospitals in Uganda.

Study population: The study population will consist of pregnant women in the first stage of labour (that is between 4-10 cm cervical dilation).

Sample size: Using the Kelsey formula for sample size calculation for randomized controlled trials, with a ratio of intervention to controls of 1:1, 1108 women will be recruited.

Randomisation Block randomisation will be done to reduce the risk of bias in the study and ensure that each of the two groups has equal numbers of participant. A sequence of variable block sizes (4, 6 and 8) will be generated using STATA 15 by a statistician who will not be part of the study team.

Allocation and concealment: The allocation concealment will be done by the pharmacist in Uganda who will be in charge of storage of the control and intervention gas cylinders and dressing them with similar tamper proof jackets. The pharmacist will receive the randomization sequence from the statistician and label each masked cylinder with a set of participant serial numbers. Copies of these serial numbers will be concealed in opaque envelopes and given to research assistants.

Blinding/Masking: This will be a double blind study where the participants, the investigator, research assistants, and outcome assessors will not know the participants allocated to either intervention or control arm of the study. Unblinding of allocation will be done if the occurrence of serious adverse events increase above the study site's rates warranting investigation by the data safety monitoring board. In this case, the unblinding will be done for the data safety monitoring board (DSMB) and not to the research team.

Study procedures Women whose fetuses develop abnormal fetal heart rate on intermittent auscultation will be evaluated for presence of fetal distress by a research assistant. Diagnosis of fetal distress will be confirmed on the basis of fetal heart monitoring for 10 minutes using a Moyo fetal monitor by the trained study research assistants. A FHR below 120 or above 160 beats per minute will be considered fetal distress Standard care: All women with fetal distress will receive current standard of care which may include, encouragement for take oral fluids, administration of intravenous fluids such as Ringer's lactate, normal saline or dextrose, maternal lateral positioning and stopping the oxytocin infusion if the woman is undergoing induction or augmentation.

The intervention: Women allocated to the intervention arm will receive 100% oxygen, 10 liters per minute via non re-breather face mask for 15 minutes. The control group will receive medical air which contains 21% oxygen. 10 liters per minute via a non re-breather face mask for 15 minutes.

Monitoring and follow up: Fetal heart rate will be monitored after the intervention is commenced every 15 minutes for 1 hour or until the woman delivers the baby depending on which one comes first. After one hour of monitoring, the fetal heart rate will be monitored according to hospital protocols that is every 15 minutes if the fetal heart rate is still abnormal or every 30 minutes if it normalizes. Follow up will be done until 7 days after delivery.

Data collection Principal investigator and research assistants will collect personal, socio-demographic and pregnancy related data. Case record forms(CRFs) will be used to collect patient demographic data, maternal characteristics and results of the fetal heart rate monitoring. These forms will be filled in by the research assistants allocated to the particular shift. Data from laboratory tests will also be filled into the CRFs. A copy of results will be kept in the patient's hospital records. A study coordinator at each study site will ensure that all data for each patient has been properly captured daily and/or before the woman is discharged.

Data management Daily electronic entry will be done from the paper CRFs by the research assistants. Data will be entered into Epidata software 3.0 during the study period. This data will be checked for completeness daily and any missing data retrieved before the participants are discharged usually in 24 hours. The CRFs will be kept in a lockable cupboard placed in the labor ward with the key only accessible to the research team. The electronic data will be stored on a password protected computer accessible only by the principal investigator.

Data analysis Analysis will be according to intention to treat. Women in the control group who for one reason or another require oxygen shall not be denied the treatment but will however still be analysed according to their original assigned groups. All participants that are randomized will be analysed including those who drop out for one reason or the other. Reasons for dropout will be provided in the patient enrollment flow chart.

Data will be entered into STATA 15 software for analysis. Descriptive statistics will be computed to analyse for similarities and differences in the intervention and control group. Simple chi-squares will be used to assess differences in the study groups. The expected outcome in the two groups will be categorically analysed and proportions of participants whose fetal heart rate normalized within 30 minutes will be determined. The effect of oxygen therapy on early neonatal outcomes will also be determined using a multi-variable logistic regression model. Odds ratios will be calculated to determine associations between the outcome and predictors variables.

Quality control: Allocation concealment and blinding will be done to improve the rigor and methodological quality of the study. Fetal heart rate will be assessed objectively using the Moyo fetal heart rate monitor. Research assistants will be trained before the study commences and ongoing support supervision will be done. All procedures will follow the prescribed standard operating procedures developed in consultation will available literature. Data collected on the CRFs will be checked for completeness by the site coordinators and entered into the software regularly for cleaning. Ongoing monitoring by the study monitor and the DSMB will be done as well as visits from the ethics committee to ensure compliance with the ethical principles.

Adverse events and serious adverse events reporting Normobaric Oxygen is a category A drug for pregnant women and is therefore associated with minimal risks. Minor adverse events (AE) may include; cough, chest pain, mild dyspnea/ hypo-ventilation, malaise, nausea or tingling sensation on the skin. These signs and symptoms only occur if the oxygen is administered for more than 6 hours.

Serious adverse events (SAE) will include but not limited to, serious maternal complications such as antepartum hemorrhage, postpartum hemorrhage, disseminated intravascular coagulopathy (DIC), still birth, neonatal death, and maternal death. The principal investigator(PI) will be notified of any AE and SAE within 48 hours. In case of a SAE, it will be reported by the PI to the primary ethics committee within 7 days. According to the Uganda National council of science and technology guidelines, SAEs will be reported to the DSMB, National drug authority and the ethics committee.

Ethical considerations Ethical approval was obtained from the higher degrees research and ethics (HDREC) committee at the School of Health Sciences, Makerere University and the National Council for higher education. Approval was also received from the Uganda National drug authority and administrative clearances from both study sites. Verbal and written informed consent will be obtained from participants in the study.

Inclusion Criteria:

• Women in active first stage of labor who develop fetal distress will be included

Exclusion Criteria:

• Women in labor with critical conditions such as eclampsia, chronic heart disease, chronic lung disease and chronic renal failure
• Women in preterm labor
• Those taking Bleomycin and Amiodarone medications

De-identified data may be shared with other researchers on request.

Similar gas tanks will be used for both the intervention and control group