Trial design
This is a 3-armed community-based randomized controlled, open-label but assessor blinded superiority trial with treatment allocation ratio of 1:1:1. Multi-arm trial is considered to understand the incremental impact of BEP to the mother, along with Azithromycin to the infants, on length velocity.
Study setting and study population
The trial is being conducted in the peri-urban communities of Karachi, Pakistan. These are impoverished coastal slums with a population of approximately 250,000 residents based on the census conducted in 2017. The annual birth cohort is around 5,000 each year. The population is multi-ethnic such as Sindhi, Pashtun, Punjabi, Bengali, and Urdu-speaking. Through a demographic surveillance system (DSS), bi-monthly visits are made to married women of reproductive age (13- 49 years) in the catchment area. During these visits, mortality, pregnancy, in and outmigration and the current number of under 5 children are recorded. From previous studies, there is a high prevalence of low birth weight (30%), stunting (52%) and wasting (18%) in the study area. The study population is all lactating women of reproductive age who have recently delivered and their infants.
Participant eligibility criteria
The lactating women between 13 to 49 years of age and their newborn are enrolled, if they fulfill the inclusion and exclusion criteria provided in Table 1. Routine measurement of mid-upper arm circumference (MUAC) is done during surveillance rounds and women with MUAC of less than 23 cm in the first week of delivery are screened for eligibility by a research team. If eligible, the written informed consent is administered in the local language.
Sample size
There is limited data available on the impact of BEP on length velocity over the period of the first 6 months for an infant. In one of the studies, the mother who received only perinatal supplement showed that the overall increase in length (cm/month) among the LNS group was b=3.289, compared to b=3.289 among the MMN group. (18) Women in this trial received less dose/energy of supplement, compared to what we are proposing, also, their intervention time is different i.e. perinatal compared to infancy. Another study showed that difference in length velocity (cm/month) was on 0.02 among the MMN group compared to IFA over the period of 0-18 months. (19) Therefore, in the absence of clear evidence on the impact of these interventions, we hypothesized the effect size of 0.12cm/month to see the impact. This is also based on learning through field experience, where when provided severely malnourished lactating women (MUAC > 19.0 CM) with chickpea-based ready-to-use supplements (100 gm/day for 3 months), they showed a difference of 0.06 cm/month at 3 months of age compared to women who did not receive any supplement(data unpublished and purely from implementation, not a study).
The null hypotheses for this trial is that an infant of a lactating woman receiving standard breastfeeding counseling with BEP alone (Arm B) for 6 months or in combination with a single prophylactic dose of Azithromycin at 42 days of age (Arm C) has a mean difference in length velocity of less than 0.12/month (primary outcome) compared to standard breastfeeding counseling alone (Arm A). Learning through local experience from our field sites and due to absence of evidence of such intervention on length velocity, we used our field data. The alternate hypothesis for this trial is that an infant of a lactating woman receiving BEP alone for 6 months or in combination with a single prophylactic dose of Azithromycin at 42 days of age has a mean length velocity of greater than 0.12 cm/month or more compared to no intervention. The sample size takes multiple comparisons into account and is based on the primary outcome of length velocity with an effect size of 0.12 cm per month at least between the arm with a 1-sided test, 0.025 alpha to account for multiple comparisons (the lower alpha). A drop out of 14% (i.e. 10 % loss to follow up and 4% infants’ deaths) is assumed in the study so a minimum total sample size required is 957 (319 LWs in each arm).
Recruitment and informed consent procedure
Leverage on existing married women surveillance system established by Department of Pediatrics and Child Health, Aga University, VITAL has access to line listing of all pregnant women at study sites. Using these listings, the research team visits those households and built a rapport with pregnant women and leave our contact number to be used to notify birth events. During each pregnancy touchpoint, research team provides standard antenatal and nutrition counseling to each pregnant woman and guides them for seeking proper care. At the time of birth, the randomization/enrollment team receives a notification of birth and team visits these households for eligibility assessment. If eligible, a written informed consent is administered by the same team in a local language (mostly Urdu and where required in Sindhi and Pashto language). Team members explain details of the trial, follow up procedures, specimen collections, and other related processes. After a voluntary agreement, consent is provided by the participant.
Randomization and allocation concealment
After written informed consent procedure, randomization is performed by the team. Stratified block randomization with varying sized blocks of 3, 6, and 9 is used. Sequence generation is done by an independent statistician using a random selection method before the beginning of the trial. Self-adhesive, pre-coded sticking labels with the unique identification numbers are applied to sealed opaque envelopes containing coded information of randomization identification number and intervention to ensure that the randomization process and allocation is blinded. Baseline information is completed with counseling regarding nutrition and exclusive breastfeeding. Anthropometry of both mother and newborn is done and follow-up procedures is explained. Figure 1 shows the trial flow in detail.
Blinding
Outcome assessors are blinded, restricted only to anthropometry with a designated non-overlapping schedule with follow-up teams. All investigators are also blinded to group allocation throughout the period of the study.
Interventions
In the control arm, lactating women receives standard nutritional counseling and promotional messages of exclusive breastfeeding, a standard-of-care. In the first intervention arm, in addition to above, lactating women receives 2 sachets of BEP supplementation per day until the infant reaches 6 months of age. BEP is a certified product of the World Food Program and is locally produced by Ismail Industries, Karachi. The manufacturers do not and will not have a role in any part of the study. Each sachet contains a caloric value of 400 kilocalories per 75 grams and protein of around 10.5 grams. The source of protein is mainly chickpea, peanuts, lentils, legumes, and skimmed milk. In the second intervention arm lactating women receive standard nutritional counseling and promotional messages of exclusive breastfeeding along with BEP supplement (same as in first intervention arm) but in addition, their infants also receive a single prophylactic dose of Azithromycin suspension 20 mg/kg at day 42 of life. All arms receive routine care including newborn care, immunization, counseling regarding newborn and infant care at home and timely referral to health facility in case of urgent need.
Data collection and data management
Case report forms (CRFs) are designed to capture details on screening, eligibility, randomization, household demography, newborn assessment, danger signs, serious adverse events, compliance with intervention, exclusive breastfeeding, 24-hours food recall (to estimate usual intake as well as diversity on a monthly basis) and anthropometry. The data is collected on tablets with inbuilt logical check and skip patterns by trained team and updated on secure servers in real-time using digital application, which is design and built in-house. Auto alerts reminds about the schedule of follow up and data tracking is done on key indicators. All the data is collected in a real-time manner and uploaded on cloud server which is password protected and only accessible to the trial data management team and manager. Participants confidentiality is highly maintained through unique ID system and participants identification is not exposed to anyone outside the trial team. Tablets which are used, are password protected and only accessible to study team. Deidentified data will be used for analysis purpose.
Follow ups
Follow up teams perform home visits to provide counselling in all arms. BEP is also provided (intervention arm 1) and compliance measured through logging of number of empty sachets since last visit. Azithromycin 20mg/kg stat dose (intervention arm 2) is given to infants at day 42 of life. Assessment of compliance of exclusive breast feeding is done in all arms since the last visit. Follow-up is done daily for first 15 days, then at alternate days for 2 weeks, next every 72 hours for two weeks and finally weekly until child reaches the age of 6 months. Monthly 24-hour food recall data is also collected. In case where participant shifts out of the catchment area, team has developed a system to follow them at new location, when possible. There is no plan of retention of these participants once their 6 months of follow-ups are completed. However, through our existing free-of-cost primary health care facilities, the standard of care is available to all participants, even after completion of trial. After enrollment (within 168 hours of birth) total follow-up duration of each participant in the trial is 6 months. Figure 2 Time schedule of enrollment, interventions, and assessments.
Growth outcome assessment
Teams are trained for anthropometry of infant and mother by a master trainer using INTERGROWTH-21st standards with monthly refreshers. The measurements include infant length, weight, MUAC and head circumference and maternal MUAC and weight. Apart from primary outcome, which is length velocity of infant at six months compared to baseline, other secondary outcome includes: weight velocity, length-for-age z-score, weight-for-age z-score and weight-for-length z-score at six months. Outcome assessment teams will perform anthropometry every 4 weeks starting from day 27 of infant life until 6 months. Two readers measure the infant and mother blinded from one another, data is entered in digital form and system calculates the average reading automatically. For infant weight we are using Laica weighing scale model PS3001, whereas SECA adult weighing scale model 874 is used to assess maternal weight, we imported MUAC tape from UNICEF. SECA scale model 417 and SECA scale model 213 are used to measure infant length and maternal height respectively. The allowable difference between the two measurers according to study standard procedure for maternal MUAC is ± 0.5 cm, maternal weight is ±0.2Kg, maternal height is ±0.5cm, infant weight is ±20 gm, and ±0.4 cm for infant length, infant MUAC and infant head circumference.
Biomarkers assessment
On day 40-42 and 56 of life, infant blood will be collected from all comers where parents/caregiver are agreed for Hemoglobin, Ferritin, Transferrin, C-reactive protein (CRP), alpha1-acid glycoprotein (AGP). Rationale for collecting Hemoglobin, Ferritin and transferrin is to assess if there is any difference in makers for iron deficiency anemia among infant in each arm. Hemoglobin assessment is being done at Hemacue machine; while Ferritin and Transferrin will be performed using immune-turbidimetric assay on Roche Cobas c-311 automated clinical chemistry analyzer. AGP and CRP will be analyzed using the same assay and equipment to see the difference in these inflammatory biomarkers in three arms. In order to complement and link the findings of infant’s biomarkers with maternal intervention, the 50 lactating women from each are also approached at same time point to collect the blood specimens for same biomarkers to complement the mother-infant dyad. These assays will be performed at Nutrition Research Laboratory (NRL) at Aga Khan University. Using the same subsample dyad, we are also planning to perform plasma proteomics to assess the element of antibiotic resistance.
Further, breast milk of same lactating women will be collected to assess quality of breast milk composition (macro- and micro-nutrients), Human Milk Oligosaccharides (HMO), immunoglobulins and microbiome analysis. This analysis will be done at Azad Lab, University of Manitoba. We will develop a Material Transfer Agreement (MTA) with the University of Manitoba shipment of the specimens.
Among the same women and their infant i.e. 50 per arm, we will also collect stool specimens at same time points. We will assess inflammatory biomarkers in stool like Calprotectin, Lipokalin-2, and Myeloperoxidase (MPO) using Elisa assay. Further, analysis of stool samples will also include detection of enteropathogens using TaqMan Array Card (TAC) system for Polymerase chain reaction (PCR) at IRDL. Moreover, we will be performing targeted Bifidobacterium Identification using real time PCR. We will also send stool samples to University of Stanford after signing MTA, for further metagenomic work.
Quality assurance
10% of data is checked by trial supervisors and associate. All research team is certified in Good Clinical Practice. Comprehensive training and refreshers are routinely conducted. Teams involved in anthropometry are standardized by the WHO trained master trainers. Aga Khan University provides full technical support and overall monitoring of the trial.
Participant safety
Close follow-up is done to ensure participant safety, and both the lactating women and infants are referred if needed to physician at the primary health care clinic, with facilitated referral to tertiary hospital if required. An independent Data Safety Monitoring Board (DSMB) monitors the safety of trial participants and provides trial oversight. Monthly reports are shared with the DSMB on serious adverse events with the authority that if a safety signal is observed, the DSMB may stop this trial prior to full recruitment. The safety net including facilitated referral and reporting will minimize the chances of harm to participants. The advisory group at the Aga Khan University is overseeing the recruitment rate and over all compliance with the standard of Good Clinical Practices.
Possible risks
There may be diarrhea, nausea, vomiting, skin rash, and abdominal distension after use of BEP. Similarly, infants may develop diarrhea, nausea, vomiting, skin rash, and abdominal distension post Azithromycin dose. The risk management includes prevention through rigorous follow-ups, continuous monitoring of danger signs and documentation and reporting of side effects to the ERC and DSMB. Every illness or danger sign reported or identified, is addressed through facilitated referral for both women and infants. Where necessary, intervention is discontinued temporary or permanently, after consultation with the technical team of the trial.
Data Safety and Monitoring Board
Independent group of experts, comprises of 5 members, formulated a Data Safety and Monitoring Board for the trial responsible for monitoring safety indicators and result of interim analysis. Interim analysis is scheduled when 50% of the enrollments complete their six months of follow-ups. Data of Serious Adverse Events is shared with the board on a monthly basis in the form of progress report.
Statistical analysis
This will be done on Stata version 15. Distribution of baseline characteristics across the arms will be done. The primary analysis will be intention-to-treat (ITT). We will compare mean length velocity (cm/month) (primary outcome), weight velocity (gm/kg/day), LAZ, WLZ, and WAZ in the two intervention arms with the control arm using repeated measures ANOVA adjusted for birth weight and age of infant at enrolment. If the outcome is missing for the intention to treat infants, means from that group will be imputed. Subgroup analysis will be done by maternal BMI and MUAC categories, and compliance with the intervention.
Participants and public involvement
Investigators has a very extensive experience of working with community and their representatives /elders. During protocol development phase, team discussed and took feedback on the research question and trial design from the community representatives. Furthermore, community perspective about the trial procedures, specially frequency and duration of follow-ups and biospecimens collection was also taken. Also, during the pilot phase the aim was not only to test the consent and questionnaires, but also to understand how community responds to different questions, and how sensitive information around antenatal and postnatal period can be collected in a more receptive and profound manner.