Study Design:
The study is an unblinded, open-labeled, non-inferiority, randomized controlled trial with a participant ratio of 1:1 in each group (the design remained the same through the length of the study and no changes were made to the methods approved in the protocol). The reason for making the study as an open labeled trial is that mothers could not feasibly be blinded from the information on allocation scheme, and the information on vaccine given to infants could not be held back from family and parents. Another non-inferiority vaccine trial in Pakistan also used a similar approach in allocation scheme Trial Profile The trial was done at a tertiary care hospital in an urban area from March 15, 2015, to May 20, 2016. Assuming an attrition rate of 25%, the study enrolled 296 expectant mothers for allocation in two groups, intervention, and control group.
Participants of the trial:
The study population was healthy newborns born to healthy females, in a health facility. These pregnant women were registered for institutional deliveries at Countess Dufferin Fund (CDF) Hospital, Hyderabad. The CDF Hospital Hyderabad is a tertiary care hospital for women and children. The hospital is located in a populous downtown area of the Hyderabad city and caters to patients coming from the urban areas of the city. The average daily out-patient visits to the hospital are 200 women. The average number of daily deliveries in the hospital is 25.
All women who attended an antenatal clinic for a routine examination by gynecologist were evaluated on following inclusion criteria;
1\. Full term pregnant and registered for delivery at CDF Hospital Hyderabad 2.1. An absence of a chronic illness (Diabetes / Tuberculosis / Hypertension) 3.1. No serological marker (HBs Ag) of prior exposure to hepatitis B Inclusion criteria for mothers and Recruitment process All women who attended an antenatal clinic for a routine examination by gynecologist were evaluated on following inclusion criteria;
1. Full term pregnant and registered for delivery at CDF Hospital Hyderabad
2. An absence of a chronic illness (Diabetes / Tuberculosis / Hypertension)
3. No serological marker (HBs Ag) of prior exposure to hepatitis B Prior to the due date for childbirth, while attending the antenatal clinic, these females were approached and informed about the study. Those who were willing to participate were given a consent form to sign. After the informed consent, blood was obtained by a finger prick and a drop of blood was put onto the Abbot's Determine ® rapid testing kit for detection of hepatitis B surface antigen. The Determine ® has sensitivity, (negative predictive value) of 98% and specificity, (positive predictive value) of 100%, compared to the Enzyme Link Immunosorbent Assay (96).
The results were interpreted as follows;
* Two red bars meant blood is positive for hepatitis B surface antigen
* Single red bar meant blood is negative for hepatitis B surface antigen Women who were negative for hepatitis B surface antigen were then allocated into an intervention or a control group (details of the methodology are given below). They were followed until they delivered and were discharged from the hospital with a healthy baby. For vaccine given at birth, other researchers have also enrolled infants as subjects of the vaccine trials(97)(98)(99). Infants have also been selected as trial participants and the hepatitis B vaccine was given to infants in trials of preventing mother to child transmission of hepatitis B(100).
Exclusion criteria for neonates:
After the delivery infants born in both groups were examined by a trained midwife for following criteria;
1. Weight, less than 2000 grams
2. Signs and Symptoms of febrile Illness/sepsis
3. Congenital Malformation
4. Signs of Cyanosis
5. Difficulty in breathing The newborns who fulfilled the inclusion criteria, i.e., weight more than 2000 grams and absence of above-mentioned criteria were followed up.
Intervention
One cc of intravenous blood was drawn from the peripheral veins of these healthy newborn. The blood was immediately centrifuged, and serum was put in a plastic container, labeled with a unique identification number. The serum containers were frozen and then sent using a cold chain container, to Pakistan Medical Research Council laboratory at Jinnah Postgraduate Medical Center Karachi. The serum was tested for anti-hepatitis B surface antigen antibody on Enzyme Immunoassay (EIA) kit DS-EIA-Anti-HBsAg, manufactured by DSI S.r.I (Italy).
After obtaining a blood sample, infants born to mothers in the intervention group were administered GSK's ® recombinant hepatitis B vaccine ENGERIX 10 micrograms (µ gm), intramuscular in the anterolateral part of either of thighs. This birth dose of hepatitis B was followed by zero dose oral polio vaccine as per hospital policy. Infants born to mother in the control group were given oral polio vaccine as per the hospital policy.
Subsequent vaccinations in both groups of infants were administered as per the expanded program on immunization schedule in Pakistan.
After discharge from the hospital, both groups of children were followed until the completion of the third dose of Diptheria - Pertussis Tetanus - Hemophilus Influenza Type B - Hepatitis B - Pneumococcal (DPT-HiB-HBV-PCV) vaccine offered under the Expanded Program on Immunization. The follow up was done with caregivers at 6, 10 and 14 weeks, follow up started at 8 weeks and duration of follow-up was up to 10 weeks after the 3rd shot of DPT-HiB-HBV-PCV vaccine. After the discharge of mothers from hospital, preformed, short text messages (SMS) were sent to parents, reminding them of the due date for as per the vaccination schedule. One day after the due date of the vaccine shot, a phone call was made to confirm the child was vaccinated or not. The parents were approached after eight weeks of the last DPT-HiB-HBV-PCV vaccine shot. They were requested to bring the child to the hospital for a follow-up blood test to evaluate the presence of anti-HBs Ag-Ab titer in the blood. Upon the follow-up visit, infants were weighed, and 1 cc venous blood was drawn from the peripheral veins. Blood was centrifuged, and serum was stored in a plastic container labeled with a unique number. All the samples were then sent in cold chain to Pakistan Medical Research Council laboratory at the Jinnah Postgraduate Medical Center Karachi.
Intervention Group:
Four (4) mothers from the intervention group were shifted to a teaching hospital, prior to delivering a baby. Baseline blood samples were taken from 144 babies and the birth dose was administered. During subsequent follow-up, cell numbers of three mothers were not responsive, after a period of three months, they were declared as lost to follow-up. Baseline antibody titer in serum of fourteen infants was found to be greater than10 mIU/ml. These children were not followed. One child died, and parents of five children refused to continue and did not allow obtaining a peripheral follow up blood sample.
Control Group:
A high attrition of study participants happened in the control group. Thirty-two expectant mothers refused to participate in the study when told that their babies would not get birth dose hepatitis B vaccination. This attrition occurred despite an explanation of trial to the control group participants, prior to the enrollment. The phenomenon of high attrition in the control group is a well-known and well-documented effect of an unblinded trial.
Margin of non-inferiority In a non-inferiority trial, if the upper limit of 95% confidence interval is less than the margin of non-inferiority i.e. delta (Δ), there is evidence of non-inferiority cases(48). At the protocol development stage, published literature was reviewed for seroprotection correlates. For the purpose of this trial, we calculated the Δ as the difference in seroprotection correlates with or without the birth dose of hepatitis B vaccine, and the Δ for this study was set to be 5%. The trial outcome set at the time of protocol design was not changed during the course of study.
Sample Size In the non-inferiority trials, where the primary response variable is dichotomous, it is assumed that the event rate for the two interventions is equal to p i.e.
p = p intervention = p control. With this assumption the formula becomes as under(48); 2N = 4p (1-p) (Zα - Zβ)2 / Δ2 A brief explanation of the variables used in the equation is given below P = 0.957 (proportion of children attaining seroprotection correlates) (104) Δ = 0.05 (5%) Zα = 2.57 (99% CI) Zβ = 0.89 (80% power) After solving the equation 2 N comes out to be 228, hence 1 N =114. Thus, to get the answer to questions posed in the study; the number of subjects in one arm of the trial is 114. The trial was planned to end after the last follow up
