BabyGel pilot: a pilot cluster randomised trial of the provision of alcohol handgel to postpartum mothers to prevent neonatal and young infant infection-related morbidity in the community

Abstract

Background: Alcohol-based hand rub (ABHR) is widely used in both health and social facilities to prevent infection, but it is not known whether supplying it for regular perinatal use can prevent newborn sepsis in African rural homes. Our study piloted a cluster randomised trial of providing ABHR to postpartum mothers to prevent neonatal infection-related morbidity in the communities.

Methods: We conducted a pilot parallel cluster randomised controlled trial across ten villages (clusters) in rural Eastern Uganda. Pregnant women of over 34 weeks' gestation were recruited over a period of 3 months. Both clusters received the standard of care of antenatal health education, Maama Kit, and clinic appointments. In addition, women in the intervention villages received ABHR, instructions on ABHR use, a poster on the 'three moments of hand hygiene', and training. We followed up each mother-baby pair for 3 months after birth and measured rates of consent, recruitment, and follow-up (our target rate was more than 80%). Other measures included ABHR use (the acceptable use was more than four times a day) and its mode of distribution (village health workers (VHWs) or pharmacy), acceptability of study protocol and electronic data capture, and the use of WHO Integrated Management of Childhood Illness (IMCI) tool to screen for newborn infection.

Results: We selected 36% (10/28) of villages for randomisation to either intervention or control. Over 12 weeks, 176 pregnant women were screened and 58.5% (103/176) were eligible. All, 100% (103/103), eligible women gave consent and were enrolled into the trial (55 intervention and 48 control). After birth, 94.5% (52/55) of mothers in the intervention and 100% (48/48) of mothers in the control villages were followed up within 72 h. Most, 90.9% (50/55), of the mothers in the intervention villages (96.2% of live births) and 95.8% (46/48) of mothers in the control villages (95.9% of live births) were followed up at 3 months. In intervention villages, the average hand rub use was 6.6 times per day. VHWs accounted for all ABHR stock, compared to the pharmacy that could not account for 5 l of ABHR. The screening tool was positive for infection among a third of babies, i.e. 29.2% (14/48) in the intervention villages versus 31.4% (16/51) in the control villages.VHWs completed the first four questions of IMCI screening tool with ease and accuracy. There were no adverse reactions with the ABHR.

Conclusion: It is feasible to conduct a cluster-randomised controlled trial (cRCT) of the provision of ABHR to postpartum mothers to prevent neonatal infection-related morbidity in the community in resource-poor settings. Our results indicate that home recruitment promotes excellent follow-up and retention of participants in community trials. The intervention was safe. This pilot study informed the substantial changes necessary in the larger cRCT, including a change in the primary outcome to a composite outcome considering multiple methods of infection detection. A large BabyGel cluster randomised controlled trial is now required.

Trial registration: ISRCTN67852437, registered March 02, 2015.

Trial funding: Medical Research Council/WellcomeTrust/DfID (Global Health Trials Scheme).

Keywords: BabyGel; Hand hygiene; Infant sepsis; Infection; Morbidity alcohol-based hand rub; Mothers; Pilot trial.

Fig. 1

The village map showing the villages drawn and the distribution of participants in each village for the BabyGel pilot trial. A map of villages around the community health facilities drawn locally. The 10 BabyGel study villages were selected from the above map. There is also a Google Earth map showing the distribution of all the study participants from each respective village. Each number (e.g. 0412) is the assigned identifier for a participant in her household. The first two digits of the number is a study village number (e.g. 04) and the preceding two digits represent the consecutive number for each participant as recruited in each village. The first five (01 to 05) were intervention while the last five (06 to 10) were control villages. This distribution of participants shows clearly the careful selection of villages to observe the effects of contamination in this study

Fig. 2

CONSORT diagram showing the flow of participants through each stage of the BabyGel pilot cluster randomised trial. This CONSORT flow chart illustrates the screening and randomisation of clusters and the flow of participants in the study

Fig. 3

The ‘three moments for community neonatal hand hygiene’ poster developed for the BabyGel pilot trial. This shows an illustrative and diagrammatic representation of the key moments of hand hygiene for newborns in the community

Fig. 4

Adapted WHO IMCI screening tool for infection. This tool was administered to newborns at particular encounters as a screening tool for infection

Fig. 5

Illustration of the method for the identification of infection-related infant outcomes in the BabyGel pilot trial. Infants participating in the study had the IMCI screening tool administered either at routine research visits or when their mothers brought them to participating hospitals or health centres. The group who screened positive were referred to the paediatric team at Mbale Hospital for clinical care, and along with those who died, composed the primary outcome for the study. Those who screened negative returned to the community. At the hospital, those who had screened positive had both a clinical and bacterial diagnosis performed; these were collected as secondary outcomes