A Postpartum Intervention for Vaccination Promotion by Midwives Using Motivational Interviews Reduces Mothers' Vaccine Hesitancy, South-eastern France, 2021 to 2022: A Randomised Controlled Trial

"This study shows that an educational intervention on vaccination of newborns using MI with mothers in maternity wards in France was significantly associated with a 32.7% reduction in their VH score and an 8.5% increase in their intention to vaccinate their infants at 2 months of age."

In France, 11 infant vaccines are required for entry to daycare and school, yet vaccine hesitancy (VH) is still frequent among parents. One approach to VH that has been studied is the motivational interview (MI), which is a collaborative conversation designed to understand reasons behind parents' hesitation to have their children vaccinated and reinforce their motivation and commitment to behaviour change. This randomised controlled trial (RCT) evaluated whether MI by midwives, carried out in two maternity wards in southeastern France with new mothers, could impact VH. Two secondary objectives were to (i) test the differential impact of MI on VH and intention to vaccinate their children (VI) according to social status and (ii) evaluate mothers' acceptance of and satisfaction with this approach.

From November 2021 to April 2022, the researchers implemented a parallel-arm, multicentre RCT with individual randomisation, comparing the impact assessed by a pre-post questionnaire of MI intervention versus standard care (a 4-page educational leaflet about vaccination) as the control. Participants were randomly assigned to the intervention group (MI, n = 365) or the control group (leaflet, n = 371). This article presents the results of the comparisons just after the MI or leaflet distribution; further results of the RCT covering a 7-month follow-up will follow this initial report.

Prior to the intervention, midwives were trained via both e-learning and a face-to-face workshop covering topics such as MI's theoretical foundations and its adaptation to vaccination and using techniques such as role-playing exercises. The trained midwives then put this approach into practice with mothers during a pilot phase the following month. The midwives performed MI individually with participants in sessions that lasted 10 to 30 minutes, as needed, applying the following principles: (i) establish, as the initial objective, a trust relationship by listening carefully and without judgement to the mothers' and, where applicable, partners' concerns, without trying to correct or argue against certain information; (ii) understand the specific reasons for their hesitation to ascertain what information would improve their perception of vaccination's importance; (iii) deliver this information with their agreement, to support their personal choice; and (iv) respect their personal autonomy and try to direct the conversation towards a more favourable position towards vaccination, while encouraging partnership and avoiding discord.

The researchers collected data through two self-administered questionnaires completed by study participants during the postpartum stay at the maternity ward: one before the MI or leaflet distribution (T0, pre-questionnaire, 34 items) and a second afterwards (T1, post-questionnaire, 35 items).

In short, the RCT found that MI of mothers in maternity wards significantly reduced their VH. The women who were initially the most hesitant to vaccinate their newborns were the most impacted by the interviews. A reduction in VH was greatest for mothers with low education levels. More specifically:

• The control group's average VH score was 33.2/100 at T0 and 27.6/100 at T1, a 16.7% decrease (-5.5 points, p = 0.0002). In the MI group, the corresponding scores were 34.8/100 at T0 and 23.4/100 at T1, a 32.7% decrease (-11.4 points, p < 0.0001). The estimated difference-in-difference (D-I-D) effect without covariables, that is, the net T1-T0 difference of the VH scores between the MI and control groups, was -5.8 points (p = 0.01); adding covariables to the overall model produced a similar result (D-I-D effect: -5.0 points, p = 0.02).
• The D-I-D effect was significant in the high and moderate VH categories (high baseline score: -9.9 points, p = 0.001; moderate: −8.3 points, p < 0.0001) but not the low VH category (-2.3 points, p = 0.14).
• The D-I-D effect of MI vs the control intervention on VH scores was significant among participants with a low education level (-9.3 points, p = 0.01) but not among those with a high level (-2.7 points, p = 0.31).

In terms of vaccination intentions:

• In the control group, the average VI score, i.e. intention to vaccinate one's infant at 2 months of age, did not increase significantly: from 8.7 to 8.8/10 (p = 0.28). In contrast, the VI score did increase in the MI group: from 8.3 to 9.0/10 (+8.5%, p < 0.0001). The D-I-D effect between MI and control groups was estimated at +0.6 point for the VI score (p = 0.005); adding covariables to the overall model produced the same results.
• In the MI group, VI scores increased significantly regardless of initial VH score category, with the highest increase (+24.2%, p < 0.0001) in the highest initial VH category. The control group showed no significant increase in any initial VH category.
• In the MI group, VI increased by 11.1% among participants with low education levels (p = 0.0002) and by 6.8% in those with a high education level (p = 0.001). Vaccine intention did not change significantly in the control group, regardless of education level.

In the MI group, 99.4% (320/322) mothers responded to the satisfaction questionnaire: 95.0% reported they appreciated participating, and the great majority would recommend extending the MI intervention to other maternity wards. Almost all found the MI intervention useful, 90.3% had no more questions about vaccination afterwards, 96.2% reported that it had respected their views on vaccination, and 94.7% said its duration was appropriate.

Per the researchers, the results suggest that MI could reduce VH especially among mothers who are part of socially deprived populations, whose VH was highest in the study. The MI approach (listening, non-judgemental) could have enabled these mothers to better express their concerns and allowed midwives to respond more appropriately than a standardised leaflet could allow (no significant VH reduction found among the less educated in the control arm). "Even mothers with little hesitation can have some concerns about vaccination, which MI allows them to express. Answering their questions may strengthen their confidence in vaccinating their infants, a useful effect in this era of vaccine misinformation."

"The impact of MI on mothers' VH and intentions to vaccinate, the apparent reduction of social disparities in vaccine acceptance and parental satisfaction with several aspects of the programme all argue for investigating conditions for larger-scale implementation of MI intervention in France possibly through an implementation study, including an assessment of its impact on vaccination delays between the recommended and actual age at vaccination and a cost-effectiveness analysis."

In conclusion, the results, combined with high participant satisfaction with the intervention overall and for several of its specific aspects, suggest that MI, as applied in maternity wards in the days after birth, can be an effective strategy to reduce VH and improve mothers' intention to vaccinate their newborn child.