The Impact of a Chatbot on COVID-19 Vaccine Perceptions and Intentions in ESPERES Cohort (ESPERES-COVID-19-CHATBOT)

The Impact of a Chatbot on COVID-19 Vaccine Perceptions and Intentions in ESPERES Cohort (ESPERES-COVID-19-CHATBOT)

The ESPERES Prospective Cohort of Healthcare Workers in France: Study on the Impact of a Chatbot on COVID-19 Vaccine Perceptions and Intentions

Vaccine hesitancy also concerns healthcare workers (HCWs). However, HCWs are at the frontline of the COVID-19 pandemic and identified as a priority target group for COVID-19 vaccines. Thus, the identification of interventions likely to improve COVID-19 vaccine attitudes and intentions among HCWs is of interests to increase the vaccine coverage among HCWs. The study hypothesis is that the use of the same chatbot as the one tested in the general population in France could also improve the COVID-19 vaccine intentions and perceptions in HCWs.

Most countries face the issue of vaccine hesitancy, with sizeable fractions, or sometimes the majority, of the public opposing some vaccines. The problem is particularly acute in the case of COVID-19 vaccination: first, a high uptake of COVID-19 vaccines is necessary to reach and sustain herd immunity; second, and to the best of our knowledge, no country is currently planning on making COVID-19 vaccination mandatory, making public approval essential. Unfortunately, hesitancy towards COVID-19 vaccines is high in many countries. It is therefore crucial that health authorities make it as easy as possible for people who want to be vaccinated to do so, but also that they devise communication strategies to reassure vaccine hesitant individuals. A recent study performed in France (Altay S, et al. 2021) introduced a novel messaging strategy: the use of a chatbot that answers people's questions about COVID-19 vaccines. Results suggest that a properly scripted and regularly updated chatbot could offer a powerful resource to help fight hesitancy towards COVID-19 vaccines. Vaccine hesitancy also concerns healthcare workers (HCWs). However, HCWs are at the frontline of the COVID-19 pandemic and identified as a priority target group for COVID-19 vaccines. Thus, the identification of tools likely to improve COVID-19 vaccine attitudes and intentions among HCWs is of interests to increase the vaccine coverage among HCWs. The study hypothesis is that the use of the same chatbot as the one tested in the general population in France could also improve the COVID-19 vaccine intentions and perceptions in HCWs. The primary objective will be to evaluate the short impact of the use of a chatbot on the evolution of the COVID-19 vaccine perceptions in HCWs. The secondary objectives will be: To evaluate the impact at 15-days of the use of a chatbot on the evolution of the COVID-19 vaccine perceptions in HCWs. To evaluate the short and at 15-days impact of the use of a chatbot on the evolution of the COVID-19 vaccine intentions in HCWs globally and by vaccine. To describe the chatbot connection rate, the interaction time with the chatbot and feedback from the device These objectives (primary and secondary) will be decline in the sub-groups of caregivers and non-caregivers. The study population will be made up of HCWs from the ESPERES cohort. For the purposes of this study, a "healthcare worker" is defined as an individual who currently works in a setting where individuals receive healthcare (i.e. individuals do not have to work directly with patients, but may have any role within a setting where individuals receive healthcare, such as administrative office, housekeeping, food service, etc.). It is a randomized, two-arm, open label, parallel trial. Participants will be asked to answer a new questionnaire on their COVID-19 vaccine intention and perception, 15 days after their inclusion in the ESPERES cohort. Participants in ESPERES cohort who meet the eligibility criteria for the trial evaluating the chatbot will be randomized into 2 arms: access to the chatbot (intervention group) or access to simple information (control group). Randomization will be carried out at a 1:1 ratio, on blocks of four, stratified on the caregiver/non-caregiver group and the health facility. Randomization will be done online which ensures allocation concealment. Depending on the randomization group, participants access the chatbot or simple information. They can consult them as long as they wish. Participants will be contacted immediately after the first connection to the chatbot and again 15-days later to complete questionnaires on their COVID-19 vaccine hesitancy and perception. The statistical analyses planned in the research protocol and those responding to specific subsequent studies nested in ESPERES and validated by the scientific committee, will be performed by the pharmaco-epidemiology center (CEPHEPI), under the responsibility of Dr Candice ESTELLAT and Pr Florence TUBACH. The CEPHEPI is backed by the public health department of the hospital group Pitié-Salpêtrière - Charles Foix, AP-HP.

Participants randomly assigned to this arm will access to a Chatbot condition, in which they had the opportunity to interact with the chatbot on COVID-19 vaccine hesitancy.

Participants randomly assigned to this arm will access to a brief text describing the way vaccines work.

This intervention will provide to participants information on COVID-19 vaccine hesitancy consisting of the possibility of accessing a long series of questions and answers online about this vaccination.

This control intervention will provide to participants a brief text decribing the way vaccines works.

The change in the COVID-19 vaccine perceptions (5 questions) using a seven-point Likert scale, immediately after the access to the intervention

"I think vaccines against COVID-19 are safe" "I think vaccines against COVID-19 are effective" "I think we know enough about the COVID-19 vaccines" "I think we can trust the people who produce the COVID-19 vaccines" "I think it is important to get vaccinated against COVID-19". The COVID-19 vaccine perceptions variable is the mean of the response to these five questions, each answered on a 7-point likert scale.

The change in the COVID-19 vaccine perceptions (5 questions) using a seven-point Likert scale, 15-days after the access to the intervention

"I think vaccines against COVID-19 are safe" "I think vaccines against COVID-19 are effective" "I think we knowenough about theCOVID-19 vaccines" "I think we can trust the people who produce the COVID-19 vaccines" "I think it is important to get vaccinated against COVID-19". The COVID-19 vaccine perceptions variable is the mean of the response to these five questions, each answered on a 7-point likert scale.

The change in the COVID-19 vaccine intention, immediately after the access to the intervention with the following question: "Do you personally intent to be vaccinated against COVID-19?"

The change in the COVID-19 vaccine intention, 15 days after the access to the intervention with the following question: "Do you personally intent to be vaccinated against COVID-19?" (yes, as soon as possible / yes but I prefer to wait / no)

Changes into the attitudes and intention towards the COVID-19 vaccines according to the vaccine immediately and 15-days after the access to the interventions

Participants in the Chatbot intervention will be ask, after the access to the intervention, to answer to a satisfaction questionnaire

Participants in the Chatbot intervention will be asked whether they had been able to access the chatbot, whether the Chatbot was intuitive, pleasant, frustrating, whether the information provided in the chatbot were too simple or too complicated, and whether they had unanswered questions that they wish the chatbot had addressed (free text entry).

Inclusion Criteria: Legal age Healthcare workers defined as individuals who currently work in a setting where individuals receive healthcare (i.e. individuals do not have to work directly with patients, but may have any role within a setting where individuals receive healthcare, such as administrative office, housekeeping, food service, etc.) Included in ESPERES cohort and with the initial questionnaire completed Agreeing to participate (digital consent) Exclusion Criteria: No internet access Vaccinated against COVID-19 (at least one dose)