Use of a Practitioner-Friendly Behavior Model to Identify Factors Associated with COVID-19 Vaccination and Other Behaviors

"The COVID-19 pandemic laid bare the urgency to design behavioral interventions to increase COVID-19 vaccine uptake. At the same time, it has highlighted the absence of behavior change models in designing immunization interventions."

Evidence suggests that theory-based behavioural interventions are more likely to have an impact, yet global health institutions have recognised the lack of capacity for the design and implementation of these types of interventions in low-and-middle income countries (LMICs). This article is a response to the identified need to translate social and behavioural science (SBC) concepts into "practitioner-friendly" models that can be used by intervention designers, implementers, and evaluators with limited technical and financial resources. Specifically, it illustrates the use of the Fogg Behavior Model (FBM), applying data on four different behaviours in three countries: COVID-19 vaccine uptake (Nigeria), condom use (Pakistan), iron folate use (India), and modern contraceptive use (Nigeria).

FBM is not per se a model of health behaviour; rather, it is a model of human behaviour that has been applied to health behaviour. The FBM can be visualised in two dimensions, with motivation on the y-axis and ability on the x-axis. For a specific behaviour, motivation can range from high to low, and ability can range from high to low. The FBM proposes that a behaviour happens when a person with high motivation and high ability is prompted. By contrast, a person with low motivation and low ability is not likely to adopt a behaviour when prompted. There is a threshold (the "action line") above which a person with sufficient motivation and ability will adopt a behaviour when prompted.

This study created a categorical variable for motivation and ability and tested whether high motivation and high ability are associated with a greater likelihood of adopting a behaviour. It tested the utility of the FBM for both formative research studies (in the case of Nigerian healthcare workers (HCWs) and adolescents and young women) and evaluations (in the case of married Pakistani men and Indian women of reproductive age). Specifically:

• In July 2021, an online survey of 496 Nigerian HCWs was conducted to identify factors associated with the uptake of a COVID-19 vaccine. Among the questions were one that asked about their motivation and another about their ability to get vaccinated.
• In 2009, a household panel survey of married men aged 15-49 was conducted in urban Pakistan to assess the effects of a television-based social marketing condom promotion campaign. The present study used a panel of 617 married men to test the applicability of the FBM to the behaviour of interest: a man's use of a condom during last sex with his wife.
• In 2020, an online survey was conducted among women aged 18-49 in the states of Uttar Pradesh and Madhya Pradesh in India. Among the questions were those that asked the 1,136 women included in this analysis asked about their motivation and ability to use iron folate.
• In 2018, a household survey of women of ages 14-24 was conducted in Lagos, Kaduna, and Kano states in Nigeria. The present study tested the FBM using data about 618 Nigerian women who reported ever having had sex. The behaviour of interest was a woman's current use of a modern contraceptive method.

The analysis shows that, in Nigeria, HCWs with high motivation and high ability had 27 times higher odds of being vaccinated against COVID-19. In Pakistan, married men with high motivation and high ability had 35 times higher odds of condom use with their wives. In India, women with high motivation and high ability had 9 times higher odds of iron folate use. In Nigeria, adolescents and young women with high motivation and high ability had 8 times higher odds of contraceptive use.

In explaining the adoption of the four behaviours, the study found that a motivation-ability variable with three categories showed a powerful correlation with behaviour. Specifically, motivation and ability had three distinct levels of effects: high motivation and high ability had the strongest effect, followed by high motivation or high ability and low motivation and low ability.

"That three empirically meaningful segments based on motivation and ability can be obtained easily by asking one question on motivation and one on ability suggests that use of the FBM will bear immediate rewards for practitioners. It will allow practitioners to measure the size of each of these segments and decide how to use project resources to influence behavior. Use of the model will inform a practitioner about whether a behavioral intervention should focus on motivation, or ability, or on both motivation and ability."

To illustrate how an FBM-inspired analysis could help inform a behavioural intervention, consider that 64% of young women in Nigeria and 56% of married men in Pakistan had low motivation and low ability to adopt contraception. In such cases, implementers might want initially to focus on prompts to action for the segment with high motivation and high ability: a segment that is on the verge of behavioural adoption. Following that early "win", implementers could focus on increasing motivation (for those with low motivation and high ability) or increasing ability (for those with high motivation and low ability). A staged approach with "wins" at each stage could help build confidence among practitioners as they take on the task of increasing motivation or increasing ability or increasing both motivation and ability.

In conclusion, the study findings suggest that the FBM has the potential to be applied in low-resource settings for the design, implementation, and evaluation of behavioural interventions. Per the author: "There is a need to understand how well practitioner-friendly models of behavior perform against more traditional models that have been tested and validated. What are the shortcomings of practitioner-friendly models? In what contexts are they most useful?....What resources would a practitioner need to implement a model such as the Fogg Behavior Model in an LMIC? This paper paves the way for a broader discussion on this topic by behavioral scientists working in low-resource settings."