Introduction

Vaccination against human papillomavirus (HPV) is an important tool in the global effort to eliminate cervical cancer [1–3], especially in countries like Malawi, which has the second-highest cervical cancer incidence and mortality globally [4] and challenges in delivering cancer screening at scale [5]. Malawi introduced the HPV vaccine in 2019 [6], and eligible girls (9–14 years) can receive the vaccine free of charge at schools, health facilities, and in communities. However, only 13% of eligible Malawian girls received any HPV vaccine doses in 2021 and 2022 [7]. Other countries face a similar challenge [7,8].

There are many reasons for low uptake of the HPV vaccine globally; our team has identified numerous important intra- and inter-personal factors that may shape parents’ decisions about whether to vaccinate their daughters, including attitudes about the HPV vaccine, perceiving pro-HPV vaccination social norms, and speaking to other people about HPV vaccination – including health workers [9,10]. Malawian girls whose parents who had spoken with a health worker about HPV vaccination had nearly three times the adjusted odds of having received the HPV vaccine than their peers whose parents had not had such a conversation [10]. A highly effective evidence-based intervention to increase HPV vaccination is health worker recommendation [11] – but there is limited evidence from lower-income countries about how to encourage health workers to recommend the HPV vaccine.

Our team is undertaking a multi-year study to develop, implement, and evaluate implementation strategies to increase health worker recommendation of HPV vaccination to girls and young women with HIV in Malawi, who are a priority group for HPV vaccination due to their high risk for cervical cancer [12–14]. Following the multi-phase optimization strategy (MOST) paradigm [15,16], this study (“Kupewa,” which means “prevent” in the Chichewa language of Malawi) will evaluate the effectiveness and implementability of three implementation strategies, alone and in combination.

Intervention optimization is a deliberate process to evaluate multi-component interventions (here, implementation strategies [17]) by balancing effectiveness against other outcomes like affordability or scalability. MOST is a specific approach to intervention optimization that involves three distinct phases: preparation (including identifying, designing, and pilot testing the components), optimization (evaluate the components alone and in combination, using an optimization randomized controlled trial design), and evaluation (use a traditional two-armed trial design to assess effectiveness of the optimized multi-component approach) [18].

Here we present results from the pilot (“preparation”) phase of the Kupewa study, which was designed to apply MOST to implementation strategies for increasing HPV vaccination in Malawi. The goal of the preparation phase of MOST is to ensure that the strategies are implementable and show promise for a subsequent optimization trial [19]. Thus, in this pilot study, we sought to ascertain if knowledge and self-efficacy in skills to counsel about HPV vaccination increased following exposure to implementation strategies designed to target these hypothesized determinants of health worker recommendation.

Materials and methods

Results

In total, 45 health workers participated in this pilot study: 19 participated in the training, and 26 viewed the videos (see S1 Table for sample characteristics).

Knowledge improved on 11 of the 14 items from pre- to post- test (Fig 1). The biggest increases were seen for items that reflected specific clinical knowledge, for example, “As a two-dose series, the minimum interval between the 1st and 2nd dose is 6 months,” which was answered correctly by only 26.7% of respondents during the pre-test and by 100% in the post-test (Fig 1). All health workers responded correctly to “A person could have HPV for many years without knowing it” during the pre- and post- tests, so there was no improvement on this item, and over 90% responded correctly to “The HPV vaccines offer protection against all sexually transmitted infections” at both pre- and post-test (Fig 1).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Knowledge item changes from pre- to post-test (n=45).

https://doi.org/10.1371/journal.pgph.0006508.g001

Overall knowledge scores increased from an average of 68.1 points to 87.8 points (Table 1). At the sites where the training was piloted, health workers’ knowledge scores increased on average by 24.1 points (from 67.3 points to 91.4 points); at the sites where videos were piloted, the corresponding increase in knowledge scores was 16.5 points (from 68.8 points to 85.2 points).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Outcomes at pre-test and at post-test, overall and by intervention assignment.

https://doi.org/10.1371/journal.pgph.0006508.t001

Self-efficacy in counseling skills increased on all items from pre- to post-test (Fig 2). The largest increases were in being able to identify issues to discuss (increased by 31.5%), and making a plan for the conversation, urging parents to expand on their concerns, and handling angry parents (all increased by 24.7%).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Self-efficacy in counseling skills item changes from pre- to post-test (n = 45).

https://doi.org/10.1371/journal.pgph.0006508.g002

Health workers reported high self- efficacy, with an average pre-test self-efficacy score of 82.6 and an average post-test score of 96.4. The score increased following exposure to the pilot intervention by 13.7 points on average across all sites (Table 1). The increase was greater at video sites (14.2 points on average) than at training sites (13.1 points on average).

We found very high acceptability of the training and the videos: nearly every element received a perfect score of 4 (possible range 1–4) (S2 Table).

Discussion

In this pilot study, these implementation strategies increased health workers’ knowledge about HPV vaccination and increased their self-efficacy in skills to discuss HPV vaccination with girls and their parents. The findings provide strong support for the Kupewa study, which will formally evaluate whether these strategies change health workers’ behavior– i.e., if these increases in knowledge and self-efficacy translate to increased recommendation of HPV vaccination to girls and young women living with HIV as hypothesized, and if this results in greater vaccine uptake among these girls and young women.

We found greater increases in knowledge among health workers who received the training intervention, and greater increases in self-efficacy among those who watched the videos. These differential changes correspond to the underlying constructs from COM-B that we hypothesized would respond to these specific activities.

In our sample, health workers lacked specific knowledge about HPV vaccination, despite being well-informed about the public health-related aspects of HPV and cervical cancer. Knowledge gaps mostly reflected clinical details (timing between doses, contraindications, etc.). This suggests the importance of training as part of an intervention to improve provider recommendation for HPV vaccination. Our full trial will evaluate training alone and in combination with video vignettes to determine which combination most increases health workers’ recommendation of HPV vaccination during discussions with girls (and parents) and young women.

Health workers generally reported high self-efficacy in counseling skills during the pre-test, but there were large improvements in thoughtfully engaging parents in the conversation and being proactive in steering the discussion, even when difficult. Given common concerns about HPV vaccine safety in Malawi [9], these more advanced counseling skills will likely be critical to achieve widespread vaccine uptake.

Some limitations should be noted. First, as this was a pilot study, it was not designed or intended to detect statistically significant differences; we therefore did not conduct these tests, but subsequent evaluation studies should. Second, because this was a pilot study, results may not generalize to other contexts. A larger study would also permit subgroup comparisons (e.g., cadres or duration of job experience). Related, we did not collect any data about who participated in the pilot study at these sites versus who did not, so we cannot assess whether there are any unique attributes of participants owing to potential selection bias. Third, knowledge of certain topics was near-universal during the pre-test, and pre-test self-efficacy was high for many items, resulting in ceiling effects. Likewise, nearly all participants reported the maximum acceptability scores for all items, suggesting that improved implementation outcome measures may be warranted.

Conclusions

Implementation strategies aiming to increase Malawian health workers’ knowledge and self-efficacy to counsel may be implementable and may result in complementary improvements. Our team will next evaluate whether these strategies increase health workers’ recommendation of HPV vaccination and HPV vaccine uptake among girls/young women living with HIV in Malawi through an optimization trial in a study informed by the MOST framework.

Supporting information

Acknowledgments

The authors gratefully acknowledge the health facilities, leadership, and health workers who supported this study and who generously shared their time and insights for this research. We thank our key partners at the Malawi Ministry of Health for their very valuable inputs to the design and implementation of this pilot intervention.