This descriptive, cross-sectional, and analytical study was conducted in public health facilities (HFs) in the Central-South region of Burkina Faso. Its primary objective was to assess healthcare workers’ attitudes toward AFP surveillance and identify factors influencing case reporting. Data collection took place from January 2 to 24, 2022.

The study focused on healthcare personnel working in these facilities, including doctors, health officers, nurses, midwives, birth attendants, and itinerant health agents. To be included, professionals had to work in a public facility, be present at the time of the survey, and provide informed consent.

Healthcare workers were selected from public health facilities using proportional stratified random sampling. First, 78 out of 155 operational HFs were selected. Each stratum corresponded to a health district, with the number of facilities drawn proportional to the size of its healthcare workforce.

A total of 278 agents were randomly selected from approximately 580 eligible personnel, giving each a 48% probability of inclusion (278/580). Each worker was assigned a unique number prior to random selection. This method ensured representativeness, minimized bias, and improved generalizability of findings.

The selected sites included Health and Social Promotion Centers (HSPC), where the surveys covered dispensary and maternity services, as well as Medical Centers with surgical units, specifically pediatrics and general medicine departments. These facilities were chosen due to the high likelihood of encountering AFP cases.

Data was collected using a structured questionnaire administered via KoBoCollect on Android phones. The instrument had two sections:

The second section included items rated on a five-point Likert scale (five levels: Strongly agree, Agree, Neutral, Disagree, strongly disagree), adapted from previously validated tools. It was informed by a literature review and pretested with 15 professionals to ensure clarity and coherence.

The questionnaire measured four key attitude dimensions:

Interest in surveillance, measuring the level of attention, importance, and perceived relevance of AFP activities.

Involvement, reflecting the perceived integration into surveillance workflows and participation in related activities.

Motivation, capturing perceptions of institutional recognition and reward systems.

Satisfaction, assessing opinions about the organization and resources available for surveillance.

The reliability of the questionnaire was assessed using Cronbach’s alpha, which yielded acceptable internal consistency for each subscale (α ranging from 0.71 to 0.80). Content validity was ensured through expert review by three epidemiologists familiar with AFP surveillance in Burkina Faso.

The entire study was conducted in compliance with ethical standards, with authorization obtained from the health research ethics committee (certificate No. 2022-12-264 dated 12/23/2022).

Data analysis was performed using STATA software. The descriptive analysis consisted of calculating the frequencies of respondents based on their sociodemographic and professional characteristics, as well as attitudinal dimensions. Regarding the association analysis, the dependent variable was whether the healthcare worker had reported at least one AFP case during the past three years (binary outcome: Yes/No). Independent variables included socio-professional characteristics (e.g., gender, education, qualification) and the four attitude dimensions.

Given the binary nature of the outcome variable (case reported vs. not reported), logistic regression was chosen as the most appropriate method for identifying associations and estimating the probability of case reporting as a function of the predictor variables.

An initial univariate logistic regression helped select candidate variables (p < 0.20), which were then entered into a multivariate model.

However, the variable “qualification” was forcibly included in the analytical model despite not reaching the 20% significance threshold due to its strong theoretical and contextual relevance in the surveillance of Acute Flaccid Paralysis (AFP). Indeed, qualification reflects the level of initial training of healthcare personnel, which directly influences their technical competencies, ability to identify clinical signs suggestive of AFP, to correctly apply the case definition, and to initiate the notification process within the timeframes required by surveillance standards [5] .

Previous studies have established a strong link between qualification level and performance in epidemiological surveillance, particularly for diseases with epidemic potential or requiring high clinical vigilance, such as poliomyelitis [6] [7] . In the context of Burkina Faso, where human resources in health are heterogeneous in terms of training, professional qualification constitutes a key structural determinant that may condition the effectiveness of the notification system.

Therefore, regardless of its statistical significance, this variable was retained in the model as a potential confounder and adjustment variable, to allow for a better interpretation of the relationships between other factors studied and the notification of AFP cases.

A multivariate logistic regression analysis was conducted using a “stepwise backward” elimination approach to identify factors associated with AFP case reporting.

Socio-professional status of the respondents

Table 1 shows that the sample was predominantly composed of women (62.3%) and nurses (45.1%). More than half of the participants had a secondary education level (59.3%). Regarding job roles, 35.8% did not have a specific responsibility. Finally, 47.8% of respondents had received training on AFP surveillance at the time of the survey.

Attitudes of Healthcare Professionals on AFP Surveillance

The results of Table 2 show that the majority of healthcare professionals are highly supportive of AFP surveillance. More than 60% of respondents strongly agree with the assessed dimensions, and approximately 25% - 30% agree. Very few people express neutral or negative opinions, indicating an overall positive attitude.

Initial selection of variables for cross-tabulation

In Table 3 , the variables significantly associated with AFP case notification at the 20% threshold were as follows: among the control variables—Gender of Respondents, level of education, position, and training on AFP surveillance; and among the attitudinal variables—interest, motivation, involvement, and satisfaction. The variables workplace, health district, and professional qualification did not show a significant association at this threshold. Only variables with a significant association at the 20% level were retained for inclusion in the multivariate analysis. However, as stated in the methodology, the variable professional qualification was included in the adjusted model due to its potential influence on case reporting.

Note: variables marked with an asterisk (*) are considered statistically significant at the 20% threshold (P < 0.20).

Final model

In Table 4 , the multivariate analysis revealed statistically significant associations between AFP case notification and selected variables. Health workers who had received training on AFP surveillance were more likely to report AFP cases (aOR = 2.303; 95% CI: 1.288 - 4.115; p = 0.005). Male respondents were also more likely to report cases compared to females (aOR = 1.897; 95% CI: 1.072 - 3.358; p = 0.028). Regarding interest in AFP surveillance, those who agreed (rather than strongly agreed) showed a lower likelihood of notification (aOR = 0.398; 95% CI: 0.181 - 0.876; p = 0.022). No other variable showed a statistically significant association with AFP case notification after adjustment.

The results of this study show that the majority of participants are women (62.3%) and nurses (45.1%), with a predominance of professionals with a secondary education level (59.3%) and training in PFA surveillance for approximately 47.8% of respondents. This demographic profile reflects a reality often observed in many similar studies, where women dominate in the healthcare sector, particularly in nursing positions, which are often considered less prestigious than medical roles [8] .

The absence of specific responsibilities in 35.8% of participants could suggest a lack of structure in health teams, which may affect the effectiveness of surveillance and case reporting. This observation aligns with a study that highlights organizational challenges in fragile health systems, where poorly defined or poorly assigned roles can hinder effective epidemiological data management [9] . However, these results contrast with another study where better structuring of health teams and clearly defined responsibilities were associated with improved reporting of vaccine-preventable diseases [10] .

Healthcare professionals’ attitudes toward PFA surveillance are generally very positive, with more than 60% of respondents expressing strong support. This result aligns with the conclusions of other studies on the importance of healthcare professionals’ engagement in disease surveillance strategies [11] . However, this enthusiasm contrasts with the relatively low level of actual case reporting, suggesting that positive attitudes alone are insufficient to ensure effective surveillance. This highlights a disconnect between attitudes and practice, pointing to possible institutional, motivational, or contextual constraints.

The analysis of factors associated with PFA case reporting identified several significant predictors: training received, interest in surveillance, and gender.

The significant impact of training (aOR = 2.303, p = 0.005) confirms findings from other contexts [10] [12] , and supports the hypothesis that regular, targeted training improves detection and notification practices. This reinforces the argument that ongoing capacity-building is essential to sustain the quality of surveillance, particularly in resource-constrained health systems where staff turnover is frequent.

The study also shows that healthcare workers who “agree” (rather than “strongly agree”) on the importance of AFP surveillance were significantly less likely to report cases (aOR = 0.398, p = 0.022). This gradient effect suggests that only a high level of conviction and prioritization of surveillance activities translates into concrete action. Moderate interest may reflect competing priorities, lack of institutional incentives, or limited resources. These findings align with studies showing that frontline staff are more reactive when surveillance tasks are clearly supported and rewarded by their institutions [13] [14] .

Regarding gender, male healthcare workers were nearly twice as likely to report AFP cases as females (aOR = 1.897, p = 0.028). This gender gap could be linked to multiple factors: men may be more often assigned to leadership or technical roles that involve surveillance tasks; they may have more access to training opportunities; or cultural norms might influence confidence and authority in clinical decision-making. In contrast, female staff may be overrepresented in subordinate or overburdened roles, which limits their ability to engage in reporting. This finding echoes research from Ghana and other low-income settings showing structural and sociocultural gender disparities in the health workforce [15] .

However, even after forcing the variable “qualification” into the multivariate logistic regression model, it did not reach statistical significance. This suggests that, in this context, the formal qualification of healthcare workers may not be the primary determinant of AFP case reporting. One possible explanation is that AFP case detecting may be more strongly influenced by individual attitudes, local management dynamics, or practical training experiences rather than by academic credentials alone. This finding is consistent with results from other studies. For instance, similar patterns were reported in Kenya [16] , Mozambique [17] , and across multiple settings in sub-Saharan Africa [6] , where formal qualification was not significantly associated with performance in surveillance activities. Instead, factors such as experience in the field, motivation, regular supervision, and continued training were found to have greater influence.

Crucially, these results address the gap identified in the introduction regarding the limited evidence on how healthcare workers’ attitudes influence AFP case reporting in Burkina Faso. They provide empirical support for the idea that beyond knowledge, engagement, motivation, and professional empowerment are critical levers for improving surveillance performance.

The findings have important implications for strengthening AFP surveillance in Burkina Faso:

Ultimately, enhancing the responsiveness of AFP surveillance systems requires a multidimensional approach: technical capacity-building, behavioral reinforcement, and structural reform. These strategies must be grounded in the realities of frontline workers to be effective and sustainable.

Like any human endeavor, this study has its limits, which may be related to the small sample size. A larger sample would have contributed to making the significance of the probabilities more expressive. 35.8% of respondents who did not have a specific role could have likely influenced our results. Additionally, the measurement of attitudes using a Likert scale might be subject to social desirability biases. Respondents may have overreported positive attitudes toward surveillance due to perceived expectations from supervisors or researchers, which could artificially inflate the level of reported engagement. While anonymity was assured, the presence of data collectors affiliated with the health system might have reinforced this effect.

Cross-sectional design also limits the ability to infer causality. While associations were observed between attitudes, training, and case reporting, the direction of these relationships cannot be firmly established. It is possible that individuals who reported cases were more likely to retrospectively express stronger attitudes, rather than the reverse.

Furthermore, certain contextual or institutional variables were not captured in the survey, such as workload, staffing levels, management support, or community pressures, which might moderate the relationship between attitudes and surveillance behavior. Their omission could have resulted in residual confounding.

Finally, the study was limited to public healthcare facilities in one region of Burkina Faso, which may restrict the generalizability of the findings. Private sector actors and community-based workers, who also play a role in surveillance, were not included. Nevertheless, the relatively high response rate (96%) and diversity of professional profiles enhance the internal validity and representativeness of the results within the studied region.