A Cross-Sectional Survey of Knowledge, Attitudes, and Practices Toward Mpox Among One Health Stakeholders in Nigeria

Simple Summary

Monkeypox (Mpox) is a disease that spreads between animals and humans and has become a growing concern around the world. In Nigeria, where people often live and work closely with animals and the environment, it is important to understand how professionals in health, veterinary, and environmental fields think about and respond to Mpox. This study asked these professionals questions about their knowledge, attitudes, and behaviors to see how prepared they are to deal with the disease. We found that while many had a good understanding and positive attitude, fewer were putting safe practices into action. Females also showed slightly lower confidence in their attitudes toward the disease compared to men. These results help us see where training and awareness need to improve. By strengthening knowledge and practical responses among health workers, Nigeria can be better prepared to prevent and control Mpox and other similar diseases.

Abstract

Mpox has re-emerged as a global public health threat, particularly in endemic regions such as Nigeria, where human, animal, and environmental health sectors intersect. To inform surveillance and control strategies, this study assessed the knowledge, attitudes, and practices (KAP) toward Mpox among One Health stakeholders in Nigeria. A cross-sectional survey was conducted among 492 participants from human, veterinary, and environmental health sectors using a structured questionnaire. Descriptive statistics, ordinal logistic regression, and margins analysis were used to evaluate levels and predictors of KAP. Results showed that 33.7% of respondents had low knowledge, 43.5% moderate, and 22.8% high. While 62.6% demonstrated high attitude scores, only 48.2% reported moderate preventive practices. Gender was significantly associated with attitudes, with females having lower odds of expressing higher attitudes than males (OR = 0.70, 95% CI: 0.49–1.00, p = 0.052), and margins analysis revealed a predicted probability of high attitude at 56% for females and 64% for males. Multivariable modeling for practice was not pursued because model fit did not improve compared to univariable results, and sparse data led to unstable estimates, thus offering no added explanatory power. These findings underscore persistent knowledge gaps and gender-related disparities that may hinder effective Mpox response. Targeted risk communication and capacity building are recommended to strengthen One Health preparedness in Nigeria.

1. Introduction

Mpox (formerly known as monkeypox) [1], a zoonotic disease caused by the Mpox virus (MPXV), has re-emerged as a significant public health concern in recent decades, particularly in Central and West Africa. Initially identified in laboratory monkeys in 1958 and later in humans in the Democratic Republic of Congo in 1970, Mpox has since become one of the most prominent orthopoxvirus infections in humans following the eradication of smallpox [2,3]. The disease presents with fever, rash, and lymphadenopathy, and while it is generally self-limiting, severe outcomes, including death, can occur, especially in vulnerable populations such as children and immunocompromised individuals [4,5,6,7]. Recent global reports confirm the ongoing spread of Mpox beyond Africa, particularly following the 2022–2023 outbreaks across Europe and the Americas [1,8]. Over the years, Nigeria has witnessed repeated outbreaks of Mpox, including a large resurgence beginning in 2017 that marked the re-emergence of the virus after nearly four decades of silence. This resurgence, after nearly four decades of silence since 1978, has been documented in outbreak investigations [8,9,10], and it underscores the urgent need to strengthen surveillance, public awareness, and intersectoral collaboration in the country.

While Mpox has historically been described as a zoonosis with sporadic animal-to-human spillover and limited secondary spread, the 2022 global outbreak was characterized by efficient human-to-human transmission, frequently linked to close and intimate contact, including sexual contact within defined social and sexual networks. This epidemiologic shift underscores the need to appraise knowledge, attitudes, and practices (KAP) not only regarding zoonotic exposure but also human-to-human transmission dynamics and risk communication for key populations [1,8,9,10,11].

The dynamics of Mpox transmission are complex and deeply embedded within the human–animal–environment interface, making it a textbook example of a disease best addressed using a One Health approach—a concept that recognizes the interconnectedness of human health, animal health, and the environment [12,13,14,15]. Human infections are often linked to contact with infected animals such as rodents or primates, either through hunting, handling, or consumption, especially in regions with high dependence on bushmeat. Furthermore, environmental changes such as deforestation, urban encroachment, and increased human–wildlife interactions have contributed to the spillover of Mpox virus from animals to humans [16,17,18]. As such, a collaborative response that integrates human health, veterinary, and environmental expertise is crucial to understanding and controlling Mpox outbreaks effectively.

The One Health framework has gained increasing attention globally as a comprehensive strategy to prevent, detect, and respond to emerging and re-emerging zoonotic diseases. It promotes transdisciplinary collaboration among health professionals from human medicine, veterinary medicine, public health, wildlife ecology, and environmental science [19,20]. In Nigeria, while the One Health approach has been endorsed at policy levels, operational implementation remains uneven. Challenges such as limited cross-sectoral communication, institutional silos, inadequate workforce training, and low community engagement continue to hinder effective integration [21,22,23,24,25].

The success of One Health implementation depends significantly on the knowledge, attitudes, and practices (KAP) of professionals and stakeholders who constitute the One Health workforce.

Despite Nigeria being one of the epicenters of Mpox resurgence, there is limited empirical evidence on KAP toward Mpox among One Health stakeholders in Nigeria [26]. Understanding the level of awareness, perception, and practical engagement of stakeholders in surveillance, reporting, and response to Mpox is essential for guiding future outbreak preparedness and One Health capacity development. Studies in other zoonotic contexts have shown that gaps in knowledge and poor attitudes among frontline workers can delay disease detection, increase risk of spread, and undermine control efforts [27,28]. Therefore, assessing the current KAP landscape related to Mpox among stakeholders across human, animal, and environmental health sectors in Nigeria is both timely and necessary.

This study was designed to fill this gap by conducting a cross-sectional survey to assess the knowledge, attitudes, and practices toward Mpox among One Health stakeholders in Nigeria. By identifying patterns and predictors of KAP and exploring the interrelationship between demographic factors and disease-related perceptions, the study aims to generate actionable insights to inform policy, training, and communication strategies for strengthening Mpox preparedness and response under the One Health umbrella.

2. Materials and Methods

2.1. Study Design and Setting

This study employed a cross-sectional survey design to assess the knowledge, attitudes, and practices (KAP) of stakeholders in the One Health workforce across various geopolitical zones in Nigeria. To maximize geographic representation and ease of participation, an electronic questionnaire was administered online, targeting professionals engaged in human, animal, and environmental health disciplines.

2.2. Study Population

The study population consisted of professionals working within the One Health framework, including medical doctors, registered nurses and midwives, environmental scientists, health information managers, laboratory scientists, microbiologists, paraveterinarians (Paravets), and veterinary doctors. Participants were eligible if they were Nigerian nationals, were professionally involved in relevant One Health domains, and consented to participate in the survey. Respondents whose professional categories did not align with core One Health disciplines, such as educationists, civil engineers, and journalists, were excluded from the analysis. Responses that were incomplete or inconsistent, as well as submissions from non-Nigerian nationals (e.g., one Ethiopian respondent), were also excluded during the data cleaning phase. Occupational titles were aggregated during data processing to enhance analytical coherence. For instance, various forms of registered nurse titles were grouped under “Registered nurse/midwife,” while laboratory technicians were combined as “Laboratory scientists.” Individuals with backgrounds in animal science or agricultural science were categorized under “Paravets.” The variable “Area of specialization” was excluded from analysis due to redundancy with the “Profession” variable.

2.3. Sample Size Determination

A total of 492 participants were included in the study. Participants were recruited using a convenience sampling strategy, leveraging professional networks across human, animal, and environmental health sectors. The required minimum sample size was determined using a standard formula for estimating proportions in cross-sectional studies, incorporating a 95% confidence level, a 5% margin of error, and an assumed population proportion of 50%, which maximizes sample size [29]. This yielded a minimum sample size of 384 participants. The sample size for this cross-sectional study was estimated using the formula for calculating sample size for proportions:

$$n={\displaystyle \frac{Z^2·p·\left(1-p\right)}{d^2}}$$

where $n$ represents the required minimum sample size; $Z$ is the Z-score corresponding to the desired confidence level, with a value of 1.96 for a 95% confidence level; $p$ denotes the estimated proportion of the population with the characteristic of interest, set at 0.5 to maximize the sample size in the absence of a prior estimate; and $d$ is the desired margin of error (precision), which was set at 0.05.

Substituting the values:

$$n={\displaystyle \frac{{\left(1.96\right)}^2·0.5·\left(1-0.5\right)}{{\left(0.05\right)}^2}}={\displaystyle \frac{3.8416·0.25}{0.0025}}={\displaystyle \frac{0.9604}{0.0025}}=384.16$$

Thus, a minimum of 384 participants was required. To enhance precision and account for possible non-responses, the final number of participants included in the analysis was 492.

2.4. Data Collection Instrument

Data were collected using a structured, self-administered questionnaire developed based on a review of relevant literature and expert input (see Appendix A). The questionnaire included three main sections: knowledge, attitude, and practice. The knowledge section consisted of 20 items assessed using nominal scale values for multiple-choice or single-response survey items. These included both multiple-choice options with combination responses and multi-select combination-type response options, where respondents could choose more than one answer to reflect their level of knowledge on specific topics. To avoid ambiguity, the knowledge item asking ‘Where is Mpox prevalent?’ (k2) was explicitly coded to reflect recognition of the historically endemic regions of Central and West Africa, which was the established epidemiological pattern prior to the 2022 global outbreak. While this coding captures pre-2022 endemicity, we acknowledge that the phrasing may not fully reflect the post-2022 global distribution of Mpox. In addition, we clarified the scoring rubric for items k8–k20. Item k8 (‘travelers from Central and West Africa were the sources of imported cases in Europe and America during the 2022 outbreak’) was reverse-scored to align with contemporary transmission dynamics, where sustained human-to-human transmission predominated outside Africa. Item k17 (‘antiviral drugs are required’) was treated as non-scored due to its context-dependent clinical applicability and is presented descriptively only. Items k9–k16 and k18–k20 were scored as correct or incorrect based on clinical and epidemiological evidence. The attitude section comprised 10 items measured on a 5-point Likert scale (Agree, Disagree, Not sure, Strongly agree, or Strongly disagree) and as ordinal categorical variables or ordinal response options (No, Not sure, and Yes). The practice section contained 5 items evaluating the frequency of specific actions related to One Health engagement. The questionnaire underwent expert review for face and content validity and was pilot-tested among 30 professionals to assess clarity and reliability. Adjustments were made based on feedback to improve item phrasing and flow. Reliability of the knowledge and attitude scales was assessed using Cronbach’s alpha (α = 0.81 and α = 0.79, respectively), indicating good internal consistency.

2.5. Variables and Measures

The independent variables in this study included sociodemographic and professional characteristics, while the dependent variables were the composite knowledge, attitude, and practice scores. Each KAP domain was evaluated by summing the scores of respective items, ensuring that all items were scored in the same direction. Negatively worded items were reverse-coded, such as item k8 in the knowledge section. The variable “Qualification (Educational qualification)” was categorized as Certificate, Diploma, Fellowship diploma, First degree, Higher National Diploma (HND), Masters, and PhD. For clarity to an international audience, brief explanations were provided: in the Nigerian system, a Diploma is a post-secondary academic award that typically precedes a first degree, while the Higher National Diploma (HND) is a professional qualification obtained after a minimum of four years of tertiary-level study, broadly comparable to a bachelor’s degree in many contexts. The summed scores were then categorized using tertiles to generate ordinal outcome variables: knowledge_cat, attitude_cat, and practice_cat.

2.6. Data Management and Statistical Analysis

All statistical analyses were conducted using STATA/IC version 15.0 [30] and R software (version 4.5.1) [31]. Initial data management included validation of responses, exclusion of ineligible participants, and recoding of string variables into numeric formats suitable for statistical computation.

Descriptive statistics—comprising frequencies and percentages—were used to summarize participants’ sociodemographic characteristics and their distribution across the Knowledge, Attitude, and Practice (KAP) domains.

To assess the strength and direction of associations among the composite KAP scores, Spearman’s rank correlation coefficient was applied. The inferential analysis was then anchored on ordinal logistic regression modeling, given the ordinal nature of the three outcome variables [32,33]. For each of the KAP domains, univariable and multivariable ordinal logistic regression models were constructed to investigate predictors of higher category responses. Categorical predictors were entered using the most frequent category as the reference group. Professional categories with extremely sparse representation (e.g., microbiologists, n = 1) were retained in regression output for transparency, but their estimates were flagged as unstable and interpreted only descriptively.

The cumulative logit model was specified as follows:

$$log\left({\displaystyle \frac{P(Y\le j)}{P(Y>j)}}\right)={\theta }_j\left({\beta }_1{X}_1+{\beta }_2{X}_2+\dots +{\beta }_k{X}_k\right)$$

where $Y$ is the ordinal outcome (knowledge, attitude, or practice), $j$ is the response category, $X_1$ to $X_k$ are predictor variables, and ${\theta }_j$ represents the category-specific thresholds. The proportional odds assumption—requiring constant coefficients across all cumulative logits—was tested using the Brant test in STATA. Where this assumption was violated, partial proportional odds models were fitted using the gologit2 command.

Model fit was evaluated using log-likelihood, Akaike Information Criterion (AIC), and Bayesian Information Criterion (BIC) [34], defined as:

$$AIC=-2·\mathit{log}L+2k$$

$$BIC=-2·log L+k·\mathrm{l}\mathrm{o}\mathrm{g}\left(n\right)$$

where $\mathit{log}L$ is the log-likelihood of the fitted model, $k$ is the number of parameters, and $n$ is the sample size.

To aid interpretation of statistically significant associations, predictive margins were estimated to determine how the probability of being in higher outcome categories changed across values of key predictors. These probabilities were derived from the formula:

$$P\left(Y=j | X\right)=logistic({\theta }_j-\eta )-logistic\left({\theta }_{j-1}-\eta \right)$$

where $\eta ={\beta }_1{X}_1+{\beta }_2{X}_2+\dots +{\beta }_k{X}_{k,}$ and $logistic\left(x\right)={\displaystyle \frac{1}{1+e^{-x}}}$.

Model results were reported as adjusted odds ratios (AORs) with corresponding 95% confidence intervals (CIs) and p-values. Odds ratios were computed as:

$$OR=e^{\beta }$$

and their confidence intervals as:

$${CI}_{95\%}=\left(e^{\beta -1.96·SE}, e^{\beta +\mathrm{1,96}·SE}\right)$$

These values were presented in summary tables for each KAP domain to highlight statistically significant predictors and key patterns in the data.

2.7. Ethical Considerations

The study protocol was reviewed and approved by the University Research and Ethics Committee, Usmanu Danfodiyo University, Sokoto with reference number UDUS/UREC/2024/009. Participation in the study was voluntary, and informed consent was obtained electronically from all participants prior to completing the questionnaire. All responses were anonymized and treated with strict confidentiality during data handling and analysis.

3. Results

3.1. Socio-Demographic Characteristics of Respondents

A total of 492 respondents participated in the study assessing the knowledge, attitude, and practice of the One Health workforce in Nigeria during the Mpox outbreak. As shown in

Table 1. Socio-demographic characteristics of respondents (N = 492).

Variable	Level	Freq.	Percent
Gender	Female	274	55.69
	Male	218	44.31
	Total	492	100
Age	20–29	161	32.72
	30–39	179	36.38
	40–49	132	26.83
	50–59	20	4.07
	Total	492	100
State of residence	Abia	4	0.81
	Abuja	28	5.69
	Adamawa	9	1.83
	Akwa Ibom	12	2.44
	Anambra	6	1.22
	Bauchi	4	0.81
	Bayelsa	1	0.2
	Benue	4	0.81
	Borno	8	1.63
	Cross River	7	1.42
	Delta	14	2.85
	Ebonyi	7	1.42
	Edo	4	0.81
	Ekiti	6	1.22
	Enugu	22	4.47
	Gombe	15	3.05
	Imo	15	3.05
	Jigawa	5	1.02
	Kaduna	10	2.03
	Kano	25	5.08
	Katsina	12	2.44
	Kebbi	60	12.2
	Kogi	8	1.63
	Kwara	7	1.42
	Lagos	20	4.07
	Nasarawa	9	1.83
	Niger	13	2.64
	Ogun	5	1.02
	Oyo	4	0.81
	Plateau	7	1.42
	Rivers	8	1.63
	Sokoto	95	19.31
	Taraba	5	1.02
	Yobe	5	1.02
	Zamfara	28	5.69
	Total	492	100
Marital status	Divorced	2	0.41
	Married	324	65.85
	Single	164	33.33
	Widow	2	0.41
	Total	492	100
Qualification	Certificate	65	13.21
	Diploma	120	24.39
	Fellowship diploma	2	0.41
	First degree	167	33.94
	HND	58	11.79
	Masters	56	11.38
	PhD	24	4.88
	Total	492	100
Profession	Environmental scientist	1	0.2
	Health information management	1	0.2
	Laboratory scientist	41	8.33
	Medical doctor	19	3.86
	Microbiologist	1	0.2
	Paravet	3	0.61
	Registered nurse/midwife	302	61.38
	Veterinary doctor	124	25.2
	Total	492	100

, the majority was female (55.69%), while males constituted 44.31%. The age distribution revealed that most respondents were between 30 and 39 years (36.38%), followed by those aged 20–29 years (32.72%), and 40–49 years (26.83%). Only 4.07% were aged 50 years or older. In terms of marital status, the majority were married (65.85%), followed by singles (33.33%), while less than 1% were widowed or divorced.

Regarding educational attainment, respondents had a wide range of qualifications: 33.94% held a first degree, 24.39% had a diploma, 13.21% held a certificate, and 11.79% had a Higher National Diploma (HND). Only 4.88% possessed a PhD, and 11.38% held a master’s degree. The professional background of respondents was diverse, with the largest group being registered nurses and midwives (61.38%), followed by veterinary doctors (25.2%) and laboratory scientists (8.33%).

The distribution of respondents by state of residence, illustrated in Figure 1, showed that the highest number of participants were from Sokoto (19.31%) and Kebbi (12.2%) states, followed by Zamfara and Abuja, each with 5.69%. Other states with notable participation included Kano (5.08%), Lagos (4.07%), and Enugu (4.47%); while several states such as Abia, Bauchi, Bayelsa, Edo, and Oyo had minimal representation (less than 5%).

3.2. Level of Knowledge, Attitude, and Practice Among One Health Workforce During the Mpox Outbreak

The findings from the study assessing the level of knowledge, attitude, and practice (KAP) of the One Health workforce during the Mpox outbreak in Nigeria are presented in

Table 2. Distribution of knowledge, attitude, and practice scores among One Health workforce respondents on Mpox outbreak, Nigeria, 2024 (N = 492).

Item	Level	Freq.	Percent
Knowledge	0	1	0.20
	2	1	0.20
	3	1	0.20
	4	2	0.41
	5	2	0.41
	6	4	0.81
	7	8	1.63
	8	11	2.24
	9	37	7.52
	10	46	9.35
	11	59	11.99
	12	64	13.01
	13	69	14.02
	14	63	12.80
	15	59	11.99
	16	33	6.71
	17	20	4.07
	18	6	1.22
	19	5	1.02
	20	1	0.20
	Total	492	100
Attitude	0	1	0.20
	1	1	0.20
	2	1	0.20
	3	3	0.61
	4	5	1.02
	5	11	2.24
	6	14	2.85
	7	51	10.37
	8	111	22.56
	9	185	37.60
	10	109	22.15
	Total	492	100
Practice	0	17	3.46
	1	48	9.76
	2	63	12.80
	3	119	24.19
	4	160	32.52
	5	85	17.28
	Total	492	100

and Figure 2. These results highlight distinct patterns across the three domains assessed.

In terms of knowledge, scores ranged from 0 to 20. The distribution reveals that the majority of respondents demonstrated moderate to high knowledge about Mpox. Scores of 13, 12, 14, 15, and 11 were the most frequently recorded, with proportions of 14.02%, 13.01%, 12.80%, 11.99%, and 11.99%, respectively. Conversely, only a few respondents had very low knowledge, with 0.20% each scoring 0, 2, or 3. These results suggest that the One Health workforce possessed a sound understanding of Mpox, likely reflecting previous training or experience with zoonotic disease outbreaks.

Attitude scores, which ranged from 0 to 10, revealed a generally positive outlook among respondents. The majority scored 9 (37.60%), 8 (22.56%), and 10 (22.15%), indicating a strong readiness and favorable disposition toward Mpox response activities. Only a small fraction of the workforce exhibited low attitude scores, with just 0.20% scoring 0, 1, or 2, and a cumulative 3.67% scoring between 0 and 4. This widespread positive attitude among respondents suggests a high level of commitment to public health responsibilities during the outbreak.

Practice scores, assessed on a scale of 0 to 5, displayed a slightly different pattern. The highest proportion of respondents (32.52%) scored 4; while 24.19% scored 3 and 17.28% scored 5, indicating a moderate to high level of appropriate practice. However, a notable proportion of respondents still fell into the lower practice categories, with 3.46% scoring 0 and 9.76% scoring 1. This distribution suggests that while knowledge and attitudes were generally high, actual practices lagged slightly behind. The gap between attitude and practice may reflect systemic challenges such as limited access to protective equipment, institutional constraints, or insufficient operational support in the field.

The findings depicted in Table 2 and illustrated in Figure 2 demonstrate a knowledgeable and motivated One Health workforce with largely positive attitudes but also underscore the need to improve practical application and support systems to enhance outbreak response effectiveness. These results reinforce the importance of continuous training, provision of adequate resources, and structural capacity-building for frontline responders during emerging infectious disease outbreaks such as Mpox.

The distribution of respondents’ levels of knowledge, attitude, and practice regarding Mpox prevention and response among the One Health workforce is presented in

Table 3. Categorical distribution of knowledge, attitude, and practice levels among One Health workforce respondents on Mpox prevention and response in Nigeria.

Item	Level (Range)	Freq.	Percent
Knowledge	Low Knowledge (0–9)	67	67.00
	Moderate Knowledge (10–14)	301	61.18
	High Knowledge (15–20)	124	25.20
	Total	492	100
Attitude	Low Attitude (0–5)	22	4.47
	Moderate Attitude (6–8)	176	35.77
	High Attitude (9–10)	294	59.76
	Total	492	100
Practice	Low Practice (0–2)	128	26.02
	Moderate Practice (3–4)	279	56.71
	High Practice (5)	85	17.28
	Total	492	100

. The majority of respondents (61.18%) demonstrated moderate knowledge scores (10–14), while 25.20% exhibited high knowledge (15–20), and a smaller proportion (13.62%) fell into the low knowledge category (0–9). In terms of attitude, a high proportion (59.76%) showed positive attitudes (scores 9–10), with 35.77% having moderate attitudes (6–8), and only 4.47% scoring low attitudes (0–5), reflecting a generally favorable disposition toward Mpox response. Regarding practice, over half of the participants (56.71%) had moderate levels of practice (scores 3–4), while 26.02% were classified as having low practice levels (0–2). Only 17.28% of respondents demonstrated high practice (score of 5), indicating a gap between knowledge and its translation into preventive behaviors.

3.3. Distribution of Participants’ Responses to Knowledge, Attitude, and Practice Items on Mpox

Table 4. Distribution of responses to knowledge items on Mpox.

Question/Item	Answer/Response	Freq.	Percent
k1 (When was the first time you heard of Mpox?)	2022/2023 outbreak	160	32.52
	Never heard of it	20	4.07
	Previous outbreaks	86	17.48
	School training	217	44.11
	None of the above	9	1.83
	Total	492	100
k2 (Where is Mpox prevalent?)	Central and West Africa	342	69.51
	Europe	47	9.55
	North Africa	56	11.38
	North America	26	5.28
	None of the above	21	4.27
	Total	492	100
k3 (Mpox is a viral disease)	Bacterial disease	22	4.47
	Fungal disease	7	1.42
	Parasitic disease	8	1.63
	Viral disease	444	90.24
	None of the above	11	2.24
	Total	492	100
k4 (How is Mpox transmitted?)	(a) Direct contact with infected animals	89	18.09
	(b) Consumption of monkey and other wild animals	24	4.88
	(c) Contact with contaminate lesions	12	2.44
	(d) a, b, and c	235	47.76
	(e) a and c only	132	26.83
	Total	492	100
k5 (Could Mpox be sustained in a human-to-human transmission?)	No	21	4.27
	Not sure	85	17.28
	Yes	386	78.46
	Total	492	100
k6 (Could Mpox be transmitted through infected monkey bite?)	No	44	8.94
	Not sure	107	21.75
	Yes	341	69.31
	Total	492	100
k7 (Are there human Mpox cases in Nigeria?)	No	32	6.5
	Not sure	123	25
	Yes	337	68.5
	Total	492	100
k8 (In the 2022/2023 global Mpox outbreak, travelers from central and western Africa are the sources of imported cases in Europe and America)	Agree	186	37.8
	Disagree	20	4.07
	Not sure	143	29.07
	Strongly agree	135	27.44
	Strongly disagree	8	1.63
	Total	492	100
k9 (Mpox symptoms cannot be distinguished from smallpox symptoms)	Agree	184	37.4
	Disagree	129	26.22
	Not sure	94	19.11
	Strongly agree	64	13.01
	Strongly disagree	21	4.27
	Total	492	100
k10 (Early signs of human Mpox infection include flu-like)	Agree	243	49.39
	Disagree	36	7.32
	Not sure	84	17.07
	Strongly agree	129	26.22
	Total	492	100
k11 (Body rashes are one of the symptoms of human Mpox infection)	Agree	215	43.7
	Disagree	10	2.03
	Not sure	27	5.49
	Strongly agree	238	48.37
	Strongly disagree	2	0.41
	Total	492	100
k12 (Papules on the skin are among the symptoms of human Mpox)	Agree	231	46.95
	Disagree	10	2.03
	Not sure	37	7.52
	Strongly agree	212	43.09
	Strongly disagree	2	0.41
	Total	492	100
k13 (Vesicles on the skin are among the symptoms of Mpox)	Agree	230	46.75
	Disagree	22	4.47
	Not sure	79	16.06
	Strongly agree	160	32.52
	Strongly disagree	1	0.2
	Total	492	100
k14 (Pustules on the skin are among the symptoms of human Mpox)	Agree	217	44.11
	Disagree	16	3.25
	Not sure	102	20.73
	Strongly agree	154	31.3
	Strongly disagree	3	0.61
	Total	492	100
k15 (Diarrhea is one of the signs of human Mpox infection)	Agree	145	29.47
	Disagree	104	21.14
	Not sure	180	36.59
	Strongly agree	50	10.16
	Strongly disagree	13	2.64
	Total	492	100
k16 (Lymphadenopathy is among the clinical signs that could differentiate between Mpox and smallpox cases)	Agree	222	45.12
	Disagree	26	5.28
	Not sure	91	18.5
	Strongly agree	151	30.69
	Strongly disagree	2	0.41
	Total	492	100
k17 (In the management of human Mpox cases, antiviral drugs are required)	Agree	234	47.56
	Disagree	38	7.72
	Not sure	55	11.18
	Strongly agree	158	32.11
	Strongly disagree	7	1.42
	Total	492	100
k18 (Specific vaccine for Mpox is available)	Agree	124	25.2
	Disagree	38	7.72
	Not sure	211	42.89
	Strongly agree	103	20.93
	Strongly disagree	16	3.25
	Total	492	100
k19 (People immunized against chickenpox can be protected against Mpox)	Agree	121	24.59
	Disagree	104	21.14
	Not sure	173	35.16
	Strongly agree	57	11.59
	Strongly disagree	37	7.52
	Total	492	100
k20 (Men who have sex with men are among the high-risk group for infection with Mpox)	Agree	157	31.91
	Disagree	54	10.98
	Not sure	139	28.25
	Strongly agree	128	26.02
	Strongly disagree	14	2.85
	Total	492	100

Scoring notes: Correct answers were defined as follows: k8 reverse-scored to reflect 2022 global transmission dynamics (Disagree/Strongly Disagree = correct); k9 false (Disagree/Strongly Disagree = correct); k10–k14 true (Agree/Strongly Agree = correct); k15 true (Agree/Strongly Agree = correct); k16 true (Agree/Strongly Agree = correct); k17 excluded from scoring due to context-dependence, presented descriptively only; k18 true; k19 false (Disagree/Strongly Disagree = correct); k20 true (Agree/Strongly Agree = correct). References: [22,23,24].

presents the distribution of responses to twenty coded knowledge questions (k1 to k20) that assessed participants’ awareness and understanding of Mpox, using categorical variables as defined in the data dictionary (see Appendix A). The responses reflect varying degrees of knowledge, misinformation, and uncertainty among the respondents (n = 492).

Regarding the first time respondents heard of Mpox (k1), the majority (44.11%) indicated they were informed through school training, followed by 32.52% who learned about it during the 2022/2023 outbreak, while a smaller fraction (17.48%) had heard about it during previous outbreaks. Only 4.07% had never heard of Mpox. When asked about Mpox prevalence (k2), most participants correctly identified Central and West Africa (69.51%) as endemic regions, though some wrongly cited Europe (9.55%), North Africa (11.38%), or North America (5.28%).

On the nature of the disease (k3), a high percentage (90.24%) correctly identified Mpox as a viral disease, while small percentages mistook it for bacterial (4.47%), fungal (1.42%), or parasitic (1.63%) diseases. For modes of transmission (k4), nearly half (47.76%) correctly selected all relevant routes (direct contact, consumption of wild animals, and contact with contaminated lesions), while another 26.83% chose the more accurate but slightly restrictive option of only direct contact and lesion contact. Others selected incomplete or incorrect options.

Participants’ knowledge of human-to-human transmission (k5) was encouraging, with 78.46% answering “Yes.” Similarly, 69.31% recognized that infected monkey bites could transmit Mpox (k6), though 21.75% were unsure. Awareness of human Mpox cases in Nigeria (k7) was fairly high (68.5%), though 25% were uncertain. On the 2022/2023 outbreak’s origins (k8), 37.8% agreed and 27.44% strongly agreed that travelers from Central and West Africa were the source of imported cases, although 29.07% were not sure.

Opinions on the similarity of Mpox and smallpox symptoms (k9) were divided, with 37.4% agreeing and 13.01% strongly agreeing that they are indistinguishable, though 26.22% disagreed. Regarding early signs such as flu-like symptoms (k10), a large majority either agreed (49.39%) or strongly agreed (26.22%). High agreement was also observed for specific clinical signs: body rashes (k11) were acknowledged by 92.07% of respondents (combined agree and strongly agree), papules (k12) by 90.04%, vesicles (k13) by 79.27%, and pustules (k14) by 75.41%. However, fewer respondents recognized diarrhea (k15) as a sign, with 29.47% agreeing and 36.59% unsure.

When asked about differentiating Mpox from smallpox by lymphadenopathy (k16), 75.81% either agreed or strongly agreed. A total of 79.67% believed that antiviral drugs are required for Mpox management (k17), and 46.13% agreed or strongly agreed that a specific vaccine for Mpox is available (k18), although 42.89% were uncertain. The belief that chickenpox immunization protects against Mpox (k19) was held by 36.18%, but 35.16% were unsure and 21.14% disagreed. Finally, 57.93% either agreed or strongly agreed that men who have sex with men are at higher risk of Mpox infection (k20), while 28.25% were unsure.

In this study, the distribution of responses to Mpox knowledge items reveals that participants possess a generally good understanding of Mpox’s virology, transmission routes, clinical signs, and public health relevance. However, uncertainties remain concerning specific transmission modes, differential diagnosis, and vaccine availability, highlighting the need for targeted awareness and education campaigns.

Table 5. Distribution of responses to attitude items on Mpox.

Question/Item	Answer/Response	Freq.	Percent
a1 (The Nigerian medical and animal healthcare workers play a critical role in the control of Mpox infection in the country)	Agree	193	39.23
	Disagree	6	1.22
	Not sure	32	6.5
	Strongly agree	258	52.44
	Strongly disagree	3	0.61
	Total	492	100
a2 (The Nigeria’s NCDC, Ministry of health, Ministry of Environment and Federal Livestock Department can locally control the Mpox outbreak in Nigeria if they work together in a “One health” approach)	Agree	148	30.08
	Disagree	2	0.41
	Not sure	31	6.3
	Strongly agree	310	63.01
	Strongly disagree	1	0.2
	Total	492	100
a3 (I think that there are currently enough prevention and control measures for Mpox in among the Nigerian medical and animal healthcare workers)	Agree	140	28.46
	Disagree	113	22.97
	Neutral	127	25.81
	Strongly agree	76	15.45
	Strongly disagree	36	7.32
	Total	492	100
a4 (I have bad feelings toward the Mpox virus that it might become a worldwide pandemic)	Agree	229	46.54
	Disagree	69	14.02
	Not sure	105	21.34
	Strongly agree	84	17.07
	Strongly disagree	5	1.02
	Total	492	100
a5 (I think that Mpox can add a new burden on the Nigerian medical and animal healthcare systems)	Agree	247	50.2
	Disagree	24	4.88
	Not sure	50	10.16
	Strongly agree	170	34.55
	Strongly disagree	1	0.2
	Total	492	100
a6 (I think that mass media coverage of Mpox may influence its worldwide prevention)	No	19	3.86
	Not sure	14	2.85
	Yes	459	93.29
	Total	492	100
a7 (I am interested in learning more about Mpox)	No	9	1.83
	Not sure	5	1.02
	Yes	478	97.15
	Total	492	100
a8 (I am interested to learn more about the epidemiology of the new emerging diseases)	No	8	1.63
	Not sure	8	1.63
	Yes	476	96.75
	Total	492	100
a9 (I am interested in learning more about epidemiology of reemerging diseases)	No	13	2.64
	Not sure	8	1.63
	Yes	471	95.73
	Total	492	100
a10 (I think that it is dangerous to travel to the countries where the current Mpox outbreak is ongoing)	No	29	5.89
	Not sure	31	6.3
	Yes	432	87.8
	Total	492	100

presents the distribution of responses to ten coded attitude questions (a1 to a10), which assess the perceptions, emotional responses, and interests of respondents regarding Mpox and related public health concerns, using five-point and three-point Likert-type scales as defined in the data dictionary (see Appendix A).

For item a1, which asked whether Nigerian medical and animal healthcare workers play a critical role in controlling Mpox, 45.8% of respondents agreed and 33.2% strongly agreed, indicating that nearly four out of five respondents (79.0%) recognized the importance of these professionals. Only 3.4% disagreed, 1.3% strongly disagreed, and 16.3% were not sure.

Regarding item a2, which assessed beliefs in the capacity of national health institutions to control Mpox through a One Health approach, 47.6% agreed and 34.0% strongly agreed—together making up 81.6% of respondents who supported the idea. A small portion disagreed (2.6%), strongly disagreed (1.6%), or were unsure (14.2%).

On item a3, addressing whether there are currently enough Mpox prevention and control measures among Nigerian health workers, only 29.5% agreed and 11.3% strongly agreed. In contrast, 28.7% disagreed and 6.3% strongly disagreed, while 24.2% were not sure. These figures reflect a mixed perception, with a notable proportion of respondents expressing doubt or uncertainty about existing measures.

For item a4, which gauged emotional concern about Mpox becoming a worldwide pandemic, 38.7% agreed and 29.2% strongly agreed, while 20.8% were not sure. A smaller number disagreed (8.4%) or strongly disagreed (2.9%). Overall, 67.9% of respondents reported a negative emotional response to the virus’s potential global spread.

In item a5, 42.1% agreed and 40.3% strongly agreed that Mpox could add a new burden to the Nigerian medical and animal healthcare systems. Only 2.9% disagreed, 1.8% strongly disagreed, and 12.9% were not sure. These responses suggest strong awareness of the strain Mpox could place on already challenged systems, with 82.4% expressing concern.

Item a6 explored the influence of mass media on Mpox prevention. Here, 77.4% answered “Yes,” acknowledging the role of media in shaping global prevention efforts, while 16.6% were not sure and 6.0% said “No.”

Interest in further learning about Mpox (item a7) was also high, with 86.3% of respondents indicating “Yes,” while 10.0% were not sure and only 3.7% said “No.” A similar trend appeared in item a8, where 88.9% said “Yes” to wanting to learn more about emerging diseases, with 7.1% not sure and 4.0% responding “No.” In item a9, 88.4% expressed interest in learning more about reemerging diseases, 7.1% were not sure, and only 4.5% said “No.”

Lastly, item a10 evaluated perceived risk associated with international travel to Mpox-affected countries. A majority (81.8%) of respondents agreed that it is dangerous, while 13.2% were not sure and 5.0% disagreed.

In this study, the distribution of responses to Mpox attitude items reveals a broadly positive attitude among respondents toward the importance of health professionals and intersectoral collaboration in controlling Mpox, with 81.6% supporting the One Health approach. Although attitudes toward the adequacy of current control measures were mixed (only 40.8% affirming their sufficiency), there was high concern about Mpox potentially becoming a pandemic (67.9%) and overwhelming health systems (82.4%). The majority of respondents also recognized the influence of mass media (77.4%) and expressed strong interest in gaining more knowledge about Mpox (86.3%), emerging diseases (88.9%), and reemerging diseases (88.4%). These findings underscore the importance of sustained public health communication, cross-sectoral collaboration, and educational outreach in managing Mpox and similar infectious disease threats.

Table 6. Distribution of responses to practice items on Mpox.

Question/Item	Answer/Response	Freq.	Percent
p1 (How do you protect yourself from contracting Mpox from suspected cases in humans/animals?)	(a) Create awareness about Mpox for the community	38	7.72
	(b) Avoid contact with persons/animals suspected/confirmed to be infected with Mpox	43	8.74
	(c) Wash/sterilize hands after contact with infected/dead animals or humans	14	2.85
	(e) All of the above	397	80.69
	Total	492	100
p2 (What samples do you collect for the laboratory diagnosis of Mpox?)	(a) Skin lesion material (swabs of lesion surface/exudate)	128	26.02
	(b) Roofs from more than one lesion	10	2.03
	(c) Lesion crust	5	1.02
	(d) Saliva	13	2.64
	(e) All except d	336	68.29
	Total	492	100
p3 (Where is Mpox reference laboratory located in Nigeria?)	(a) Lagos	219	44.51
	(b) Abuja	217	44.11
	(c) Kaduna	24	4.88
	(d) Kano	13	2.64
	(e) Enugu	19	3.86
	Total	492	100
p4 (How do you handle suspected human/animal Mpox cases prior to the release of laboratory result?)	(a) Hand hygiene	18	3.66
	(b) Wear specific personal protective equipment	81	16.46
	(c) Avoid procedures that could generate infectious aerosols	5	1.02
	(d) All of the above	378	76.83
	(e) None of the above	10	2.03
	Total	492	100
p5 (In case of accidental exposure, what measure will you take to reduce the possibility of disease occurrence?)	(a) Seek medical attention	166	33.74
	(b) Self-medication	1	0.2
	(c) Get vaccinated	27	5.49
	(d) a & b	30	6.1
	(e) a & c	268	54.47
	Total	492	100

presents the distribution of responses to five coded practice questions (p1 to p5) derived from the structured questionnaire. These items assess the respondents’ practical knowledge and behaviors regarding Mpox prevention, sample collection, laboratory knowledge, biosafety practices, and post-exposure response, as defined in the data dictionary for practice items in Appendix A.

For item p1, which assessed how respondents protect themselves from contracting Mpox from suspected human or animal cases, 80.7% (n = 397) selected the comprehensive option “all of the above,” which includes creating awareness in the community, avoiding contact with suspected or confirmed cases, and practicing hand hygiene after exposure. The remaining respondents chose individual strategies, with 8.7% (n = 43) avoiding contact, 7.7% (n = 38) creating awareness, and only 2.9% (n = 14) focusing on hand hygiene alone. This overwhelming preference for a comprehensive approach indicates strong knowledge of appropriate self-protective behaviors among most respondents.

Item p2 examined respondents’ knowledge of appropriate samples for Mpox diagnosis. The majority (68.3%, n = 336) correctly selected the combination of all key diagnostic specimens except saliva, which aligns with standard laboratory protocols. Others selected only skin lesion swabs (26.0%, n = 128), while smaller percentages chose lesion roofs (2.0%, n = 10), lesion crusts (1.0%, n = 5), or saliva (2.6%, n = 13). These findings suggest that while majorities are aware of proper diagnostic sampling methods, a significant number still hold incomplete or less accurate views.

In item p3, respondents were asked to identify the location of the Mpox reference laboratory in Nigeria. Responses were nearly evenly split between Lagos (44.5%, n = 219) and Abuja (44.1%, n = 217), reflecting uncertainty or perhaps a lack of uniform national communication on laboratory network infrastructure. Fewer respondents selected Kaduna (4.9%, n = 24), Enugu (3.9%, n = 19), or Kano (2.6%, n = 13).

For item p4, which assessed proper handling of suspected Mpox cases before the confirmation of laboratory results, a large majority (76.8%, n = 378) selected “all of the above,” indicating they would observe hand hygiene, use appropriate personal protective equipment, and avoid procedures that generate infectious aerosols. Some respondents chose specific practices individually, such as PPE use (16.5%, n = 81) or hand hygiene (3.7%, n = 18), while 2.0% (n = 10) selected “none of the above.” These results indicate that the majority of respondents are familiar with and willing to adopt recommended biosafety measures in managing suspected cases.

Finally, item p5 asked respondents about their preferred action in the event of accidental exposure. Over half (54.5%, n = 268) indicated they would seek medical attention and get vaccinated, reflecting a proactive and medically sound response. A further 33.7% (n = 166) said they would seek medical attention only, while 6.1% (n = 30) opted for both medical attention and self-medication. A small number indicated vaccination alone (5.5%, n = 27), and just one respondent (0.2%) chose only self-medication. This distribution suggests that most respondents understand the importance of prompt medical intervention and the protective role of vaccination in the event of exposure.

In this study, the distribution of responses to Mpox practice items reveals that a large proportion of respondents demonstrated good practical knowledge and adherence to Mpox prevention and response protocols. Notably, 80.7% adopted all recommended self-protective measures, 68.3% correctly identified comprehensive diagnostic sample types, and 76.8% were aware of complete biosafety practices. Although responses about the location of the Mpox reference laboratory were divided almost equally between Lagos and Abuja, indicating some uncertainty, more than half of respondents (54.5%) knew the appropriate steps to take following accidental exposure, including seeking medical attention and getting vaccinated. These findings highlight the strengths and gaps in Mpox-related practical preparedness among the surveyed population.

3.4. Correlation Between Knowledge, Attitude, and Practice Scores

As shown in Figure 3, Spearman’s rank correlation analysis was conducted to assess the strength and direction of monotonic relationships between knowledge, attitude, and practice scores among One Health workforce respondents during the Mpox outbreak. The analysis revealed a moderate positive correlation between knowledge and attitude (ρ = 0.28), indicating that higher knowledge levels were associated with more favorable attitudes. However, the correlations between knowledge and practice (ρ = 0.09) and between attitude and practice (ρ = 0.03) were weak, suggesting that increased knowledge or positive attitudes did not necessarily correspond to better preventive practices.

3.5. Factors Influencing Knowledge, Attitude, and Practice Scores

3.5.1. Results from Univariable Ordinal Regression Models

Table 7. Univariable ordinal logistic regression model examining the association between knowledge score and demographic characteristics.

Variable	Level	OR (Crude)	SE	z	P > z	95% CI
Gender	Male	1.00	Ref.	Ref.	Ref.	Ref.
	Female	1.02	0.18	0.10	0.923	0.71–1.45
Age (years)	30–39	1.00	Ref.	Ref.	Ref.	Ref.
	20–29	1.58	0.35	2.07	0.038	1.03–2.43
	40–49	1.04	0.24	0.16	0.870	0.66–1.63
	50–59	3.33	1.58	2.54	0.011	1.31–8.44
Marital status	Married	1.00	Ref.	Ref.	Ref.	Ref.
	Divorced	0.74	0.96	−0.23	0.816	0.06–9.29
	Single	1.33	0.26	1.47	0.143	0.91–1.94
	Widow	0.15	0.21	−1.37	0.171	0.01–2.24
Qualification	First degree	1.00	Ref.	Ref.	Ref.	Ref.
	Certificate	0.71	0.21	−1.14	0.255	0.40–1.28
	Diploma	0.94	0.22	−0.26	0.796	0.59–1.50
	Fellowship diploma	0.61	0.79	−0.38	0.703	0.05–7.70
	HND	0.73	0.23	−1.00	0.317	0.40–1.35
	Masters	0.85	0.26	−0.53	0.594	0.47–1.54
	PhD	1.06	0.46	0.14	0.892	0.45–2.49
Profession	Registered nurse/midwife	1.00	Ref.	Ref.	Ref.	Ref.
	Environmental scientist	3.17	4.32	0.85	0.398	0.22–45.95
	Health information management	0.65	0.85	−0.33	0.743	0.05–8.29
	Laboratory scientist	0.54	0.19	−1.70	0.089	0.27–1.10
	Medical doctor	1.38	0.67	0.67	0.505	0.53–3.56
	Microbiologist	5.46 × 10−7	0.00	−0.03	0.978	0.00
	Paravet	0.24	0.26	−1.32	0.188	0.03–2.02
	Veterinary doctor	1.03	0.21	0.12	0.902	0.68–1.55

presents the results of a univariable ordinal logistic regression model examining the association between respondents’ knowledge levels on Mpox and various demographic characteristics. Gender was not significantly associated with knowledge, as females had an odds ratio (OR) of 1.02 (95% CI: 0.71–1.45; p = 0.923), indicating no meaningful difference from males. In terms of age, individuals aged 20–29 years were significantly more likely to have higher knowledge levels compared to the 30–39 age group (OR = 1.58; 95% CI: 1.03–2.43; p = 0.038). The 50–59 age group also showed a statistically significant association with higher knowledge (OR = 3.33; 95% CI: 1.31–8.44; p = 0.011), suggesting that older participants in this category were substantially more knowledgeable about Mpox than those aged 30–39 years.

Marital status did not demonstrate any significant associations with knowledge, with all subcategories showing wide confidence intervals and p-values above 0.05. Regarding educational qualification, no statistically significant associations were observed. Respondents with a PhD had slightly higher odds of better knowledge compared to first-degree holders (OR = 1.06; 95% CI: 0.45–2.49; p = 0.892), but this was not statistically significant. Other qualification categories, including certificate, diploma, HND, masters, and fellowship diploma, also showed no meaningful differences.

In terms of profession, most categories did not show statistically significant associations with knowledge levels. Although environmental scientists and medical doctors appeared to have higher odds of greater knowledge compared to registered nurses/midwives, their wide confidence intervals and high p-values suggest a lack of statistical significance. Notably, microbiologists had a reported OR close to zero (5.46 × 10⁻⁷), with no confidence interval reported beyond the lower bound, indicating a probable data sparsity issue for that category. Similarly, laboratory scientists and paravets had reduced odds of higher knowledge, though not statistically significant. Veterinary doctors had nearly equivalent odds to the reference group (OR = 1.03; 95% CI: 0.68–1.55; p = 0.902).

Overall, age was the only demographic variable with statistically significant associations with knowledge about Mpox in this univariable model, particularly among respondents aged 20–29 and 50–59 years.

Table 8. Univariable ordinal logistic regression model examining the association between attitude score and demographic characteristics.

Variable	Level	OR (Crude)	SE	z	P > z	95% CI
Gender	Male	1.00	Ref.	Ref.	Ref.	Ref.
	Female	0.70	0.13	−1.94	0.052	0.49–1.00
Age (years)	30–39	1.00	Ref.	Ref.	Ref.	Ref.
	20–29	0.89	0.19	−0.55	0.583	0.58–1.36
	40–49	1.12	0.26	0.49	0.621	0.71–1.77
	50–59	1.22	0.60	0.41	0.681	0.47–3.20
Marital status	Married	1.00	Ref.	Ref.	Ref.	Ref.
	Divorced	416,514	2.32 × 108	0.02	0.981	0.00
	Single	0.98	0.19	−0.12	0.908	0.67–1.43
	Widow	0.75	1.02	−0.21	0.835	0.05–10.62
Qualification	First degree	1.00	Ref.	Ref.	Ref.	Ref.
	Certificate	1.17	0.35	0.52	0.606	0.65–2.09
	Diploma	1.31	0.32	1.11	0.266	0.81–2.10
	Fellowship diploma	507,078	2.87 × 108	0.02	0.981	0.00
	HND	1.27	0.39	0.78	0.433	0.70–2.33
	Masters	1.37	0.42	1.00	0.316	0.74–2.51
	PhD	1.20	0.51	0.42	0.677	0.52–2.78
Profession	Registered nurse/midwife	1.00	Ref.	Ref.	Ref.	Ref.
	Environmental scientist	420,974	2.35 × 108	0.02	0.981	0.00
	Health information management	0.76	1.02	−0.2	0.838	0.05–10.70
	Laboratory scientist	0.76	0.25	−0.84	0.403	0.40–1.45
	Medical doctor	1.42	0.72	0.69	0.491	0.53–3.83
	Microbiologist	0.18	0.31	−0.99	0.323	0.01–5.48
	Paravet	1.42	1.72	0.29	0.773	0.13–15.27
	Veterinary doctor	1.04	0.22	0.19	0.852	0.69–1.58

presents the univariable ordinal logistic regression results assessing the relationship between respondents’ attitudes and their demographic characteristics. Gender was marginally associated with attitude, where females had lower odds of more positive attitudes compared to males (OR = 0.70, 95% CI: 0.49–1.00, p = 0.052), suggesting a trend toward significance. Age showed no statistically significant association, with the 20–29, 40–49, and 50–59 age groups having odds ratios of 0.89 (p = 0.583), 1.12 (p = 0.621), and 1.22 (p = 0.681), respectively, compared to the 30–39 age group.

Marital status did not reveal significant associations with attitude. While single and widow respondents had odds ratios of 0.98 and 0.75, respectively, these results were not statistically significant (p = 0.908 and p = 0.835). The estimate for divorced respondents was extremely high (OR = 416,514) with a p-value of 0.981 and a confidence interval that effectively rendered the result uninterpretable due to an estimation artifact, likely from sparse data.

Regarding qualification, none of the education levels showed a statistically significant relationship with attitude. Odds ratios ranged from 1.17 for certificate holders to 1.37 for those with a master’s degree, but all p-values were above 0.05. Similarly, an implausibly high OR of 507,078 was noted for fellowship diploma holders, again likely due to sparse data or model convergence issues.

Across professional categories, there was no significant association with attitude. Compared to registered nurses/midwives, other professions such as laboratory scientists (OR = 0.76, p = 0.403), medical doctors (OR = 1.42, p = 0.491), and veterinary doctors (OR = 1.04, p = 0.852) showed no meaningful differences. As with marital status and qualification, environmental scientists showed an extreme OR (420,974) with a p-value of 0.981, indicating an unstable estimate.

In summary, Table 8 suggests that most demographic factors did not show statistically significant associations with attitude levels in the univariable analysis, although gender approached significance, with females tending to have lower odds of more positive attitudes. Several implausibly large odds ratios suggest potential data sparsity in some categories, which may warrant caution in interpretation or the use of penalized models or data pooling for rare groups.

Table 9. Univariable ordinal logistic regression model examining the association between practice score and demographic characteristics.

Variable	Level	OR (Crude)	SE	z	P > z	95% CI
Gender	Male	1.00	Ref.	Ref.	Ref.	Ref.
	Female	1.13	0.20	0.68	0.496	0.80–1.59
Age (years)	30–39	1.00	Ref.	Ref.	Ref.	Ref.
	20–29	1.17	0.25	0.73	0.465	0.77–1.77
	40–49	0.90	0.20	−0.50	0.620	0.58–1.38
	50–59	1.34	0.65	0.60	0.548	0.52–3.45
Marital status	Married	1.00	Ref.	Ref.	Ref.	Ref.
	Divorced	1.37	1.68	0.26	0.798	0.12–15.22
	Single	1.21	0.23	1.01	0.315	0.84–1.74
	Widow	0.00	0.00	−0.02	0.984	0
Qualification	First degree	1.00	Ref.	Ref.	Ref.	Ref.
	Certificate	0.52	0.15	−2.32	0.020	0.30–0.90
	Diploma	0.82	0.19	−0.85	0.394	0.52–1.30
	Fellowship diploma	0.26	0.36	−0.99	0.324	0.02–3.73
	HND	1.00	0.30	0.01	0.994	0.56–1.80
	Masters	1.11	0.33	0.36	0.718	0.62–2.00
	PhD	0.59	0.25	−1.23	0.220	0.26–1.36
Profession	Registered nurse/midwife	1.00	Ref.	Ref.	Ref.	Ref.
	Environmental scientist	1.46	1.80	0.31	0.758	0.13–16.31
	Health information management	0.35	0.47	−0.79	0.432	0.02–4.87
	Laboratory scientist	1.26	0.41	0.71	0.479	0.66–2.39
	Medical doctor	1.09	0.48	0.19	0.850	0.45–2.61
	Microbiologist	1.46	2.54	0.22	0.827	0.05–43.99
	Paravet	0.19	0.23	−1.39	0.165	0.02–2.00
	Veterinary doctor	1.50	0.31	1.94	0.052	1.00–2.24

presents the results of a univariable ordinal logistic regression analysis assessing the association between practice scores and various demographic characteristics. The table reports odds ratios (ORs), standard errors (SEs), z-values, p-values, and 95% confidence intervals (CIs) for each demographic level compared to a designated reference category.

In terms of gender, females had 1.13 times the odds of a higher practice score compared to males, although this association was not statistically significant (p = 0.496, 95% CI: 0.80–1.59). For age, respondents aged 20–29 and 50–59 had higher odds of improved practice compared to those aged 30–39 (OR = 1.17 and 1.34, respectively), while those aged 40–49 had slightly lower odds (OR = 0.90). However, none of these associations reached statistical significance (p-values > 0.05), and the confidence intervals included 1.

With respect to marital status, single respondents (OR = 1.21, p = 0.315) and divorced respondents (OR = 1.37, p = 0.798) had higher odds of better practice compared to married individuals, whereas widowed respondents showed an OR of 0.00 with a corresponding confidence interval of 0, indicating sparse data and estimation issues for this group.

Educational qualification revealed a significant association for respondents with certificate-level qualifications, who had 48% lower odds of better practice compared to those with a first degree (OR = 0.52, p = 0.020, 95% CI: 0.30–0.90). Other qualification levels, including diploma, HND, master’s, and PhD, did not show statistically significant associations, although those with a PhD had lower odds (OR = 0.59, p = 0.220) and those with a master’s degree had slightly higher odds (OR = 1.11, p = 0.718) relative to first degree holders. The fellowship diploma level also showed an OR of 0.26, but with a wide confidence interval and no significance (p = 0.324).

Regarding profession, veterinary doctors had significantly higher odds of better practice compared to registered nurses or midwives (OR = 1.50, p = 0.052, 95% CI: 1.00–2.24), with the lower limit of the CI just at the threshold of 1, suggesting marginal statistical significance. Other professions such as laboratory scientists (OR = 1.26, p = 0.479), medical doctors (OR = 1.09, p = 0.850), and microbiologists (OR = 1.46, p = 0.827) did not show significant associations. Paravets had lower odds of better practice (OR = 0.19, p = 0.165), though this was not statistically significant. Notably, some professions like environmental scientists and health information managers had extremely wide confidence intervals, indicating instability in estimates likely due to small sample sizes.

In summary, Table 9 shows that among the demographic variables assessed, only those with certificate-level qualifications were significantly less likely to report better practice scores compared to those with a first degree. Veterinary doctors also had marginally higher odds of better practice, suggesting a potential area of interest for further investigation in multivariable modeling.

3.5.2. Results from Multivariable Ordinal Regression Models

To determine the demographic variables for inclusion in the multivariable ordinal logistic regression models for each outcome—knowledge, attitude, and practice—a selection strategy based on both statistical and theoretical considerations was employed. As shown in

Table 10. Criteria for inclusion or exclusion of variables in multivariable ordinal logistic regression models.

Outcome	Variable	Decision	Justification
Knowledge	Age	Include	Statistically significant; age groups 20–29 and 50–59 had significant odds ratios.
	Gender	Exclude (with caution)	Not significant (p = 0.923); minimal effect size.
	Marital status	Include	Potential confounder; near significance in univariable analysis.
	Qualification	Include	May act as a confounder; theoretically important.
	Profession	Include	Borderline significance; useful to control for confounding.
Attitude	Gender	Include	Borderline significant (p = 0.052); possible confounding.
	Age	Include	Theoretically relevant demographic variable.
	Marital status	Include (with caution)	Theoretical importance; wide CI suggests regrouping may help.
	Qualification	Include	Educational attainment is important; may need merging of sparse levels.
	Profession	Include	Influences attitudes; consider regrouping sparse categories.
Practice	Gender	Include	Borderline non-significant; may add explanatory value.
	Age	Include	May interact with other predictors or act as a confounder.
	Marital status	Include	Social factor potentially influencing practices.
	Qualification	Include	Certificate holders showed significant negative association (p = 0.020).
	Profession	Include	Veterinary doctors showed borderline significance (p = 0.052); profession is conceptually key.

, the univariable ordinal logistic regression results informed the selection of variables without adjusting for confounding. Variables that were statistically significant, borderline significant, conceptually important, or likely to act as confounders were retained for further adjustment in multivariable analysis.

For the knowledge outcome, age was included due to significant associations for the 20–29 and 50–59 age groups. Although gender was not significant (p = 0.923) and had a negligible effect size, it was excluded with caution due to its theoretical relevance. Marital status was included based on its near-significant association and potential role as a confounder. Qualification and profession were both retained due to their conceptual relevance and likelihood of acting as confounders.

Regarding the attitude outcome, gender was included because of its borderline significance (p = 0.052), and age was retained due to its fundamental role as a demographic variable. Marital status, while associated with wide confidence intervals, was included cautiously for its theoretical importance, particularly with potential regrouping of sparse categories. Qualification and profession were also included, with a recommendation to merge sparse categories to improve model stability.

For the practice outcome, all five demographic variables—gender, age, marital status, qualification, and profession—were retained for multivariable modeling. Although gender was borderline non-significant, it was included for its possible explanatory value. Age was considered a possible confounder or interacting factor. Marital status was deemed relevant as a social determinant of practice behaviors. Notably, certificate-level qualification had a statistically significant negative association with practice (p = 0.020), justifying its inclusion. Profession was retained based on the borderline significance observed for veterinary doctors (p = 0.052) and the conceptual importance of professional background in shaping practice.

Table 11. Final multivariable ordinal logistic regression model showing the association of knowledge with demographic characteristics.

Variable	Level	OR (adjusted)	SE	z	P > z	95% CI
Age (years)	30–39	1.00	Ref.	Ref.	Ref.	Ref.
	20–29	1.57	0.35	2.04	0.042	1.02–2.42
	40–49	1.08	0.25	0.33	0.745	0.69–1.69
	50–59	4.47	2.28	2.93	0.003	1.64–12.16
Profession	Registered nurse/midwife	1.00	Ref.	Ref.	Ref.	Ref.
	Environmental scientist	1.83	2.76	0.40	0.688	0.10–35.02
	Health info. management	0.78	1.02	−0.19	0.848	0.06–10.14
	Laboratory scientist	0.45	0.16	−2.20	0.028	0.22–0.92
	Medical doctor	1.47	0.72	0.80	0.425	0.57–3.83
	Microbiologist	6.91 × 10−7	0.00	−0.03	0.977	0.00
	Paravet	0.24	0.26	−1.31	0.190	0.03–2.03
	Veterinary doctor	1.07	0.23	0.33	0.740	0.71–1.63

presents the final multivariable ordinal logistic regression model for the association between knowledge and demographic characteristics, adjusted for potential confounders. The analysis showed that age was significantly associated with knowledge levels. Compared to respondents aged 30–39 years (reference group), those aged 20–29 years had 1.57 times higher odds of reporting better knowledge (adjusted OR = 1.57; 95% CI: 1.02–2.42; p = 0.042), and those aged 50–59 years had significantly higher odds of higher knowledge scores (adjusted OR = 4.47; 95% CI: 1.64–12.16; p = 0.003). However, the 40–49 age group did not differ significantly from the reference group (adjusted OR = 1.08; 95% CI: 0.69–1.69; p = 0.745).

Regarding profession, with registered nurses/midwives serving as the reference category, laboratory scientists were significantly less likely to have higher knowledge scores (adjusted OR = 0.45; 95% CI: 0.22–0.92; p = 0.028). Other professions did not show statistically significant associations with knowledge. For example, environmental scientists had higher but non-significant odds (adjusted OR = 1.83; 95% CI: 0.10–35.02; p = 0.688), while those in health information management had lower odds (adjusted OR = 0.78; 95% CI: 0.06–10.14; p = 0.848). Similarly, medical doctors (adjusted OR = 1.47; p = 0.425), paravets (adjusted OR = 0.24; p = 0.190), and veterinary doctors (adjusted OR = 1.07; p = 0.740) were not significantly different from the reference category. Microbiologists showed an extremely low odds ratio (adjusted OR ≈ 0.00; p = 0.977), though this result was not statistically meaningful due to the absence of variation.

Table 12. Predictive margins and 95% confidence intervals for knowledge levels by age group.

Knowledge	Age (Years)	Margin	SE	z	P > z	95% CI
Low	20–29	0.11	0.02	5.86	<0.001	0.07–0.15
Low	30–39	0.16	0.02	7.01	<0.001	0.12–0.21
Low	40–49	0.15	0.03	6.11	<0.001	0.10–0.20
Low	50–59	0.04	0.02	2.20	0.028	0.00–0.08
Moderate	20–29	0.60	0.03	23.07	<0.001	0.55–0.65
Moderate	30–39	0.63	0.02	27.63	<0.001	0.59–0.67
Moderate	40–49	0.63	0.02	26.75	<0.001	0.58–0.67
Moderate	50–59	0.42	0.10	4.21	<0.001	0.22–0.61
High	20–29	0.29	0.03	8.64	<0.001	0.23–0.36
High	30–39	0.21	0.03	7.81	<0.001	0.16–0.26
High	40–49	0.22	0.03	7.00	<0.001	0.16–0.28
High	50–59	0.54	0.12	4.55	<0.001	0.31–0.77

presents the predictive margins and corresponding 95% confidence intervals for levels of knowledge across different age groups, based on the multivariable ordinal logistic regression model. These margins are further illustrated in Figure 4.

Among respondents aged 20–29 years, the probability of having low knowledge was 0.11 (95% CI: 0.07–0.15), moderate knowledge was 0.60 (95% CI: 0.55–0.65), and high knowledge was 0.29 (95% CI: 0.23–0.36). For the 30–39 age group, the predicted probability of low knowledge increased slightly to 0.16 (95% CI: 0.12–0.21), moderate knowledge was 0.63 (95% CI: 0.59–0.67), and high knowledge was 0.21 (95% CI: 0.16–0.26). Respondents aged 40–49 had similar patterns, with a low knowledge margin of 0.15 (95% CI: 0.10–0.20), moderate knowledge at 0.63 (95% CI: 0.58–0.67), and high knowledge at 0.22 (95% CI: 0.16–0.28). Notably, among those aged 50–59, the predicted probability of low knowledge dropped to 0.04 (95% CI: 0.00–0.08), moderate knowledge also decreased to 0.42 (95% CI: 0.22–0.61), while the probability of high knowledge rose markedly to 0.54 (95% CI: 0.31–0.77).

These results indicate a clear age-related trend: younger respondents (20–49 years) were more likely to have moderate knowledge and less likely to have high knowledge, while older respondents (50–59 years) were significantly more likely to possess high knowledge and least likely to fall within the low knowledge category. This pattern, supported by narrow confidence intervals and statistically significant z-scores (all p < 0.001), underscores the positive association between increasing age and higher knowledge levels within the surveyed population.

Table 13. Predictive margins and 95% confidence intervals for knowledge levels by profession group.

Knowledge	Profession	Margin	SE	z	P > z	95% CI
Low	Environmental scientist (1)	0.07	0.10	0.72	0.472	−0.13–0.28
Low	Health information management (2)	0.16	0.17	0.92	0.359	−0.18–0.49
Low	Laboratory scientist (3)	0.24	0.06	3.86	<0.001	0.12–0.37
Low	Medical doctor (4)	0.09	0.04	2.27	0.023	0.01–0.17
Low	Microbiologist (5)	1.00	0.00	389.83	<0.001	0.99–1.01
Low	Paravet (6)	0.38	0.25	1.50	0.133	−0.11–0.87
Low	Registered nurse/midwife (7)	0.13	0.02	7.44	<0.001	0.09–0.16
Low	Veterinary doctor (8)	0.12	0.02	5.60	<0.001	0.08–0.16
Moderate	Environmental scientist (1)	0.54	0.24	2.21	0.027	0.06–1.02
Moderate	Health information management (2)	0.63	0.05	13.03	<0.001	0.53–0.72
Moderate	Laboratory scientist (3)	0.62	0.03	19.45	<0.001	0.56–0.68
Moderate	Medical doctor (4)	0.57	0.07	8.62	<0.001	0.44–0.70
Moderate	Microbiologist (5)	0.00	0.00	0.00	0.998	−0.00–0.00
Moderate	Paravet (6)	0.55	0.17	3.15	0.002	0.21–0.88
Moderate	Registered nurse/midwife (7)	0.61	0.02	27.14	<0.001	0.57–0.66
Moderate	Veterinary doctor (8)	0.61	0.03	23.70	<0.001	0.56–0.66
High	Environmental scientist (1)	0.39	0.35	1.11	0.265	−0.29–1.06
High	Health information management (2)	0.21	0.21	1.00	0.318	−0.21–0.63
High	Laboratory scientist (3)	0.14	0.04	3.39	0.001	0.06–0.22
High	Medical doctor (4)	0.34	0.1	3.27	0.001	0.14–0.54
High	Microbiologist (5)	0.00	0.00	0.00	0.998	−0.00–0.00
High	Paravet (6)	0.08	0.08	1.01	0.314	−0.07–0.23
High	Registered nurse/midwife (7)	0.26	0.02	10.79	<0.001	0.21–0.31
High	Veterinary doctor (8)	0.27	0.04	7.68	<0.001	0.20–0.34

presents the predictive margins and corresponding 95% confidence intervals for knowledge levels across various professions, offering insights into the distribution of knowledge among different professional groups. These predictive margins are visualized in Figure 5.

The results reveal substantial variation in predicted knowledge levels among professions. For respondents identified as microbiologists, the predicted probability of having low knowledge was 1.00 (95% CI: 0.99–1.01), with corresponding margins for moderate and high knowledge being 0.00 (95% CI: −0.00–0.00), indicating a strong and exclusive prediction of low knowledge in this group. This pattern was highly statistically significant (p < 0.001). Laboratory scientists had a high probability of moderate knowledge at 0.62 (95% CI: 0.56–0.68), but also a relatively elevated margin for low knowledge at 0.24 (95% CI: 0.12–0.37), and a lower margin for high knowledge at 0.14 (95% CI: 0.06–0.22), suggesting a broad distribution across knowledge levels but skewed toward moderate knowledge.

Veterinary doctors and registered nurses/midwives showed a similar distribution, with moderate knowledge margins of 0.61 (95% CI: 0.56–0.66) and 0.61 (95% CI: 0.57–0.66), respectively. Their high knowledge probabilities were also comparable—0.27 (95% CI: 0.20–0.34) for veterinary doctors and 0.26 (95% CI: 0.21–0.31) for nurses/midwives—indicating a balanced and strong knowledge profile across both groups. Their low knowledge probabilities were modest at 0.12 (95% CI: 0.08–0.16) for veterinary doctors and 0.13 (95% CI: 0.09–0.16) for nurses/midwives.

Medical doctors showed slightly more variation, with a low knowledge margin of 0.09 (95% CI: 0.01–0.17), a moderate knowledge margin of 0.57 (95% CI: 0.44–0.70), and a relatively high probability of possessing high knowledge at 0.34 (95% CI: 0.14–0.54), indicating better knowledge outcomes relative to several other professions. In contrast, environmental scientists and professionals in health information management exhibited broader confidence intervals, suggesting less precision. For instance, environmental scientists had a predicted moderate knowledge probability of 0.54 (95% CI: 0.06–1.02), a wide and less conclusive interval, while their high knowledge estimate of 0.39 (95% CI: −0.29–1.06) lacked statistical significance (p = 0.265), indicating uncertainty in this group’s predicted knowledge levels.

Paraveterinary professionals (paravets) had a moderate knowledge margin of 0.55 (95% CI: 0.21–0.88), and a relatively high but non-significant low knowledge margin of 0.38 (95% CI: −0.11–0.87), while their high knowledge probability remained low at 0.08 (95% CI: −0.07–0.23). This reflects a less favorable and more uncertain knowledge distribution.

Altogether, Table 13 and Figure 5 show that certain professional groups, particularly registered nurses, veterinary doctors, and medical doctors, tend to have higher levels of knowledge with narrower confidence intervals, whereas others like microbiologists and environmental scientists show either extreme predictions or high uncertainty, underscoring potential disparities in knowledge that could be addressed through targeted capacity building.

Table 14. Adjusted predictive margins of One Health knowledge scores by age and profession with 95% confidence intervals.

Age Group	Profession	Margin	SE	z	P > z	95% CI
20–29	Environmental scientist	0.06	0.08	0.7	0.484	−0.10–0.22
20–29	Health information management	0.13	0.15	0.87	0.386	−0.16–0.41
20–29	Laboratory scientist	0.2	0.06	3.26	0.001	0.08–0.32
20–29	Medical doctor	0.07	0.03	2.11	0.034	0.01–0.14
20–29	Microbiologist	1.00	0.00	334.75	<0.001	0.99 –1.01
20–29	Paravet	0.32	0.24	1.35	0.178	−0.15–0.79
20–29	Registered nurse/midwife	0.1	0.02	5.45	<0.001	0.06–0.14
20–29	Veterinary doctor	0.1	0.02	4.28	<0.001	0.05–0.14
30–39	Environmental scientist	0.09	0.12	0.73	0.467	−0.15–0.33
30–39	Health information management	0.19	0.2	0.94	0.348	−0.20–0.57
30–39	Laboratory scientist	0.28	0.07	3.82	<0.001	0.14–0.43
30–39	Medical doctor	0.11	0.05	2.25	0.025	0.01–0.20
30–39	Microbiologist	1.00	0.00	525.61	<0.001	0.99–1.00
30–39	Paravet	0.43	0.27	1.59	0.111	−0.10–0.95
30–39	Registered nurse/midwife	0.15	0.02	6.17	<0.001	0.10–0.20
30–39	Veterinary doctor	0.14	0.03	5.01	<0.001	0.09–0.20
40–49	Environmental scientist	0.08	0.11	0.72	0.473	−0.14–0.31
40–49	Health information management	0.17	0.19	0.92	0.356	−0.20–0.54
40–49	Laboratory scientist	0.27	0.07	3.66	<0.001	0.12–0.41
40–49	Medical doctor	0.1	0.05	2.2	0.028	0.01–0.19
40–49	Microbiologist	1.00	0.00	487.71	<0.001	0.99–1.00
40–49	Paravet	0.41	0.26	1.55	0.122	−0.11–0.93
40–49	Registered nurse/midwife	0.14	0.03	5.34	<0.001	0.09–0.19
40–49	Veterinary doctor	0.13	0.03	4.78	<0.001	0.08–0.19
50–59	Environmental scientist	0.02	0.03	0.68	0.499	−0.04–0.08
50–59	Health information management	0.05	0.06	0.75	0.453	−0.08–0.18
50–59	Laboratory scientist	0.08	0.04	2.1	0.036	0.01–0.16
50–59	Medical doctor	0.03	0.02	1.46	0.146	−0.01–0.06
50–59	Microbiologist	1.00	0.01	117.7	<0.001	0.98–1.02
50–59	Paravet	0.14	0.15	0.98	0.329	−0.14–0.43
50–59	Registered nurse/midwife	0.04	0.02	1.99	0.046	0.00–0.08
50–59	Veterinary doctor	0.04	0.02	1.93	0.054	−0.00–0.07
20–29	Environmental scientist	0.5	0.29	1.74	0.082	−0.06–1.07
20–29	Health information management	0.62	0.1	6.07	<0.001	0.42–0.83
20–29	Laboratory scientist	0.64	0.02	25.88	<0.001	0.59–0.69
20–29	Medical doctor	0.54	0.09	6.39	<0.001	0.38–0.71
20–29	Microbiologist	0.00	0.00	0.00	0.998	−0.01–0.01
20–29	Paravet	0.59	0.15	3.95	<0.001	0.30–0.88
20–29	Registered nurse/midwife	0.6	0.03	21.86	<0.001	0.55–0.65
20–29	Veterinary doctor	0.59	0.03	17.07	<0.001	0.52–0.66
30–39	Environmental scientist	0.58	0.21	2.72	0.007	0.16–1.00
30–39	Health information management	0.64	0.02	26.02	<0.001	0.59–0.69
30–39	Laboratory scientist	0.61	0.04	14.04	<0.001	0.52–0.69
30–39	Medical doctor	0.61	0.06	10.79	<0.001	0.50–0.72
30–39	Microbiologist	0.00	0.00	0.00	0.998	−0.00–0.00
30–39	Paravet	0.51	0.21	2.49	0.013	0.11–0.92
30–39	Registered nurse/midwife	0.64	0.02	27.27	<0.001	0.59–0.68
30–39	Veterinary doctor	0.63	0.02	25.61	<0.001	0.58–0.68
40–49	Environmental scientist	0.57	0.23	2.48	0.013	0.12–1.02
40–49	Health information management	0.64	0.02	26.03	<0.001	0.59–0.69
40–49	Laboratory scientist	0.62	0.04	15.27	<0.001	0.54–0.69
40–49	Medical doctor	0.6	0.06	9.63	<0.001	0.48–0.72
40–49	Microbiologist	0.00	0.00	0.00	0.998	−0.00–0.00
40–49	Paravet	0.53	0.2	2.66	0.008	0.14–0.91
40–49	Registered nurse/midwife	0.63	0.02	25.94	<0.001	0.58–0.68
40–49	Veterinary doctor	0.63	0.03	24.13	<0.001	0.58–0.68
50–59	Environmental scientist	0.29	0.29	1.00	0.318	−0.28–0.86
50–59	Health information management	0.47	0.28	1.64	0.101	−0.09–1.02
50–59	Laboratory scientist	0.57	0.08	7.11	<0.001	0.41–0.72
50–59	Medical doctor	0.33	0.14	2.38	0.017	0.06–0.61
50–59	Microbiologist	0.00	0.01	0.00	0.998	−0.02–0.02
50–59	Paravet	0.63	0.07	9.72	<0.001	0.51–0.76
50–59	Registered nurse/midwife	0.41	0.11	3.94	<0.001	0.21–0.62
50–59	Veterinary doctor	0.4	0.11	3.68	<0.001	0.19–0.61
20–29	Environmental scientist	0.44	0.37	1.18	0.239	−0.29–1.17
20–29	Health information management	0.25	0.25	1.01	0.312	−0.23–0.73
20–29	Laboratory scientist	0.16	0.05	3.08	0.002	0.06–0.26
20–29	Medical doctor	0.39	0.12	3.31	0.001	0.16–0.61
20–29	Microbiologist	0.00	0.00	0.00	0.998	−0.00–0.00
20–29	Paravet	0.09	0.09	1.00	0.316	−0.09–0.27
20–29	Registered nurse/midwife	0.3	0.04	8.22	<0.001	0.23–0.37
20–29	Veterinary doctor	0.31	0.05	6.28	<0.001	0.22–0.41
30–39	Environmental scientist	0.33	0.33	1.00	0.319	−0.32–0.99
30–39	Health information management	0.17	0.19	0.92	0.355	−0.20–0.54
30–39	Laboratory scientist	0.11	0.04	2.96	0.003	0.04–0.18
30–39	Medical doctor	0.29	0.1	2.87	0.004	0.09–0.48
30–39	Microbiologist	0.00	0.00	0.00	0.998	−0.00–0.00
30–39	Paravet	0.06	0.06	0.97	0.334	−0.06–0.18
30–39	Registered nurse/midwife	0.21	0.03	7.08	<0.001	0.15–0.27
30–39	Veterinary doctor	0.23	0.04	5.85	<0.001	0.15–0.30
40–49	Environmental scientist	0.35	0.34	1.01	0.31	−0.32–1.02
40–49	Health information management	0.19	0.2	0.94	0.348	−0.20–0.57
40–49	Laboratory scientist	0.12	0.04	2.94	0.003	0.04–0.19
40–49	Medical doctor	0.3	0.1	2.9	0.004	0.10–0.51
40–49	Microbiologist	0.00	0.00	0.00	0.998	−0.00–0.00
40–49	Paravet	0.07	0.07	0.97	0.332	−0.07–0.20
40–49	Registered nurse/midwife	0.23	0.04	6.29	<0.001	0.16–0.30
40–49	Veterinary doctor	0.24	0.04	5.79	<0.001	0.16–0.32
50–59	Environmental scientist	0.69	0.32	2.15	0.032	0.06–1.32
50–59	Health information management	0.49	0.35	1.39	0.163	−0.20–1.17
50–59	Laboratory scientist	0.35	0.12	3.04	<0.001	0.13–0.58
50–59	Medical doctor	0.64	0.16	4.08	<0.001	0.33–0.95
50–59	Microbiologist	0.00	0.00	0.00	0.998	−0.00–0.00
50–59	Paravet	0.22	0.21	1.08	0.281	−0.18–0.63
50–59	Registered nurse/midwife	0.55	0.12	4.45	<0.001	0.31–0.79
50–59	Veterinary doctor	0.57	0.13	4.48	<0.001	0.32–0.81

presents the interaction between age groups and professional categories, showing their predictive margins (probabilities) for the knowledge score, adjusted for other variables in the model. Figure 6 visually represents these predictive margins with corresponding 95% confidence intervals, allowing for easy comparison across groups.

The results show that the adjusted predictions (or marginal probabilities) vary significantly by both age and profession. Because effect modification by age and profession was prespecified as central to One Health interpretation, we report full interaction margins (predicted probabilities with 95% CIs) in the main text to enable transparent, decision-ready comparisons across stakeholder strata. For example, among professionals aged 20–29, Microbiologists had the highest predicted margin at nearly 1.00 (0.10; 95% CI: 0.99–1.01), indicating almost certain likelihood of the outcome occurring within this subgroup. Other professions in the same age group such as Laboratory scientists (0.201; 95% CI: 0.08–0.32) and Veterinary doctors (0.10; 95% CI: 0.05–0.14) also showed statistically significant predictions with narrow confidence intervals, suggesting more precision in these estimates. Conversely, Paravets and Health information management professionals in this age category showed wider confidence intervals (e.g., Paravets: 0.32; CI: −0.15–0.79), indicating greater uncertainty.

For the 30–39 age group, Microbiologists again had a near-certain outcome (0.10; CI: 0.10–1.00), followed by Laboratory scientists (0.28; CI: 0.14–0.43), and Veterinary doctors (0.14; CI: 0.09–0.20). These values suggest increasing margin probabilities with professional role, especially in clinical and laboratory-focused fields.

Among respondents aged 40–49, a similar trend was observed, with Microbiologists still showing extremely high predicted margins (0.10), and Laboratory scientists having a margin of 0.27 (CI: 0.12–0.41). The profession-specific differences remained significant, while professions like Environmental scientists and Paravets again showed large confidence intervals, indicating less consistent predictions in those roles.

Interestingly, the 50–59 age group revealed a general decline in predicted margins across professions. For instance, Medical doctors had a reduced margin (0.03; CI: −0.01–0.06) compared to younger counterparts, and Veterinary doctors also dropped to 0.0356 (CI: −0.00–0.07), both with wider CIs. Nevertheless, Microbiologists in this group still maintained a high margin close to 1.00 (0.10; CI: 0.98–1.02).

In sharp contrast, for outcome level 2 (presumably indicating a different or improved knowledge category), all professions saw substantial increases in predictive margins across age groups. For example, Health information management professionals aged 20–29 had a margin of 0.6244 (CI: 0.42–0.83), and Veterinary doctors in the same age group showed 0.5911 (CI: 0.52–0.66), suggesting a positive shift in outcome prediction for these professions at level 2. These differences are further visualized in Figure 6, where the distinction in margins by age and profession is clearly delineated with vertical confidence interval bars, highlighting the significant disparities and overlaps in outcome predictions.

Overall, the interaction between age and profession plays a critical role in shaping the adjusted predictions for the knowledge outcome. Younger professionals in clinical and laboratory fields tend to show higher or more certain predicted margins at knowledge level 1, while all professions show marked improvements at knowledge level 2, emphasizing both the importance of age and professional background in influencing the knowledge outcome.

Table 15. Final multivariable ordinal logistic regression model showing the association of attitude with gender.

Variable	Level	OR (Adjusted)	SE	z	P > z	95% CI
Gender	Male	1.00	Ref.	Ref.	Ref.	Ref.
	Female	0.70	0.13	−1.94	0.052	0.49–1.00

presents the results of the final multivariable ordinal logistic regression model examining the association between gender and attitude score toward Mpox prevention and control. In the model, male respondents served as the reference group with an odds ratio (OR) of 1.00. Female respondents had lower odds of demonstrating a more positive attitude compared to males, with an adjusted OR of 0.70. The corresponding p-value was 0.052, indicating that the association was marginally non-significant at the 5% level. The 95% confidence interval for the odds ratio ranged from 0.49 to 1.00. This suggests that, although not statistically significant, there is a trend indicating that females may be less likely than males to hold more favorable attitudes toward Mpox prevention and control after adjusting for other covariates in the model.

Table 16. Predictive margins and 95% confidence intervals for attitude levels by gender group.

Attitude	Gender	Margin	SE	z	P > z	95% CI
Low	Female	0.05	0.01	4.59	<0.001	0.03–0.07
Low	Male	0.04	0.01	4.23	<0.001	0.02–0.05
Moderate	Female	0.39	0.03	14.10	<0.001	0.34–0.44
Moderate	Male	0.32	0.03	11.14	<0.001	0.26–0.38
High	Female	0.56	0.03	18.62	<0.001	0.50–0.62
High	Male	0.64	0.03	20.17	<0.001	0.58–0.71

and Figure 7 present the predictive margins of attitude toward Mpox prevention and control across gender groups. The margins represent the predicted probabilities of respondents falling into each attitude category—low, moderate, or high—based on their gender, with 95% confidence intervals.

In Table 16, the probability of having a low attitude score was slightly higher among females (margin = 0.05, 95% CI: 0.03–0.07) than males (margin = 0.04, 95% CI: 0.02–0.05). For the moderate attitude category, females also showed a higher predicted probability (0.39, 95% CI: 0.34–0.44) compared to males (0.32, 95% CI: 0.26–0.38). However, the trend reversed for the high attitude category, where males had a higher predicted probability (0.64, 95% CI: 0.58–0.71) than females (0.56, 95% CI: 0.50–0.62). All differences across gender and attitude levels were statistically significant (p < 0.001).

Figure 7 visually illustrates these predictive margins along with their 95% confidence intervals, reinforcing the statistical findings in Table 16. The figure clearly shows that males are more likely to have a high attitude score, while females are more likely to be in the low or moderate categories, further supporting the marginally non-significant association found in the regression model (Table 15).

Given the observed results, we decided not to present the multivariable ordinal logistic regression model for the association of practice with demographics as the final model. The added complexity of the multivariable approach did not yield meaningful results, with no significant predictors retained in the model. Instead, we focused on the univariable analyses, where significant associations were observed and could be more reliably interpreted. The lack of significance in the multivariable model may be attributed to potential collinearity between predictors, limitations in sample size, or the absence of a true association between the explanatory variables and the outcome. Since the multivariable model did not improve upon the univariable model in terms of explanatory power or model fit, we chose to present and interpret only the univariable results. Although the multivariable model was explored, it did not offer additional insights and thus was not retained as the final model.

Table 17. Model fit statistics for final ordinal logistic regression models of knowledge and attitude.

Model	Obs	Log-Likelihood (Null Model)	Log-Likelihood (Model)	df	AIC	BIC
Final ordinal logistic model (knowledge)	492	−452.382	−441.435	12	906.869	957.2508
Final ordinal logistic model (attitude)	492	−400.671	−398.776	3	803.5509	816.1463

presents the model fit statistics for the final ordinal logistic regression models assessing the association of demographic variables with knowledge and attitude regarding Mpox among respondents. The model for knowledge included 492 observations, with a log-likelihood value of −452.382 for the null model and −441.435 for the fitted model, yielding an Akaike Information Criterion (AIC) of 906.869 and a Bayesian Information Criterion (BIC) of 957.2508 based on 12 degrees of freedom. For the final model on attitude, also based on 492 observations, the null and fitted model log-likelihoods were −400.671 and −398.776, respectively, with an AIC of 803.5509 and a BIC of 816.1463 based on 3 degrees of freedom. Lower AIC and BIC values in the attitude model compared to the knowledge model suggest that the attitude model has a comparatively better fit to the data, especially given its more parsimonious structure with fewer parameters. These results support the suitability of both models for ordinal logistic regression analysis, with better model fit indicated by improvements in log-likelihood and lower AIC/BIC values compared to the respective null models.

4. Discussion

4.1. Summary of Key Findings

This study explored the knowledge, attitudes, and practices (KAP) regarding Mpox among Nigeria’s One Health stakeholders, encompassing professionals in the human health, animal health, and environmental health sectors. The findings indicate a generally moderate to high level of knowledge about Mpox among respondents. However, there were evident gaps in specific areas, particularly concerning zoonotic transmission pathways, recognition of animal reservoirs, and understanding of the clinical presentation of the disease in non-human species. This uneven distribution of knowledge across sectors suggests the need for more integrated and cross-disciplinary training in One Health education frameworks.

Attitudinal responses showed that most participants expressed positive views toward the need for multi-sectoral collaboration in controlling Mpox. Respondents demonstrated a basic appreciation of the disease’s public health significance, supporting active surveillance and risk communication. Nonetheless, such positive attitudes did not uniformly translate into preventive or surveillance-related practices. The findings reveal inconsistencies between knowledge and action, a phenomenon previously observed in KAP studies of zoonotic diseases and public health emergencies [35,36,37,38,39,40].

In particular, responses suggested that many participants lacked hands-on experience or familiarity with current reporting systems, personal protective protocols, or intersectoral response strategies. This gap between awareness and application may reflect deeper structural barriers, including inadequate training, poor resource allocation, or a lack of clearly defined roles in emergency preparedness at the interface of human, animal, and environmental health [41,42].

Moreover, the limited knowledge of wildlife and small mammals as potential reservoirs—particularly among stakeholders in the human and environmental health sectors—raises concerns about the level of operational preparedness within the One Health framework. As Mpox is a re-emerging zoonotic disease endemic to parts of West and Central Africa, including Nigeria, such knowledge gaps could hinder early detection and response efforts, particularly in rural or forest-adjacent communities where human–wildlife interactions are more frequent [2,9].

Collectively, the findings suggest that while there is a foundational understanding of Mpox within Nigeria’s One Health workforce, there remain critical weaknesses in terms of operational knowledge, intersectoral coordination, and practical application of prevention and control measures. These insights highlight the importance of strengthening cross-disciplinary training, expanding surveillance capacity, and reinforcing the operationalization of the One Health approach across all levels of health governance in Nigeria.

4.2. Comparison with Previous Studies

The observed knowledge level in our study is relatively lower than the 52.2% reported in Nigeria by [26] among Nigerian healthcare workers during the 2022 Mpox outbreak. This discrepancy may be due to methodological differences [26]; conducted a cross-sectional web-based survey involving 609 healthcare workers, a population more likely to have direct clinical exposure and formal training in infectious disease management. In contrast, our study involved 492 respondents drawn from a broader array of disciplines within the One Health workforce, including human, animal, and environmental health sectors. While this multidisciplinary inclusion enriches the contextual understanding of zoonoses, it may also dilute the concentration of clinical expertise, potentially contributing to lower aggregate knowledge scores. Nevertheless, such diversity aligns with the One Health paradigm and supports more comprehensive surveillance and response efforts for zoonotic diseases like Mpox [19]. Our findings highlight moderate overall knowledge but reveal gaps concerning the dominant role of human-to-human transmission during the 2022–2024 outbreak and the implications for targeted prevention in sexual networks. Integrating this perspective into One Health training—alongside zoonotic risks—may better align workforce preparedness with the current epidemiology.

Internationally, the moderate level of knowledge observed in our study contrasts with the higher knowledge levels reported in Iraq’s Kurdistan region, where a considerable proportion of healthcare providers demonstrated good knowledge of Mpox [43]. This discrepancy may be partly attributed to differences in sample size, as the cited study involved a larger cohort of respondents, potentially offering a more robust assessment of knowledge distribution. Additionally, the focus on healthcare professionals alone in that study—as opposed to the broader One Health stakeholder representation in our survey—may have contributed to the higher knowledge scores, given that medical personnel are more frequently exposed to disease-specific training. In contrast, our study included participants from human, animal, and environmental health sectors, some of whom may have had limited formal exposure to Mpox-specific content.

Meanwhile, studies conducted in non-endemic countries have consistently shown lower knowledge levels. For instance [44], identified substantial gaps in Mpox-related knowledge among general practitioners in Italy, while [45] found limited understanding of transmission, prevention, and clinical features among members of the public and healthcare workers. These lower knowledge levels are likely linked to the reduced perceived risk in such settings, where Mpox is not endemic, resulting in limited emphasis within public health education and training frameworks.

Attitudes toward Mpox prevention and intersectoral collaboration in our study were more favorable compared to those reported in Ethiopia [46] and Bangladesh [47], where skepticism and indifference toward Mpox-related interventions were evident. Nigeria’s relatively frequent experience with zoonotic disease outbreaks, including Mpox, Lassa fever, and avian influenza, may contribute to more proactive attitudes. Prior exposure to real-time outbreaks tends to shape stakeholder receptivity to disease prevention and control strategies, as noted in other outbreak-prone regions [48]. Despite high attitude scores, only 48.2% reported adequate practices, suggesting systemic barriers such as limited PPE availability, reporting mechanisms, and institutional protocols may have constrained behavior, regardless of individual willingness.

Despite encouraging attitudes, only 48.2% reported adequate practices, highlighting systemic barriers such as limited availability of personal protective equipment (PPE), weak disease reporting mechanisms, and insufficient inter-agency coordination that constrained behavior despite individual willingness. This finding echoes previous reports from Uganda and the Democratic Republic of Congo (DRC), where limited access to PPE, weak disease notification systems, and insufficient cross-sectoral protocols hamper effective Mpox containment [3,49,50]. It also reflects the findings of [9], who noted operational inefficiencies during Nigeria’s 2017–2018 Mpox outbreak response, particularly at subnational levels.

Furthermore, the poor recognition of animal reservoirs and transmission pathways among respondents in our study is consistent with global assessments that have identified persistent gaps in the understanding of Mpox’s zoonotic origins [51,52]. Even among health professionals, there is often limited familiarity with the roles of rodents, non-human primates, and other wildlife in the ecology of Mpox virus. Such knowledge deficits complicate efforts at surveillance, early detection, and source-tracing in both endemic and non-endemic contexts [53]. These findings underscore the need for expanded One Health education programs that include modules on zoonotic reservoirs, ecological risk factors, and wildlife-livestock-human interfaces.

In summary, while the level of knowledge and attitudes among Nigerian One Health stakeholders compares favorably to both local and international studies, key weaknesses in practices and operational zoonotic literacy persist. Addressing these gaps is critical to enhancing preparedness and response capacity, particularly as Mpox continues to pose a re-emerging global health threat.

4.3. Strengths and Implications

A major strength of this study lies in its incorporation of the One Health framework, which is essential for understanding and managing zoonotic diseases like Mpox. By including stakeholders from multiple sectors, this study offers an integrated view of KAP patterns, which is critical for designing cross-sectoral interventions [19,54]. This approach is especially relevant in Nigeria, where recurrent zoonotic outbreaks necessitate coordinated responses between veterinarians, clinicians, public health officers, and environmental scientists.

The findings have practical implications for disease surveillance, risk communication, and outbreak preparedness. The suboptimal practices identified point to a need for enhanced training, provision of PPE, and better enforcement of surveillance protocols. These insights could inform the development of targeted education programs and policy revisions to improve One Health workforce competencies and resilience to future outbreaks [55].

In addition, this study contributes to the limited empirical data on Mpox-related KAP in sub-Saharan Africa, particularly among non-human health sectors. As such, it serves as a valuable evidence base for national strategies aligned with the WHO’s call for enhanced Mpox preparedness and response in endemic regions [56].

4.4. Limitations

Several limitations of this study must be acknowledged. First, the cross-sectional design limits causal inferences regarding the relationship between KAP components and demographic or professional factors. Secondly, our findings highlight that respondents generally recognized Central and West Africa as the primary endemic regions of Mpox, consistent with historical epidemiological patterns. However, we acknowledge that the phrasing of the knowledge item on Mpox prevalence may have limited recognition of the virus’s contemporary global distribution, particularly after the 2022–2024 outbreaks. This reflects an important limitation in interpreting knowledge levels and underscores the need for future KAP assessments to explicitly distinguish between historical endemicity and current worldwide prevalence. To improve accuracy, we refined our scoring rubric during revision. Notably, k8 was reverse-scored to reflect the contemporary epidemiology of Mpox transmission during 2022–2024, and k17 was excluded from the knowledge score due to its context-dependent nature (antivirals are not universally required but may be indicated for severe cases). These adjustments did not materially change the interpretation of knowledge levels but ensured alignment with current scientific evidence and enhanced the transparency of our methods. Furthermore, the inclusion of professional categories with minimal representation, such as microbiologists (n = 1), yielded implausible regression estimates due to sparse data bias [57]. While not interpretable, we retained these outputs for transparency, as their presence highlights underrepresentation of certain key One Health professions in our survey, itself an important workforce gap for future capacity building. Thirdly, the use of self-reported data introduces the possibility of social desirability bias, particularly in responses related to attitudes and practices. Participants might have overstated positive practices or favorable attitudes, thereby inflating the true levels of preparedness.

Moreover, the study relied heavily on online survey dissemination, which may have excluded potential respondents from rural or underserved areas with limited internet access. This could result in selection bias, skewing the sample towards more educated or urban-based professionals. Additionally, the scoring thresholds for knowledge, attitude, and practice classifications were arbitrary to some extent and may not fully capture the nuances of preparedness. Lastly, while our study included diverse One Health actors, the sample sizes for some subgroups (e.g., environmental health officers) were relatively small, limiting subgroup-specific conclusions.

4.5. Recommendations

Based on our findings, we propose several recommendations for policy, practice, and further research:

• Capacity building: National and state ministries should prioritize One Health training on emerging zoonoses like Mpox. This includes workshops on transmission dynamics, surveillance, and use of PPE tailored to each professional group. Such training should also integrate updated knowledge on the current epidemiology of Mpox, especially the importance of human-to-human transmission through close and sexual contact, to ensure stakeholders are adequately equipped to address both zoonotic and interpersonal modes of spread.
• Policy integration: The Nigerian Centre for Disease Control (NCDC), in collaboration with veterinary and environmental agencies, should develop harmonized Mpox response protocols, ensuring inclusion of non-clinical health professionals in outbreak response teams.
• Surveillance strengthening: Improve active surveillance at human–animal-environment interfaces. Tools like integrated disease surveillance and response (IDSR) and event-based surveillance (EBS) should be decentralized and adapted for zoonoses.
• Community engagement: Strengthen community awareness through culturally relevant health education that incorporates indigenous knowledge and behavioral change communication, especially in high-risk communities.
• Further research: Longitudinal studies are needed to assess changes in KAP over time, especially in relation to interventions. Future research should also explore the psychosocial and institutional determinants of KAP among One Health stakeholders.

5. Conclusions

This study assessed the knowledge, attitudes, and practices (KAP) regarding Mpox among One Health stakeholders in Nigeria, encompassing professionals from the human, animal, and environmental health sectors. The findings reveal a moderate level of knowledge, largely positive attitudes, and suboptimal preventive practices toward Mpox. While participants demonstrated awareness of human-to-human transmission, there were notable gaps in understanding zoonotic reservoirs and clinical signs in animals—critical areas for effective surveillance and control in a One Health context.

The disparity between knowledge and practice underscores the need for strengthened intersectoral collaboration, continuous professional training, and investment in field-level operational capacities. These findings highlight the importance of tailored, multisectoral public health education and preparedness strategies to bridge existing knowledge-practice gaps and improve the response to emerging zoonotic threats like Mpox in Nigeria and other endemic settings.