Epidemiological profile, trend, seasonality, and factors associated with Yellow fever disease and barriers to vaccination: a retrospective observational study of multi-season national surveillance data in Nigeria, 2017 - 2021

Data Source Description

We retrospectively analyzed YF surveillance data obtained from the Surveillance, Outbreak, Response, Management, and Analysis System (SORMAS) platform managed by the Nigeria Centre for Disease Control (NCDC) to which State ministries and Local Government departments of health reported. SORMAS in Nigeria, part of the integrated disease surveillance and response system (IDSR), has been previously described [5,11]. In brief, once a patient meets the standard case definition for a suspected case of YF, the attending physician notifies the Local Government Area (LGA) Disease Surveillance and Notification Officer (DSNO); the LGA DSNO then notifies the State Epidemiologist and the State DSNO, who verifies the case and then inform the focal person at the NCDC. The NCDC, in conjunction with the State and the LGA officers, constitutes an outbreak investigation team to investigate the case using a case investigation form (CIF). The cases are line-listed, and blood samples are collected for laboratory confirmation in line with the National Guidelines for YF Preparedness and Response [5,18]. Serological testing involved the detection of YF-specific antibodies in the serum samples of suspected cases using Enzyme-linked Immunosorbent Assay (ELISA; U.S. CDC MAC protocol; done in-country) and plaque reduction neutralization test (PRNT; done at the Regional Reference Laboratory, Dakar, Senegal). ELISA positive results were recorded as “presumptive positive” while negative results were recorded as “negative cases”. Both ELISA positive samples and “inconclusive samples” were sent to the Regional Reference Laboratory for confirmation by PRNT. Molecular testing by real-time polymerase chain reaction (RT-PCR) was done in-country for samples collected from suspected cases within ten days of onset of symptoms to detect YF viral genome in blood. PCR negative samples were further tested by serology to rule out false negative results while positive samples are reported as confirmed YF cases [18]. Information on the CIF is logged into the SORMAS platform using the epidemiological number issued by the State Epidemiologist as a unique identifier. The data abstracted from the SORMAS platform include epidemiological number, socio-demographic information (age, sex, education, occupation), vaccination status, clinical information (date of onset, date of admission, symptoms), location (state), trave history, and laboratory test status (negative, positive, unconfirmed) from July 2017 to December 2021.

Case definition

The case definitions applied for YF surveillance in Nigeria are as follows: A suspected case of YF was defined as “any person with acute onset of fever or history of fever and/or general body weakness, vomiting, bleeding from any part of the body, and jaundice appearing within 14 days of onset of the first symptom” [18]. A presumptive case was defined as a “suspected case in whom was found the presence of YF IgM antibody in the absence of YF vaccination within the last 30 days before onset of illness or positive post-mortem liver histopathology or epidemiological link to a confirmed case or an outbreak”. A confirmed case was defined as “ presumptive case with one of the following: detection of YF IgM; detection of four-fold rise in IgM or IgG antibody titers between acute and convalescent serum samples or both; detection of YF-specific neutralizing antibodies and absence of YF immunization within 30 days before onset of illness or one of the following: detection of YF viral genome in blood, liver or other organs by immunoassay, isolation of YF virus and absence of YF immunization within 14 days before onset of illness [18]. However, for this study, a positive case was defined as a suspected case that was confirmed positive for YFV by either serology or RT-PCR. A negative case is a suspected case that was confirmed negative by both serology and RT-PCR. An unconfirmed case is a suspected case for which laboratory test result was missing in the line list.

Additional data sources for the analysis

Administratively, Nigeria has 774 LGAs, which are clustered into 36 states and the Federal Capital Territory and zoned into six geo-political zones (North-Central, North-East, North-West, South-East, South-West, and South-South). The geo-political zones were created based on political and socio-cultural considerations from the northern and southern regions. There are two climatic seasons in Nigeria (dry season and rainy season). The rainy season begins in April and ends in October while the dry season commences from November and terminates at about the end of March [19]. Forests spanned over nine ecological zones [20] (S1 Fig). The northern part of the country is dotted with Savannah-type vegetation whereas the southern region is characterized by weather conditions that promote luxuriant rainforest and proliferation of insect populations. It is estimated that forest reserves occupy 10% of the total land mass, with 78.7% in the Savannah zone, 7.7% in the derived Savannah zone, and 13.6% in the rain forest zone [19,21]. The ecological zones have been described as running from the northern to the southern parts in a contiguous fashion, in which natural ranges extending from the mangrove swamps and tropical forests in the coastal flanks become shrubs and Savannah woodlands in the low Plateau, which extends into the semi-arid plains of the northern region in central Nigeria and the eastern highlands. This complex mesh of forest pattern had resulted in a variety of vegetation in the landscape, including Sahel, Guinea, and derived Savannah as well as lowland rain forest, freshwater and mangrove forests, coastal marshes, and estuaries. The ecological zones are repositories of a wide variety of plants and animal species, including 20,000 insect and 247 mammalian species [19]. The variables geo-political zone, geographical region, and ecological zone were generated based on the states in which cases were reported.

Definition of variables

Time-oriented variables such as annual quarter, epidemic period (pre-pandemic, pandemic) and climatic season (dry season, rainy season) were generated by grouping cases by date of onset of illness. Epidemic period was defined as whether a case developed symptoms before (pre-pandemic) or during (pandemic) the period of the COVID-19 pandemic (January 2020 and beyond) while climatic season was defined as whether the case occurred during the months of the dry season or the rainy season. Socio-demographic factors presumed to be associated with outcome variables were identified from previous studies [2,4,22]. Age in years and sex were analyzed as reported by cases and relatives. Age was used to create two sets of age group variables: For the analysis of factors associated with outcome variables (YFD and being unvaccinated) without disaggregation, age was categorized into seven levels (0-4, 5-9, 10-14, 15-29, 30-44, 45-59, and 60+ years); however, when YFD was disaggregated by vaccination status, age was categorized into two levels (<15 and >=15 years) to accommodate small sample sizes. Geo-political zone, geographical region, and ecological zone were generated using the state in which cases were detected. Educational status was created from the educational level reported by cases – primary, secondary, tertiary, Islamic education, no education. These were collapsed into two levels: all those who had any formal education were defined as “educated” and those with no education were defined as “no education” alongside those who did not report their educational status, considering the current illiteracy level of nearly 40.0% in Nigeria. Occupational group was generated from the occupation reported by cases, which were loosely defined. Those working in office settings, housewives, and kids below school age were grouped as “home-based/office workers”; those who reported business, trading, plumbing, electrician etc. were grouped as “artisans/traders”. In Nigeria the term “business” loosely includes people involved in any form of buying and selling of goods and services. Cases who were of school age (6-17 years) but were not reported as students were grouped as “students” alongside those who reported being students based on the premise that basic education is mandatory by law in Nigeria. Those involved in animal sales, butchers, hunters, and farmers were grouped as “butchers/farmers/hunters”; and those working in healthcare settings were defined as health workers. Cases who reported no occupation were defined as unemployed in view of the high unemployment rate in Nigeria.

Statistical analysis

De-identified and anonymized national line-list of YF outbreak cases from 1st July 2017 to 31^st December 2021 was abstracted from the SORMAS platform in compliance with NCDC data sharing protocol. We received the dataset in Microsoft Excel format from the data management on 5^th February 2022 with personal identifiers of the cases removed. Categorical variables were summarized using frequencies and percentages while numerical variables were summarized using median and inter-quartile range. Our main outcome variables were: (1) YFD, defined as a suspected case of YF confirmed for YFV by either serology or RT-PCR (a categorical variable with two levels: No = No YFD, Yes = YFD) and (2) unvaccinated, defined as a suspected with no previous YF vaccination before the onset of symptoms (a categorical variable with two levels: No/Yes). Comparison of percentages between outcome and explanatory variables was done using Pearson’s Chi square statistic and Fisher’s exact test for categorical variables and Kruskal-Wallis rank sum test for numerical variables with p-value <0.05 considered to be statistically significant. We used date of onset of illness to plot epidemic curves with trend lines (histogram with line plots) and time series seasonality plots (line plot in which cases were aggregated by months over the entire study period) to show trends of cases over time. A population-adjusted age-sex pyramid of cases was also plotted using the ages of cases and country age-sex population structure.

At bivariate analysis, odds ratio (OR) was used as a measure of association between seasonal, geographical, socio-demographic, and clinical characteristics and respective outcome variables (YFD: No = 0, Yes = 1; unvaccinated: No = 0; Yes = 1) and the 95% confidence intervals with p-values (<0.05) were estimated. Subsequently, multivariate logistic regression analysis was done using stepwise selection process, which was based on Akaike Information Criterion (AIC) [23] - implemented in R using the stepAIC function in the MASS package [24] with backward selection. Here, explanatory variables with level of significance p<0.15 at bivariate analysis were included in the multivariate logistic regression model to determine adjusted odds ratios, 95% confidence interval, and p-values of predictors of the respective outcomes. We used variance inflation factor to assess multicollinearity of the variables in the final model and variables that exhibited collinearity values >= 5.0 were excluded from the final model. For variables that exhibited collinearity, the one with a higher variance inflation factor was dropped from the final model. For both outcome variables, bivariate and multivariate analysis was first done with the whole dataset and then disaggregated by vaccination status (YFD) and laboratory test result (being unvaccinated). We conducted our analysis with RStudio version 4.4.2 and R version 4.4.2 (http://posit.co/download/rstudio-desktop/). The main R packages used in our analysis were the tidyverse package (data wrangling) [25], tableone package (descriptive table) [26], gtsummary and finalfit packages (bivariate and multivariate analysis)[27], and TSstudio package (seasonality analysis)[28].

Ethical consideration

Approval for this study was sought by submission of the study proposal (see Other supporting material) to the Research Ethics Committee of the Nigeria Centre for Disease Control and Prevention, which was reviewed and approved. The Research Ethics Committee then requested the authors to complete, sign, and submit a data sharing agreement form to access the dataset (see Other supporting material). The data has been previously collected, stored, and analyzed as part of outbreak response activity by the NCDC during which a case investigation form was administered by trained clinical staff to obtain relevant information from suspected cases of YF to which they consented in writing following detailed explanation about the use of their information by appending their signature and witnessed by an attending nurse and a family relative. In the case of minors, informed consent was given by either a parent or guardian. The data was confidentially stored in passworded computer systems by NCDC data management team and was shared in de-identified and anonymized Microsoft Excel format for this study. This necessitated a waiver of the need for informed consent for this study by the Research Ethics Committee.

Trend and seasonality of case incidence

Case reporting was higher during the 2019 epidemic season (34.0%) while YFV positive case was higher during the 2020 epidemic season (44.0%). While suspected case reporting was higher during the third and fourth quarter of the year, YFV positivity was more in the fourth quarter of the year (Table 1).

The highest peak of case reporting and YFV case positivity were seen around Week 43 in the 2020 epidemic season (Fig 1A). The highest incidence of involvement of vaccinated cases was seen during week 34-43 of the 2019 epidemic season (Fig 1B). The initiation of YF vaccination (reactive and preventive) campaign between 2017 and 2018 apparently resulted in a flattening of the 2018 epidemic peak (around week 35-40).

• Download figure
• Open in new tab

Fig 1. Time trends of incidence of YF cases in Nigeria, 2017-2021.

Lines indicate weekly mean incidence. In: (A) average weekly incidence were stratified by laboratory test result; (B) average weekly incidence disaggregated by vaccination status. Light blue histogram background is the epidemic curve of all suspected cases.

The North-west geo-political zone presented a somewhat characteristic trend of case incidence pattern with several peaks seen over the entire period, usually during the dry season (highest peak seen between week 34 and 43 of the 2019 epidemic season), which were much larger in magnitude and longer in duration, except during the 2020 epidemic season (Fig 2A). This pattern was found to be inherent in the Sahel Savannah ecological zone (Fig 2B).

• Download figure
• Open in new tab

Fig 2. Trends of incidence of reported YF cases in Nigeria, 2017-2021.

Lines indicate weekly mean incidence. In: (A) cases were disaggregated by geo-political zone; and (B) cases were disaggregated by ecological zone. Light blue histogram background is the epidemic curve of all reported cases.

The 2020 epidemic season presented the most distinct trend pattern throughout the entire period in which the highest peak incidence seen in the South-east geo-political zone reflected the outbreak pattern of the Derived Savannah. Annually, YF case reporting increased progressively from about the middle of the rainy season between June and July with the peak incidence seen sometime between September and November (Figs 3A and B).

• Download figure
• Open in new tab

Fig 3.

Seasonality plot of reported Yellow fever cases in Nigeria, 2017-2021. (A) all suspected cases, (B) YFV confirmed cases. Cases were aggregated by month in a line plot over the entire period.

Overall case reporting was higher during the pre-pandemic era (2016-2019; 58.3%) while YFV positive case was more during the COVID-19 pandemic era (2020-2021; 55.0%; <0.001) (Table 1).

Socio-demographic and clinical characteristics

Between July 1, 2017 and December 31, 2021, Nigeria recorded a total of 13,014 suspected YF cases, of which 7640 (58.7%) had laboratory confirmation by either serology or RT-PCR. Overall YFV case positivity was 7.02% (536/7640). Median age was higher among YFV positive cases (median= 20.0 years; IQR= 12.0 – 30.0 years) (Table 1). The most affected age group were those 15-29 years, who also accounted for about 30.0% (64,883,373) of the total population and about 32.0% (3936) of all suspected cases (Fig 4A). They were also the most unvaccinated group (Fig 4C) and had the largest proportion of persons with unknown vaccination status (Fig 4D).

• Download figure
• Open in new tab

Fig 4. Age-sex pyramid of Yellow fever cases in Nigeria, 2017-2021.

From left to right: (A) all reported cases, (B) vaccinated cases, (C) unvaccinated cases, and (D) cases of unknown vaccination status. The percentage of cases were superimposed on country age-sex population structure. Pink = female and blue = male. The light-shaded bars indicate country demographic age-sex structure.

Cases were mostly males (all: 56.5%; positive: 63.6%; negative: 54.8%; unconfirmed: 58.1%; p<0.001). Case reporting was higher in the Sudan Savannah (28.2%) whereas YFV positive case was more in the Derived Savannah (38.8%, p<0.001) ecological zone. They were predominantly of uneducated (all: 84.5%; positive: 82.6%; negative: 85.3%; unconfirmed: 83.7%; p=0.029); mostly home-based or working in office setting (41.3%), but the unemployed had higher YFV case positivity (43.3%, p<0.001). They were largely of unknown vaccination status (all: 69.4%; positive: 83.2%; negative: 71.7%; unconfirmed: 65.0%; p<0.001) (Table 1).

Clinical symptoms commonly presented among reported cases were fever (63.9%), jaundice (28.7%), headache (16.1%), fatigue (11.6%), dark urine (7.1%), and vomiting (6.1%). YFV positive cases presented more with fever (70.9%) and had a higher admission rate (11.9%) (Fig 4).

• Download figure
• Open in new tab

Fig 5. Clinical characteristics of reported Yellow fever cases in Nigeria, 2017-2021.

(A) all reported cases, and (B) YFV confirmed cases. Yellow bar = symptom is present (Yes), purple bar = symptom not presented, and green bar = missing data.

Seasonal, geographical, socio-demographic, and clinical factors associated with YFD

Factors associated with YFD are shown in Table 2. Independent predictors of YFD were male sex compared to female (aOR: 2.36, 95% CI: 1.45–3.91; p<0.001); age group being 15-29 years compared to under-fives (aOR: 4.13, 95% CI: 1.59–13.00, p=0.007); residing within the Derived Savannah, Lowland/Mangrove/Freshwater rainforest, and the Guinea Savannah/Jos Plateau compared to the Sahel/Sudan Savannah (aOR: 30.10, 95% CI: 11.50–104.00; aOR: 8.84, 95% CI: 3.24–31.10; aOR: 6.13, 95% CI: 1.90–23.50 respectively, p<0.001). Working in outdoor setting was associated with YFD but not statistically significant (aOR: 1.76, 95% CI: 0.96–3.22, p=0.067). Clinical predictor of YFD was presenting with vomiting (aOR: 2.62, 95% CI: 1.39–4.83; p=0.002). The rainy season was protective against YFD (aOR: 0.32, 95% CI: 0.19–0.52, p<0.001).

Considering the higher YFV positivity rate among vaccinated cases compared to the unvaccinated and the fact that being unvaccinated emerged as a protective factor in our final multivariate logistic regression analysis, we further disaggregated the dataset by vaccination status and analyzed the predictors of YFD by sub-group. The odds of YFD were higher among the unvaccinated and those with unknown vaccination status compared to those with previous YF vaccination (Table 3). Factors associated with YFD among those with unknown vaccination status were the first, third and last quarter of the year compared to the second quarter (OR: 4.27, 95% CI: 2.22–9.26; OR: 9.24, 95% CI: 4.95–19.80, and OR: 24.50, 95% CI: 13.38–51.60, p<0.001); the covid-19 pandemic period (OR: 1.84, 95% CI: 1.51–2.23; p<0.001); the rainy season (OR: 1.39, 95% CI: 1.12–1.71; p=0.003); male sex (OR: 1.45, 95% CI: 1.19–1.78; p<0.001); age being 5 years or older compared to being four years and below (5-9 year: OR: 2.16, 95% CI: 1.33–3.62; 10-14 years: OR: 2.60, 95% CI: 1.57–4.41; 15-29 years: OR: 4.37, 95% CI: 2.88-6.98; 30-44 years: OR: 2.96, 95% CI: 1.86–4.89; 45-59 years: OR: 3.56, 95% CI: 2.07–6.21; p<0.001); residing within the Derived Savannah, the Lowland/Freshwater/Mangrove rainforest, and the Guinea Savannah/Jos Plateau compared to the Sahel/Sudan Savannah (OR: 3.78, 95% CI: 2.80–5.16, OR: 2.01, 95% CI: 1.45–2.83, OR: 4.35, 95% CI: 3.15–6.07 respectively; p<0.001); being unemployed, student, and butcher/hunter/farmer (OR: 1.26, 95% CI: 1.01–1.57; OR: 2.40, 95% CI: 1.69–3.35; and OR: 3.33, 95% CI: 2.15–5.03; p<0.001); working in outdoor settings (OR: 2.21, 95% CI: 1.55–3.17; p<0.001); having traveled within last two weeks preceding onset (OR: 5.53, 95% CI: 3.45–8.65), presenting with fatigue, jaundice, and vomiting (OR: 1.97, 95% CI: 1.38–2.77, p<0.001; OR: 1.50, 95% CI: 1.12–2.02; p=0.007; and OR: 2.37, 95% CI: 1.55–3.56, p<0.001).

The predictors of YFD among those with previous YF vaccination were the first quarter of the year compared to the second (aOR: 4.04, 95% CI: 1.48–12.95; p=0.010), residing in the southern region (aOR: 14.03, 95% CI: 4.09–88.27; p<0.001). Clinical symptoms such as jaundice (aOR: 2.05, 95% CI: 0.88–5.25; p=0.110) and vomiting (aOR: 2.29, 95% CI: 0.75–6.25; p=0.120) were associated with YFD among the vaccinated but not statistically significant. Predictors among the unvaccinated were male sex (aOR: 6.35, 95% CI: 1.66–32.16; p=0.012), residing in the southern region (aOR: 76.64, 95% CI: 13.57–1465.89; p<0.001), working in outdoor settings (aOR: 3.11, 95% CI: 0.91–12.03; p=0.079), presenting with fatigue (aOR: 3.87, 95% CI: 1.10–15.08; p=0.039), hemorrhagic symptoms (aOR: 368.00, 95% CI: 14.29–14759.72; p<0.001), and vomiting (aOR: 6.53, 95% CI: 1.57–27.23; p= 0.009). Predictors of YFD among cases with unknown vaccination status were the third and fourth quarter of the year (aOR: 4.44, 95% CI: 1.92–12.11; aOR: 3.38, 95% CI: 1.08–10.90; p=0.001); age being 15 years or older (aOR: 1.87, 95% CI: 1.17–3.10, p=0.012);and residing within the Derived Savannah and the Guinea Savannah/Jos Plateau (aOR: 7.09, 95% CI: 3.35–17.45 and aOR: 5.24, 95% CI: 2.41–13.14 respectively; p<0.001) Fig 6.

• Download figure
• Open in new tab

Fig 6. Seasonal, geographical, socio-demographic, clinical predictors of YFD among reported cases in Nigeria, 2017-2021 disaggregated by vaccination status.

(A) Unvaccinated (B) Unknown vaccination status. The first column shows the explanatory variables, second column show the levels and third column is the adjusted odds ratios with corresponding confidence intervals.

Factors associated with being unvaccinated (barriers to YF vaccination)

Information on vaccination status was available for 3983 out 13014 reported cases, of which 1735 (13.3%) were vaccinated, 2248 (17.3) were unvaccinated, and 9031 (69.4%) were of unknown status (Table 1). The most unvaccinated age group were those 15-29 years, which also accounted for 30.0% of the total population, 32.0 % of unvaccinated cases, and 27.0% of cases of unknown vaccination status (Fig 4 and S1 Table). A comparison of the characteristics of the cases by vaccination status is shown in S1 Table. Independent predictors of being unvaccinated for YF, irrespective of laboratory test status, were the rainy season compared to the dry season (aOR: 1.29, 95% CI: 1.11-1.66, p=0.003); being 15years or older compared to under-five (15-29 years: aOR: 2.06, 95% CI: 1.51-2.83; 30-44 years: aOR: 2.11, 95% CI: 1.45-3.07; 45-59 years: aOR: 2.72, 95% CI: 1.63-4.58; 60+ years: aOR: 6.55, 95% CI: 2.76-17.50; p<0.001); residing in the northern region compared to the southern region (aOR: 3.71, 95% CI: 3.01-4.58, p<0.001); and occupation being either butcher, farmer or hunter compared to being home-based/officer worker (aOR: 2.30, 95% CI: 1.52-3.50, p<0.001). Working in outdoor setting was protective against being unvaccinated (aOR: 0.46, 95% CI: 0.31-0.69, p<0.001) (Table 4).

To further investigate the effect of the COVID-19 pandemic on YF vaccination, we disaggregated the analysis by laboratory test status. The COVID-19 pandemic period was associated with being unvaccinated among YFV positive (OR: 2.43, 95% CI: 0.96-6.48, p=0.066) and negative cases (OR: 2.06, 95% CI: 1.73-2.47, p<0.001) compared to the pre-pandemic period, just as working in outdoor settings in comparison to working indoors (OR: 3.81, 95% CI: 1.16-13.6, p=0.032) among YFV positive cases (Table 5).

More independent predictors were seen among YFV negative cases, which include the COVID- 19 pandemic period (aOR: 1.40, 95% CI: 1.14-1.72, p=0.002), age group being 15-29, 30-44, 60+ years (aOR: 1.81, 95% CI: 1.28-2.57; aOR: 1.79, 95% CI: 1.19-2.69; and aOR: 2.25, 95% CI: 1.03-5.02; p=0.001), residing in the northern region (aOR: 4.39, 95% CI: 3.58-5.40, p<0.001); and occupation being butcher/farmer/hunter, and health worker (aOR: 1.63, 95% CI: 1.08-2.47; aOR: 4.03, 95% CI: 1.33-17.47; p=0.021) (Fig 7).

• Download figure
• Open in new tab

Fig 7. Seasonal, geographical, and socio-demographic predictors non-vaccination among reported cases in Nigeria, 2017-2021 disaggregated by laboratory test status.

(A) Negative (B) Unconfirmed cases. The first column shows the explanatory variables, second column show the levels and third column is the adjusted odds ratios with corresponding confidence intervals.

Among YFV positive cases, independent predictors of being unvaccinated were COVID-19 pandemic period (aOR: 3.08, 95% CI: 0.86-12.55, p=0.09) and working in outdoor settings (aOR: 3.89, 95% CI: 1.13-14.64, p=0.03).

Limitation

This study represents the first attempt in Nigeria to use multi-season national surveillance data to characterize the seasonal, geographical, socio-demographic, and clinical characteristics associated with YFD and to identify barriers to vaccination. The dataset was generated from different health facilities across Nigeria, making it quite representative. However, just like every other surveillance system, the national YF surveillance in Nigeria is not without limitations. First, it is a passive, health facility-based surveillance system, which relies on the judgement of clinicians to identify cases. Although, under outbreak conditions, an active and community-based component of the surveillance system is activated. Considering the broad range of clinical presentation and symptoms associated with YF and the passive nature of the surveillance system, less severe cases may have been missed. However, clinicians and focal persons were trained regularly on using a WHO standard case definition to identify YF cases. Second, the national line list does not capture information on co-morbidities and pre-existing illnesses such as liver disease or previous toxic exposures such as alcohol consumption that may be confounding factors in YF disease. Third, there were issues related to data quality, including missing data, however, the development and deployment of the SORMAS platform, has strengthened the surveillance of diseases captured under the IDSR system, including YF in Nigeria, providing real time data entry and analysis. In any case, secondary data is useful for planning preventive measures, healthcare assessment, and research[41]. Findings from our study has the potential to be generalized as the dataset used had national representation, covering the six geo-political zones in Nigeria. However, it is important to note that the dataset used for the study was largely hospital-based. This implies that cases who felt they were not ill enough to access care in a health facility setting were completely missed by this study.