Introduction
Monkeypox (Mpox) is a relatively rare zoonotic disease,caused by the Mpox virus, a virus closely related to the variola virus (responsible for the small-pox disease) by belonging to the same genus of Orthopoxviruses [
1]. As suggested by the name, the monkeypox virus was first discovered during an outbreak amongst monkeys at a Danish laboratory in 1958 [
2], but it was not until 1970 when the first human, a 9-month-old baby, was diagnosed in the Democratic Republic of Congo (formerly Zaire) [
3], and since then been referred to as human Mpox virus.
Mpox has been endemic to West and Central Africa, with the most affected country being the Democratic Republic of Congo (DRC), where regular outbreaks have been the norm for the past five decades [
1,
4,
5]. More recently the WHO reported 4,594 new suspected cases between January and September 2020, in the DRC alone, suggesting a steady rise in incidence [
6]. This was followed by cases being reported in other parts of the world with around 3413 Mpox virus infections being reported across 50 countries. This led to the WHO declaring Mpox as an “evolving threat of moderate public health concern” in June 2022 [
7,
8].
Till the 1980s transmission to humans originated from contact with wildlife reservoirs [
9,
10]. More recently, most cases outside of Africa were due to animal-to-human transmission, imported from endemic countries, or associated with imported pets [
1,
11,
12]. Only in the 1990s when the number of secondary cases by contact with an infected person began to increase was Mpox considered an important worldwide health concern [
10]. The transmission of the virus occurs mostly through large respiratory droplets, close or direct contact with skin lesions, and possibly through contaminated fomites [
7,
13]. Vertical transmission and fatal deaths have also been described [
14]. The current spread has been shown to disproportionately affect men who are gay or bisexual and other men who have sex with men (GBMSM), which may suggest amplification of transmission through sexual networks [
15]. According to the UK Health Security Agency, of the 152 male confirmed reported cases 151 were identified as GBMSM [
16]. At present, it is still not clear whether Mpox can be transmitted through semen or vaginal fluid.
Although not as severe, Mpox disease shares many clinical characteristics with the smallpox disease such as an initial febrile prodrome lasting between 1 to 4 days with generalized headaches and fatigue. The initial prodromal period is followed by (or concomitant with) the development of a maculopapular rash, that often first appears on the face and then appears in a centrifugal distribution on the body [
17,
18]. These lesions may occur in the oral cavity causing difficulty swallowing [
17]. The disease also characteristically results in maxillary, cervical, or inguinal lymphadenopathy which is unique when compared to smallpox and suggests a more effective immune recognition and response (a hypothesis that requires further study) [
17,
19]. Smallpox (
Variola majorvariant) had a case fatality rate of 30%, fortunately the symptoms from Mpox disease are much less severe and self-limiting lasting with symptoms usually resolving within 2 to 4 weeks. Severe cases usually only appear in children and immunocompromised. Complications are rare but include pneumonitis, encephalitis, keratitis, and secondary bacterial infections [
7].
The clinical differential diagnosis includes other rash illnesses, such as chickenpox, measles, bacterial skin infections, scabies, syphilis, and medication-associated allergies. All suspected cases of Mpox disease should be reported immediately to a local health department for proper infection control and contact tracing but given the current rarity of the disease and wide range of clinical differential diagnoses, reaching a diagnosis of Mpox poses a challenge for physicians.
Combined with clinical and epidemiological information diagnostic assays are the most powerful and important components for the identification of Orthopoxviruses as recommended by the WHO [
19,
20]. McCollum et al [
17] discussed the pros and cons of multiple diagnostic assays, the most reliable of which is real-time polymerase chain reaction (PCR). These assays are highly sensitive and efficiently detect viral DNA but require high-quality laboratories that either use in rural and low-resource regions [
17,
20].
The COVID-19 outbreak in 2019 has changed the way we view zoonotic infections but experts have been warning the public about the threat of zoonotic infections as far back as 2003 during the 2003 SARS outbreak [
21]. The fear of another pandemic has led to taking measures to control the disease early on. Statistical analyses play an important role in the prediction of disease spread and can help prepare and control outbreaks through planning and policies. Since its inception as a field of study more than a century ago, infectious disease epidemiology has placed a high priority on the statistical representation and analysis of infectious diseases [
22,
23]. In recent years, in particular, for newly emerging disease outbreaks, forecasting modelling is in great demand and significantly contributes to emerging disease outbreaks and public health [
8,
24,
25]. The goals of modelling include identifying epidemiological characteristics to comprehend infectious diseases, forecasting disease trends, assessing control measures to guide decision-making, and investigating uncertainty. To study the spread of infectious diseases, numerous models have been developed, examined, and used [
8,
26‐
31], a few cited therein.
The time series modelling has long and rich history in epidemiological pursuits. For example [
32], studied the dynamics of influenza epidemics in space and time where disease counts were considered to follow multivariate autoregressive process. Further [
28], proposed a dynamic model based on SIR-type differential equation to enumerate impact of early health interventions in context of COVID-19 pandemic. Furthermore, SIR model was used by [
26] to predict that H1N1 (the swine flu) would pose a serious threat to Israel's public health. Moreover [
29], used the Poisson-lognormal, Poisson-generalized Gamma, and Poisson-Weibull distributions to enumerate the spread proportion of COVID-19 in Hong Kong, India, and Rwanda. Also, using a zero-truncated negative binomial model [
33], conducted a study to infer the super spreading potential COVID-19 flow around the globe. Additionally [
34], proposed a new zero-state coupled Markov switching negative binomial model in which the disease alternates between periods of presence and absence in each area using a series of partially hidden nonhomogeneous Markov chains coupled between nearby locations. The distribution of COVID-19 confirmed cases in China was examined by [
35] With respect to the dynamics of Power law.
Mpox cases, on the same lines, are time-series data with some dynamic fluctuation trend in the various circumstances with epidemic prevention and control, making it appropriate to create a time-series model for prediction. Predicting the daily new cases and total confirmed cases of Mpox for all the most affected countries is therefore extremely important from a practical standpoint.
However, as time-series data, Mpox cases have some dynamic fluctuation trends in the various situation with epidemic prevention and control, which is suitable for establishing a time-series model for prediction. The Automatic Regressive Integrated Moving Average (ARIMA) model, which has a simple structure and immediate applicability, is one of the most popular time series models. The capability of ARIMA model in extracting the trends in the data by considering moving averages and then obtain of the stationarity of the series by differentiating, is well documented in research literature [
36]. The ARIMA model has been widely used to predict and estimate the prevalence of common diseases, including COVID-19 [
37,
38], typhoid fever [
39], tuberculosis [
22], and influenza [
23,
40,
41]. ARIMA methods are capable of correlating regulation with short-term changing trends in time series despite their lack of reliance on mathematics and statistics. Therefore, the model is more suitable for predicting short-term epidemic diseases.
Therefore, it is of great practical significance to predict the daily new cases and cumulative confirmed cases of Mpox for all the most affected countries. This study develops best-fitted ARIMA models to predict daily new cases and cumulative confirmed cases of monkeypox in Spain, the United States, Germany, the United Kingdom, France, the Netherlands, Colombia, Mexico, Brazil, and Canada over the next 20 days and evaluates the model's prediction accuracy to provide a further reference for the prediction and early warning of infectious diseases. These models may also be used to predict the more days by incorporating the more days of data.
Conclusions
In this study, the 10 most affected countries— United States, Brazil, Spain, France, Colombia, Mexico, Peru, United Kingdom, Germany and Canada—were examined about the current and short-term predicted possible daily confirmed and cumulative cases of the Mpox epidemic. The persistent trend and scope of the epidemic were estimated using ARIMA models. It has been revealed that, among other countries, the United States will most likely be affected by Mpox in the future, prompting people to be more vigilant of this virus.
The current work can help respected governments develop emergency plans and allocate medical resources. Whereas in this study, authors used data from almost three months to forecast the next twenty-day scenario. If the data set is sizable, it can also accurately predict long periods.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.