Understanding the drivers of continued mpox transmission in the United States: a modeling study

doi:10.21203/rs.3.rs-3817998/v1

Download PDF

Article

Understanding the drivers of continued mpox transmission in the United States: a modeling study

https://doi.org/10.21203/rs.3.rs-3817998/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

The 2022 mpox epidemic in the United States (U.S.) peaked in the summer of 2022, but over a year later, new cases continue to emerge. To understand why mpox continues to spread, albeit at low levels, we explored three mechanisms: (1) underdetection of cases, (2) waning immune protection from vaccination, and (3) waning immune protection from prior infection. We refined a multilayer epidemic network model to simulate these mechanisms and fit this model to the 2022 mpox outbreaks in Chicago, New York City, and San Francisco. Analyses across mechanisms suggest that underdetection was the most likely driver of prolonged transmission. Furthermore, the duration of simulated outbreaks was most sensitive to changes in the proportion of undetected cases. Our findings underscore the need for enhanced surveillance to capture the true scope of mpox transmission in the U.S., reduce the possibility of future resurgence, and enable effective outreach to key affected populations.

Health sciences/Diseases/Infectious diseases/Viral infection

Health sciences/Health care/Public health/Epidemiology

mpox

MPXV

monkeypox

transmission dynamics

Mpox is caused by infection with the monkeypox virus (MPXV). Historically, mpox outbreaks have been small in size and primarily driven by animal-to-human transmission.¹ The ongoing mpox pandemic marked a departure from our understanding of this zoonotic disease by the scale, geographic scope, and primary mode of transmission. As of December 2023, there have been 92,000 cases spanning 116 countries, and 82.9% of reported transmission modes compromised sexual encounters.²

In the United States, the 2022 mpox epidemic emerged in May 2022 and peaked in August at over 400 reported cases per day.³ The rapid decline in incidence was attributed to public health interventions aimed at harm reduction through temporary reductions in sexual activity, and vaccination of high-risk communities, primarily men who have sex with other men (MSM).⁴ While the mpox emergency was declared over at the end of January 2023, mpox transmission continues with nationwide incidence ranging from 7 to 63 cases per week through November 2023.^3,5 The continued transmission is surprising, given that past mpox outbreaks in nonendemic countries have been limited in magnitude with no evidence of long-term persistence or long tail of low-level transmission.^6–9 Travel history among a Chicago cluster in the Spring of 2023 revealed that only a small fraction reported recent travel to New York City, New Orleans, or Mexico, suggesting persistent transmission; however, the underlying drivers remain unclear. ¹⁰

Here, we explore three possible transmission drivers. The first is underdetection of infections in local communities. Factors such as stigma associated with mpox or self-identifying as MSM, as well as the broader lack of healthcare access could impact individual decision making in seeking care.^11,12 These barriers to care therefore impact our ability to accurately capture the true incidence within communities. Furthermore, several studies have revealed some individuals may incur asymptomatic infection, which can further drive underdetection.^13–17 Prior work has suggested that the underdetection of cases during an mpox outbreak can significantly increase the duration of an outbreak.¹³ The second and third drivers we evaluated were waning immunity from vaccination and infection, respectively. Evidence for waning immunity from vaccination for mpox is uncertain; however, studies of vaccinia virus have found waning immunity from partial and complete vaccination in under one and two years, respectively.¹⁸ Moreover, a recent mpox cluster among HIV-negative individuals in Chicago revealed 39% (n = 17) were fully vaccinated within the last year.¹⁹ Consistent with other sexually transmitted infections (STIs) that do not confer life-long immunity, reinfections among a core group of individuals could contribute to continued transmission within key affected populations.^20–22 For mpox, reinfections among healthy and non-immunocompromised individuals have been reported.^23–25

Given the limited empirical data on case detection rates or waning immunity, dynamic modeling allows us to investigate the potential of these possible mechanisms. To evaluate each hypothesis, we modified an existing transmission network model of mpox to represent each hypothesis and evaluated how well each model fit the data of three major U.S. cities from May 2022 to August 2023.^26,27

Study Design

We adapted a previously developed MSM contact network model²⁷ for mpox to include each of the aforementioned mechanisms. The model was independently fitted to data from Chicago, New York City, and San Francisco (May 2022-August 2023), the U.S. cities with the highest incidence in the 2022 outbreak. We compared the model fits from each city to assess the posited mechanisms and identify potential drivers of continued transmission.

Data Sources and Processing

Incidence and vaccination data were taken from the New York City Department of Health and Mental Hygiene GitHub Repository,²⁸ the Chicago Department of Health mpox dashboard,²⁹ and the City of San Francisco Data Depot.³⁰ Given the computational cost of simulating on a full-sized network, we scaled down the network to a fraction of each city’s total MSM population by reducing the number of nodes in the network to reflect a smaller, representative sample. To preserve accuracy in this scaled-down network, we proportionally adjusted the incidence and vaccination data. Since available data were absolute counts rather than proportions of the MSM population, it was necessary to first establish the total number of MSM in each city to serve as a denominator for these adjustments. We based our adjustments on previously published 2013 estimates of the MSM populations in each U.S. county.³¹ To ensure geographic alignment, we verified that the boundaries of each city aligned with those of an entire county. Furthermore, recognizing population growth, we projected these numbers to 2022, assuming a consistent 0.5% annual increase, in line with the general growth of the U.S. population.³² For 2022, we estimated the number of MSM to be 131,704 in Cook County, 225,236 in New York City, and 69,643 in San Francisco County. Yes, your revised text is clear and concise. During the outbreak, vaccination guidelines broadened to include various sexual and gender identities. In cities with vaccination data by sex and gender, we focused on the vaccinated proportion of cisgender men and adjusted our estimates accordingly. For NYC, we only had population estimates for Bronx, Queens, New York, and Kings Counties, excluding Richmond County (Staten Island: <1.5% of cases and < 10% of vaccination recipients). Lacking borough-specific time-series data, we utilized summary data to adjust our daily vaccination coverage and incidence estimates to only include the represented counties.

Epidemic Network Model

The epidemic network model²⁷ consisted of a system of 10,000 nodes (i.e., individuals in the model) that existed across three layered networks, each network capturing specific types of partnerships (Fig. 1a-b):³³ (1) One-time contacts (i.e., existing for a single timestep) were captured in a cross-sectional, exponential random graph model (ERGM).^27,34 Each node belonged to one of six sexual activity groups dictating the probability of one-time contacts in the cross-sectional network (Table 1). To capture reported reductions in sexual activity during the height of the mpox epidemic, we assumed a temporary reduction of 50% in one-time contacts between July 1, 2022 and September 30, 2022.³⁵ (2) Casual partnerships (i.e., not with a partner or significant other and lasting approximately 5.5 months on average) were captured in a separable-temporal random graph model (STERGM).^27,34 (3) Main partnerships, defined as those with a partner or significant other and lasting 15.9 months on average, were represented in a third STERGM network.^27,34 The use of a STERGM allowed the dissolution and formation of these partnerships to be captured as separate processes and allowed for the existence of temporal edges. Each node could have one to three types of partnerships, and the probability and duration of these partnerships were informed by previous descriptive network studies of MSM in the United States.^34,36

Table 1

Fixed parameters of the mpox network epidemic model: descriptions, values, and sources.
Parameters	Value	Source
Disease Related
Rate of moving from latent to infectious state ($ϵ$)	6.6 days^− 1	^27,37
Rate of moving from symptomatic to recovered (${\gamma }_{I}$)	8 days^− 1	²⁷
Rate of moving from undetected to recovered (${\gamma }_{A}$)	27 days^− 1	³⁸
Vaccination
Scaled daily rate of first dose vaccination (${\nu }_{1}$)*
Chicago	0–49 per 10,000 people per day	²⁹
New York City	0–470 per 10,000 people per day	²⁸
San Francisco	0–224 per 10,000 people per day	⁴⁴
Scaled daily rate of second vaccination (${\nu }_{2}$)*
Chicago	0–22 per 10,000 people per day	²⁹
New York City	0–228 per 10,000 people per day	²⁸
San Francisco	0–130 per 10,000 people per day	⁴⁴
Effectiveness of partial vaccination (${\psi }_{1}$)	37%	^45,46
Effectiveness of complete vaccination (${\psi }_{2}$)	67%	^45,46
Effectiveness of delayed complete vaccination (${\psi }_{D}$)	52%	Assumed
Network Demography
Population size	10,000	Assumed
Age-range	18–40	Assumed
Sexual Partnerships and Acts
Daily probability of sexual acts in main partnerships	0.22	^27,34,36,47
Daily probability of sexual acts in casual partnerships	0.14	^27,34,36,47
Daily probability of sexual acts in one-time partnerships	1.0	^27,34,36,47
Probability of one-time sexual partnerships across the sexual activity stratum	0-0.286	^27,34,36,47
Temporary reduction in one-time contacts	50%	^35,48
Time period of temporary reduction	July 1, 2022 – September 30, 2022	Assumed
Initial Conditions
Initial number of infections	10	Assumed
Initial partial vaccination coverage	0%	Assumed
Initial complete vaccination coverage	0%	Assumed

Nodes in the network had a disease-state attribute that followed the natural history of MPXV infection and mpox disease progression (Fig. 1c). At the beginning of each simulation, except the seeded infections, all individuals were susceptible (S). Infected nodes then moved from the susceptible state to a latent ($E$) state. Individuals who were exposed moved to a symptomatic and infectious state ($I$) and eventually recovered ($R$). The rate of moving from the E to I states was informed both by prior modeling work and more recent estimates of the incubation period for mpox in the U.S.^27,37 Specifically we took the mean of the incubation period until symptom onset (5.6 days) and rash onset (7.5 days).³⁷ Susceptible individuals who remained uninfected could be randomly sampled for vaccination.

These individuals moved through the vaccination process from susceptible to partially (${V}_{1}$) vaccinated and then fully (${V}_{2}$) vaccinated. For vaccination, we assumed a leaky vaccine mechanism in which vaccinated individuals could still become infected:

$${\beta }_{{V}_{i} }=\beta \times \left(1-V{E}_{i}\right)$$

where ${\beta }_{{V}_{i} }$ is the probability of infection among those vaccinated ${V}_{i}$, computed as the product of the transmission rate ($\beta$) and the effectiveness of the i-th dose of the vaccine ($1-V{E}_{i}$).

Hypothesized Mechanisms

To represent each mechanism, we altered the corresponding disease and/or vaccination process of the base epidemic model and kept all other features and/or parameters the same as described above.

(1) Underdetection:

Some proportion ($\theta$) of cases could go undetected (Fig. 1c) and transition to an undetected but infectious state ($U$). We assumed undetected individuals were unaware of infectious status and remained sexually active for the full period of infectiousness (Table 1).³⁸

(2) Vaccine-Derived Immunity Wanning:

In the second mechanism, individuals could transition from a vaccinated to a waned and fully susceptible state (${W}_{1}, {W}_{2}$). The probability of vaccine waning was presumed to follow a logistic distribution, which informed the likelihood of waning for each individual based on the time elapsed since their vaccination. Individuals losing partial vaccination immunity could later receive a second dose, entering a delayed full vaccination state (V_D2) with a slightly reduced VE assumed to be the average of partial and full VE.

(3) Infection-Derived Immunity Wanning:

In the third mechanism, we assumed that all individuals who recover from infection return to fully susceptible at an infection-derived immunity waning rate (ξ).

Epidemic Simulation

To initiate the epidemic, ten nodes were randomly selected from two strata with the highest sexual activity and were set as symptomatic and infected on the first day of the simulation. We did not account for any importation of subsequent cases into the system (i.e., the system was closed). All simulations were run on a daily timestep for 456 days (May 1, 2022 to July 31, 2023). Simulations were run through the EpiModel and EpiModel HIV packages in R,^39,40 on an Amazon Linux server hosted on Amazon Web Services.

Model Evaluation

Simulated incidence (i.e., symptomatic cases) was fit to the scaled incidence of each city. For all mechanisms, we explored a search grid that included the transmission probability per sexual contact ($\beta$) ranging from 0.3 to 0.9 in 0.1 increments. For the first mechanism, we also explored the fraction undetected (θ), with possible values ranging from 0.0 to 0.9 in 0.1 increments. For the second mechanism, we also explored the shape parameter ($\mu$) of the probability density functions for partial and full vaccination, while the scale parameter was fixed to 20. We explored values of $\mu$ (corresponding to the mean vaccine-derived immunity period) that equated to 6–16 months in 1-month increments. Only combinations where the time to waning from full vaccination was greater than partial vaccination were explored. For the third mechanism, we explored the infection-derived immunity period (1/ξ) ranging from 6 to16 months, in approximately 1-month increments. Each parameter combination was run 100 times, and stochastic runs with no outbreak were omitted (i.e., cumulative incidence less than twice the initial number of seed infections). The mean, median, and percentile distribution of each parameter set was computed. Fits were assessed by comparing the root mean square error (RMSE) and coefficient of determination (R²) between the simulated and observed incidence. Fits with the smallest RMSE and highest R² were chosen while also considering the relative size of the simulated epidemic peak compared to the observed peak.

Estimating R₀ and R_e

The basic reproductive number (${R}_{0}$) was estimated as the average number of secondary infections among the first generation of infections (i.e., when the population was still largely susceptible, and no behavior changes were adopted). The R₀ was computed for each stochastic run and summarized across all runs. We repeated this process among the first two and three generations to see how sensitive our estimates were relative to the generations selected. In a second approach, we approximated ${R}_{0}$by capturing the maximum effective reproductive number (${R}_{e}$) between May and December 2022, a period encompassing most infections during the outbreak. For each day, the ${R}_{e}$ was estimated by averaging the number of secondary infections among nodes that recovered on that particular day. To smooth the results, we then took a 7-day moving average of ${R}_{e}$ for each stochastic run. Finally, we summarized the ${R}_{e}$ for each mechanism and city, at every timestep and across all simulations.

Statistical Analysis

We assessed the duration of outbreaks across mechanisms by comparing outbreak duration within each city using the Wilcoxon rank-sum test. To account for the multiple pair-wise comparisons across mechanisms we employed a Bonferroni correction. To evaluate the sensitivity of outbreak duration to the fitted parameters (θ, ω₁, ω₂, ζ, β), we conducted a partial rank correlation coefficient analysis, estimating the associations between these parameters and outbreak duration, while adjusting for other non-fixed parameters.

Model Evaluations and Parameter Selection

Across all modeled mechanisms, combinations with a 70–90% transmission probability generally exhibited smaller errors and better explained the variance in the data (Fig. 2; Supplemental Fig. S1-S4). For each city, the model including underdetection performed the best, with often the least error and best fit when compared to observed weekly incidence (Fig. 2a-c). The explained variance of models with underdetection ranged from 76–84% for all three cities, compared to 52–75% for the other mechanisms. Furthermore, models with underdetection produced longer outbreaks than other mechanisms, with the median last infection was between March and May 2023 for all three cities, compared to the other mechanisms where the median last infection was in October 2022 (Fig. 2).

The Proportion of Undetected Cases

Based on the best-fitting model, the proportion of undetected cases was estimated to be 90% in Chicago and San Francisco (Fig. 2a,c) and slightly lower for New York City at 70% (Fig. 2b). We extracted and summarized the number of new undetected cases per week on the basis of the best-fitting model (Fig. 3). For all three cities, the mean weekly incidence of undetected cases persisted above one until the late Spring and early summer of 2023.

Epidemiological Parameters

We estimated the R₀ for all three cities based on the best-fitting models (Fig. 4a). Estimates were consistent between cities, each below unity (i.e., one) and accompanied by wide confidence intervals. Our sensitivity analysis showed that estimates from the first two or three generations exceeded unity but covered a longer time frame, violating the assumption of a near-fully susceptible population (Supplemental: Table S1). Our second approach of approximating the R₀ yielded higher point estimates that were all above unity, but with equally wide confidence intervals (Fig. 4a).

The R_e was estimated to initially peak in early-to-mid June 2022 in each city, a month behind the peak of detected and undetected cases (Fig. 4b). By September 2022, the R_e for all three cities appeared to drop below unity and fluctuated above and below this threshold thereafter. By the end of July 2023, although the R_e in all three cities continued to hover around one, the weekly incidence of both reported and unreported cases had dropped substantially. Given the lower incidence towards the end of the simulation, the fluctuations in R_e were more pronounced.

Sensitivity of the Length of Outbreak to Proportion Undetected

The length of the outbreak was most sensitive to changing the proportion of undetected cases (Fig. 5). It was evident that increasing the proportion of undetected cases ($\theta$) was associated with outbreaks that persisted into the summer of 2023 (Fig. 5). The longest outbreak across all three cities was for 80% undetected cases with a median duration of 374 days (IQR: 269, 436). In contrast, the sensitivity to the other two mechanisms and their respective parameters (${\omega }_{1}, {\omega }_{2},\zeta$) was negligible (Supplemental: Fig. S5). The infection parameter for all three mechanisms demonstrated a trend of higher values resulting in longer outbreaks (Supplemental: Fig. S6). Finally, to assess the strengths of these relationships, we computed the partial rank correlation coefficients of each of these parameters on the duration of the outbreak which confirmed the trends observed in the figures (Supplemental: Table S2).

Our analysis suggests that underdetection of mpox cases may be the primary driver behind the prolonged mpox outbreaks in several U.S. cities, characterized by a large initial peak followed by a sustained low-level incidence into the subsequent year. In contrast, the waning of vaccine- or infection-derived immunity alone was less capable of explaining the observed epidemics in each city. Notably, the duration of the outbreak was most sensitive to increasing the proportion of undetected cases, whereas decreasing the length of protection for vaccine- or infection-derived immunity had negligible effects on outbreak duration.

Our estimations revealed strong agreement in the proportion of undetected cases—90% in Chicago and San Francisco versus 70% in New York City. Underdetection of cases may be driven by inadequate mpox surveillance. Enhanced surveillance is needed to capture the full extent of community-level transmission. For example, current testing recommendations are limited to those with specific symptoms or known exposures, which may overlook asymptomatic or mildly symptomatic cases that go undiagnosed.³⁸ Expanding testing criteria could address some of the underdetection. This could include mpox screening for key affected populations during routine STI check-ups, a practice some health departments have begun to adopt, ⁴¹ particularly in areas with high mpox incidence. In addition, contact tracing remains crucial in capturing chains of transmission. However, the primary diagnostic approach, PCR testing of lesion swabs, is not applicable to individuals without symptoms, highlighting the need for diagnostic advancements.

Wastewater surveillance offers a less biased, population-level view of transmission, complementing traditional surveillance methods. The National Wastewater Surveillance System currently produces wastewater data on mpox across 650 + U.S. sites, predominantly in California. Monthly sample submissions have varied widely, from 1 to 313.⁴² Enhancing this approach requires more frequent, consistent sample collections and the expansion of the network geographically. Such robust surveillance could better reveal the geographic spread of transmission and the extent of underdetection, while potentially serving as an early warning system for future largescale mpox outbreaks.

There are several limitations in this study. We assumed that undetected cases remained sexually active throughout their infectious period, although in reality some may have reduced activity due to mild symptoms without a formal diagnosis. Secondly, our model only captured male-identifying MSM, excluding other sexual and gender minority groups affected by the outbreak.⁴³ Third, our model assumed a static population and demographics appropriate for the short timescale studied but neglected the potential impact of large influxes of travelers nor does it consider the recruitment of new susceptibles into the population. Fourth, we assumed a leaky mechanism for the vaccine but did not explore an alternative (i.e., an all-or-nothing mechanism). Fifth, given the computational cost of running these models, we explored only a limited set of parameters on a scaled network size. Other parameters may contribute to continued transmission, and the size of the network may need to be increased for greater power to detect sporadic incidence post-epidemic peak. Furthermore, we only considered each mechanism independently and not in combination. We assumed the proportion of undetected cases was a fixed value and not time-varying. Furthermore, there was an absence of data regarding the volume of tests conducted in each city, and therefore we cannot empirically evaluate the efficacy of testing strategies between cities. Lastly, we assumed only a temporary change in behavior during the epidemic peak and did not consider the seasonality of behavior.

Our findings suggest that the persistence of mpox within the U.S. may be due largely to the underdetection of cases. To uncover and address ongoing transmission, surveillance methods must be improved. As the perceived risk of acquiring mpox diminishes, it's imperative for public health entities to persist in educating at-risk populations about the varying manifestations of mpox and to encourage testing when infection is suspected. Although our results indicate that diminished vaccine- or infection-derived immunity had a lesser impact on transmission persistence, the limited data on both immunity mechanisms call for a more thorough investigation to fully understand the disease's trajectory and understand the potential for future outbreaks.

Data and Coding Availability: Data for mpox incidence and vaccination uptake were obtained from publicly available health dashboards or technical reports. For New York City, incidence data were taken from, and vaccination data were taken from https://github.com/nychealth/mpv-vaccine-data. For Chicago, data for both incidence and vaccination were taken from https://www.chicago.gov/city/en/sites/monkeypox/home/data.html. For San Francisco, incidence and vaccination data were taken from https://sf.gov/resource/2022/mpox-data-and-reports. R scripts to run this model and recreate our analyses can be found in the dedicated GitHub repository: https://github.com/CDDEP-DC/mpox-modeling-study.

Conflicts of Interest: The authors declare that they have relevant conflicts of interest.

Acknowledgments: This study was supported in part by the CDC MInD-Healthcare Network (grant number U01CK000589).

Van Dijck, C. et al. Emergence of mpox in the post-smallpox era - a narrative review on mpox epidemiology. Clinical Microbiology and Infection S1198743X23003890 (2023) doi:10.1016/j.cmi.2023.08.008.
2022-23 Mpox (Monkeypox) Outbreak: Global Trends. https://worldhealthorg.shinyapps.io/mpx_global/ (2023).
2022 Outbreak Cases and Data | Mpox | Poxvirus | CDC. https://www.cdc.gov/poxvirus/mpox/response/2022/index.html (2023).
Intervention Services for People with or Exposed to Mpox | Mpox | Poxvirus | CDC. https://www.cdc.gov/poxvirus/mpox/health-departments/intervention-services.html (2023).
Affairs (ASPA), A. S. for P. Statement From HHS Secretary Becerra on mpox. HHS.gov https://www.hhs.gov/about/news/2022/12/02/statement-from-hhs-secretary-becerra-on-mpox.html (2022).
Mpox (monkeypox): background information. GOV.UK https://www.gov.uk/guidance/monkeypox (2023).
Reynolds, M. G. et al. Clinical Manifestations of Human Monkeypox Influenced by Route of Infection. The Journal of Infectious Diseases 194, 773–780 (2006).
Past U.S. Cases and Outbreaks | Mpox | Poxvirus | CDC. https://www.cdc.gov/poxvirus/mpox/outbreak/us-outbreaks.html (2023).
Pages - Travel-Associated Monkeypox virus infection confirmed in Maryland resident. Maryland.gov Enterprise Agency Template https://health.maryland.gov/newsroom/Pages/default.aspx.
RESURGENCE OF MPOX - Provider Update: May 9, 2023. (2023).
Turpin, R. E. et al. Monkeypox-Related Stigma and Vaccine Challenges as a Barrier to HIV Pre-Exposure Prophylaxis among Black Sexual Minority Men. Int J Environ Res Public Health 20, 6324 (2023).
Keum, B. T., Hong, C., Beikzadeh, M., Cascalheira, C. J. & Holloway, I. W. Mpox Stigma, Online Homophobia, and the Mental Health of Gay, Bisexual, and Other Men Who Have Sex with Men. LGBT Health 10, 408–410 (2023).
Van Dijck, C., Hens, N., Kenyon, C. & Tsoumanis, A. The Roles of Unrecognized Mpox Cases, Contact Isolation and Vaccination in Determining Epidemic Size in Belgium: A Modeling Study. Clinical Infectious Diseases 76, e1421–e1423 (2023).
Hampel, B. et al. Low prevalence of asymptomatic mpox in populations at high risk. The Lancet Microbe S2666524723002483 (2023) doi:10.1016/S2666-5247(23)00248-3.
De Baetselier, I. et al. Retrospective detection of asymptomatic monkeypox virus infections among male sexual health clinic attendees in Belgium. Nat Med 28, 2288–2292 (2022).
Cassir, N. et al. Observational Cohort Study of Evolving Epidemiologic, Clinical, and Virologic Features of Monkeypox in Southern France. Emerg. Infect. Dis. 28, (2022).
Mizushima, D. et al. Prevalence of Asymptomatic Mpox among Men Who Have Sex with Men, Japan, January-March 2023. Emerg Infect Dis 29, 1872–1876 (2023).
Ilchmann, H. et al. One- and Two-Dose Vaccinations With Modified Vaccinia Ankara-Bavarian Nordic Induce Durable B-Cell Memory Responses Comparable to Replicating Smallpox Vaccines. J Infect Dis 227, 1203–1213 (2023).
Faherty, E. A. G. Notes from the Field: Emergence of an Mpox Cluster Primarily Affecting Persons Previously Vaccinated Against Mpox — Chicago, Illinois, March 18–June 12, 2023. MMWR Morb Mortal Wkly Rep 72, (2023).
Yorke, J. A., Hethcote, H. W. & Nold, A. Dynamics and Control of the Transmission of Gonorrhea: Sexually Transmitted Diseases 5, 51–56 (1978).
Hethcote, H. W. & Yorke, J. A. Gonorrhea Transmission Dynamics and Control. vol. 56 (Springer, 1984).
Gesink, D. C., Sullivan, A. B., Miller, W. C. & Bernstein, K. T. Sexually Transmitted Disease Core Theory: Roles of Person, Place, and Time. American Journal of Epidemiology 174, 81–89 (2011).
Golden, J. et al. Case of apparent mpox reinfection. Sex Transm Infect 99, 283–284 (2023).
Musumeci, S. et al. A Case of Mpox Reinfection. Clinical Infectious Diseases 77, 135–137 (2023).
Raccagni, A. R. et al. Two individuals with potential monkeypox virus reinfection. Lancet Infect Dis 23, 522–524 (2023).
Yang, W. & Shaman, J. Development of a model-inference system for estimating epidemiological characteristics of SARS-CoV-2 variants of concern. Nat Commun 12, 5573 (2021).
Spicknall, I. H. et al. Modeling the Impact of Sexual Networks in the Transmission of Monkeypox virus Among Gay, Bisexual, and Other Men Who Have Sex with Men — United States, 2022. MMWR Morb. Mortal. Wkly. Rep. 71, 1131–1135 (2022).
NYC mpox Vaccination Data.
Mpox Data Dashboard. https://www.chicago.gov/content/city/en/sites/monkeypox/home/data.html.
Mpox data and reports | San Francisco. https://sf.gov/resource/2022/mpox-data-and-reports.
Grey, J. A. et al. Estimating the Population Sizes of Men Who Have Sex With Men in US States and Counties Using Data From the American Community Survey. JMIR Public Health and Surveillance 2, e5365 (2016).
Bureau, U. C. Growth in U.S. Population Shows Early Indication of Recovery Amid COVID-19 Pandemic. Census.gov https://www.census.gov/newsroom/press-releases/2022/2022-population-estimates.html.
Cimini, G. et al. The statistical physics of real-world networks. Nat Rev Phys 1, 58–71 (2019).
Weiss, K. M. et al. Egocentric sexual networks of men who have sex with men in the United States: Results from the ARTnet study. Epidemics 30, 100386 (2020).
Delaney, K. P. Strategies Adopted by Gay, Bisexual, and Other Men Who Have Sex with Men to Prevent Monkeypox virus Transmission — United States, August 2022. MMWR Morb Mortal Wkly Rep 71, (2022).
Jenness, S. M. et al. Impact of the Centers for Disease Control’s HIV Preexposure Prophylaxis Guidelines for Men Who Have Sex With Men in the United States. J Infect Dis 214, 1800–1807 (2016).
Madewell, Z. J. et al. Serial Interval and Incubation Period Estimates of Monkeypox Virus Infection in 12 Jurisdictions, United States, May–August 2022. Emerg. Infect. Dis. 29, 818–821 (2023).
Clinical Recognition | Mpox | Poxvirus | CDC. https://www.cdc.gov/poxvirus/mpox/clinicians/clinical-recognition.html (2023).
Jenness, S. M., Goodreau, S. M. & Morris, M. EpiModel: An R Package for Mathematical Modeling of Infectious Disease over Networks. Journal of Statistical Software 84, 1–47 (2018).
EpiModelHIV. (2023).
mariagomez. Integrating Mpox into Sexual Health and HIV Care (VDH). Mpox https://www.vdh.virginia.gov/monkeypox/integrating-mpox-vaccination-into-care/ (2023).
U.S. Mpox Wastewater Data | National Wastewater Surveillance System | CDC. https://www.cdc.gov/nwss/wastewater-surveillance/mpox-data.html (2023).
Mpox Cases by Age and Gender, Race/Ethnicity, and Symptoms | Mpox | Poxvirus | CDC. https://www.cdc.gov/poxvirus/mpox/response/2022/demographics.html (2023).
Mpox Vaccinations Given to SF Residents Over Time.
Deputy, N. P. et al. Vaccine Effectiveness of JYNNEOS against Mpox Disease in the United States. N Engl J Med 388, 2434–2443 (2023).
Jynneos Vaccine Effectiveness. https://www.cdc.gov/poxvirus/mpox/cases-data/JYNNEOS-vaccine-effectiveness.html (2023).
Hernández-Romieu, A. C. et al. Heterogeneity of HIV Prevalence Among the Sexual Networks of Black and White Men Who Have Sex With Men in Atlanta: Illuminating a Mechanism for Increased HIV Risk for Young Black Men Who Have Sex With Men. Sex Transm Dis 42, 505–512 (2015).
Clay, P. A. et al. Modelling the impact of vaccination and sexual behavior change on reported cases of mpox in Washington D.C. http://medrxiv.org/lookup/doi/10.1101/2023.02.10.23285772 (2023) doi:10.1101/2023.02.10.23285772.

There is NO Competing Interest.

UnderstandingthedriversofcontinuedmpoxtransmissionintheUnitedStatesamodelingstudySupplement.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Understanding the drivers of continued mpox transmission in the United States: a modeling study

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study Design

Data Sources and Processing

Epidemic Network Model

Hypothesized Mechanisms

(1) Underdetection:

(2) Vaccine-Derived Immunity Wanning:

(3) Infection-Derived Immunity Wanning:

Epidemic Simulation

Model Evaluation

Estimating R₀ and R_e

Statistical Analysis

Results

Model Evaluations and Parameter Selection

The Proportion of Undetected Cases

Epidemiological Parameters

Sensitivity of the Length of Outbreak to Proportion Undetected

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

Understanding the drivers of continued mpox transmission in the United States: a modeling study

Status:

Version 1

Abstract

Figures

Introduction

Methods

Study Design

Data Sources and Processing

Epidemic Network Model

Hypothesized Mechanisms

(1) Underdetection:

(2) Vaccine-Derived Immunity Wanning:

(3) Infection-Derived Immunity Wanning:

Epidemic Simulation

Model Evaluation

Estimating R0 and Re

Statistical Analysis

Results

Model Evaluations and Parameter Selection

The Proportion of Undetected Cases

Epidemiological Parameters

Sensitivity of the Length of Outbreak to Proportion Undetected

Discussion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1

Estimating R₀ and R_e