Evidence for influenza and RSV interaction from 10 years of enhanced surveillance in Nha Trang, Vietnam, a modelling study.

Influenza and Respiratory Syncytial Virus (RSV) interact within their host posing the concern for impacts on heterologous viruses following vaccination. We aimed to estimate the population level impact of their interaction. We developed a dynamic age-stratified two-pathogen mathematical model that includes pathogen interaction through competition for infection and enhanced severity of dual infections. We used parallel tempering to fit its parameters to 11 years of enhanced hospital-based surveillance for acute respiratory illnesses (ARI) in children under 5 years old in Nha Trang, Vietnam. The data supported either a 41% (95%CrI: 36-54) reduction in susceptibility following infection and for 10.0 days (95%CrI 7.1-12.8) thereafter, or no change in susceptibility following infection. We estimate that co-infection increased the probability for an infection in <2y old children to be reported 7.2 fold (95%CrI 5.0-11.4); or 16.6 fold (95%CrI 14.5-18.4) in the moderate or low interaction scenarios. Absence of either pathogen was not to the detriment of the other. We find stronger evidence for severity enhancing than for acquisition limiting interaction. In this setting vaccination against either pathogen is unlikely to have a major detrimental effect on the burden of disease caused by the other.

Introduction

Influenza and Respiratory Syncytial Virus (RSV) have large health and economic impacts globally, particularly in young children where they cause 870 000 [1] and 3.2 million [2] hospitalisations in <5 year olds per year respectively. While pediatric influenza vaccines are licensed for use in some countries, global uptake is poor and efficacy depending on the match to the circulating strains. RSV vaccines are in development, with close to 20 vaccine candidates being evaluated in pre-licensure trials [3].

The impact of vaccination may be enhanced if concurrent co-infections increase the propensity of severe disease beyond that of either pathogen [4]. However, the impact of vaccination may be lessened if vaccination reduces competitive pressure between influenza and RSV and thus leads to increased circulation of the other pathogen. Such competitive pressure has been observed in the form of cross protection in mouse studies that showed e.g. a protective effect of live attenuated influenza vaccine administration on RSV replication [5] and influenza infection on RSV severity [6], mediated by the innate immune system. Population level evidence for the effect of cross-protection on influenza and RSV epidemiology, however, is largely of observational nature: a lack of coincidence in peak timings [7,8], changes in RSV peak timing following unusual influenza seasons [9-14] and alternating infection patterns [15].

Cross-protection following a primary infection could occur through a variety of mechanisms including viral competition for resources in the host [16], the activation of the innate immune system such as through toll-like receptors (TLRs) 3 and 7 [17,18] or short term immune memory through surviving cells in an antiviral state (e.g. epithelial cells following influenza infection [19]). These interactions could result in a reduction in subsequent infection with the other viruses, and estimates of the duration of cross-protection and its biological pathway vary. Experimental infection of ferrets estimated less than 2 weeks protection between influenza A and B viruses [20], yet cells forming the respiratory epithelium can survive in a state of heightened antiviral activation for 3 to 12 weeks after influenza A infection, with waning of the conferred protection observable at 6 weeks [19].

In Nha Trang, Vietnam, for more than 10 years children admitted to the single public hospital with acute respiratory illness have been tested for presence of influenza and RSV infection as part of an enhanced surveillance. RSV circulation is highly seasonal, and influenza circulation varies year on year, thus giving a unique opportunity to systematically investigate evidence for their interaction at population level. We use this data in combination with a dynamic transmission model to estimate the strength of cross-protection and the effects of co-infections on disease severity (defined as the proportion of infections that require hospital attendance).

Methods

Ethics statement

This study was approved by the Institutional Review Boards of the London School of Hygiene & Tropical Medicine (16166 /RR/12988) and the National Institute of Hygiene and Epidemiology in Vietnam (VN01057-28/2015).

Study population

We used data from a hospital-based enhanced surveillance study of children with respiratory disease, as previously described [21,22]. In brief, children younger than 5 years old who resided in 16 out of the 27 communes of Nha Trang (resident population of 210’739) and attended he paediatric ward in Khanh Hoa General Hospital (KHGH) in Nha Trang, central Vietnam, with Acute Respiratory Infection (ARI) were enrolled and offered a suite of additional diagnostics. ARI was defined as Cough and / or difficulty breathing. Khanh Hoa hospital is a tertiary care facility and is the only public hospital for the catchment area of the study. More than 94% of all pediatric ARI admissions are typically enrolled. Upon admission, Nasopharyngeal (NP) samples were taken from patients, nucleic acid was extracted and multiplex-PCR assays were performed in order to detect infection with up to 13 respiratory viruses, including influenza A and RSV. Positive samples underwent a second, confirmatory PCR test and only individuals who tested positive in both PCRs were included. We use aggregate weekly data, from 5th February 2007 until 4th December 2017. We assumed that an ARI episode, for which Influenza or RSV were detected from the nasopharynx on admission, was caused by the respective pathogen. The dataset excludes neonatal cases under 28 days old.

To inform transmission pathways in the population we used age specific contact-patterns, based on a contact study in the same area, conducted in 2010[23]. In total 2002 Nha Trang residents of all ages participated in the study. A contact was defined as either skin-to-skin contact or a two-way conversation.

Data analysis

We calculated the correlation between all reported influenza and RSV cases each week using a Pearson’s Correlation test.

Assuming no interaction (in susceptibility to or severity of dual infections), we calculated the required annual RSV infection attack rate in order to observe the reported number of dual infections (Eqs 1–3). We estimate the RSV attack rate rather than the influenza attack rate, as RSV is more consistent year on year (see section 1 in S1 Text for influenza equivalent). Using a negative binomial likelihood with Brent optimization we estimated the RSV reporting rate that would correspond to the maximum likelihood of observing the reported weekly number of dual infections. We then used this estimate of the reporting rate to calculate the annual RSV population attack rate required in order to observe this many coinfected ARI admissions. The credible intervals for the attack rate were calculated using the Hessian matrix from the optimisation. If the estimated attack rate is high, this may suggest that influenza and RSV co-infection increase severity (and hence reporting). If the estimated attack rate is low, this could suggest that co-infection is less likely that at random due to competition between the viruses.

With parameters: Incidence of reported cases (I), Prevalence of Infection (P), Duration of Infection (1/γ, 9 days—see Table 1), attack rate (AR) and estimated reporting rate (υ).

Table 1

Parameter definitions, values and priors.

Parameter	Symbol	Value	Prior	Source
Transmission rate INF	β INF	Fitted	Based on R0,INF	See section 4 in S1 Text for calculations
Basic Reproduction Rate Influenza	R 0,INF	Fitted	Between 1 and 8	[26]
Transmission rate RSV	β RSV	Fitted	Based on R0,RSV	See section 4 in S1 Text for calculations
Basic Reproduction Rate RSV for strain y	R 0,RSV,y	Fitted	Between 1 and 8	[26]
RSV age group susceptibility (0–1, 2–4, 5–15, 16–64,65+	τ i	Fixed	1, 0.75, 0.65, 0.65, 0.65	Based on Henderson et al (1979) [27], see section 5 in S1 Text
Infectious period Influenza	1/γINF	3.8 days	-	Cauchemez et al (2004) [28]Range from published papers: 1–4.5 days [28–31]
Infectious period RSV	1/γRSV	9 days	-	Weber et al (2001) [25]Range from published papers 6.7–12 days [25,32,33]
Strength of cross-protection	σ	Fitted	0–1	Assuming competitive [34]
Duration of cross-protection	1/ρ	Fitted	0—Inf days
Proportion of each age group infected with Influenza, at the start of the season	δ INF,s	Fitted	0–1
Proportion of each age group infected with RSV, at the start of the season	δ RSV,s	Fitted	0–1
Proportion susceptible to influenza at the start of the season	η s	Fitted	0–1	Exponential function, see section 6 in S1 Text for details
Influenza proportion reported in ages 0–1	κ INF	Fitted	0–0.4
Influenza multiplier for proportion reported ages 2–4 vs 0–1	κ INF,m	Fitted	0–5
RSV proportion reported in age group i	κ RSV,i	Fitted	0–0.4	No additional severity from dual infection [35]
RSV ON-1 reporting multiplier	κ RSV,2012	Fitted	1–5	ON-1 clinically more severe [24]
Dual infection multiplier on RSV proportion reporting	κ Dual	Fitted	1—Inf	Based on analysis of expected RSV Attack Rate above.
Overdispersion parameter	k	Fitted	0-Inf

Model

We created an age-structured deterministic transmission model for influenza and RSV, allowing for short-term cross-protection. Individuals are either Susceptible (S), Infectious (I), cross-Protected (P) and Recovered (R) for influenza (INF) and (RSV).

Susceptibles become infected at force of infections λ and λ, and move into the I states. They then remain infectious for 1/γ and 1/γ days and during the infectious period and 1/ρ days thereafter they are cross-protected and thus their propensity for heterologous infection is reduced by factor σ, the strength of cross-protection. There is no difference in inherent susceptibility by age for influenza, but there is reduced susceptibility to RSV in older age groups, determined by parameter τ.

The force of infection includes age-specific contact rates derived from a local contact survey [23]. Modelled age-groups are: 0–1 years (infants), 2–4 years (pre-school), 5–15 years (school), 16–64 years (adults) and 65 + (older adults). Infection reporting rates vary by age-group and virus, and for RSV reporting rates are increased by a multiplier from 2012 onwards, due to the circulation of a new genotype that has increased the average severity of infection and thus the proportion of reported infections (ON-1) [24]. There is also a multiplier on the RSV reporting rates for dual infections, allowing them to be reported more frequently, for example because of increased propensity for respiratory disease that would require healthcare seeking (as observed in adults [4]). Model equations are shown in section 3 in S1 Text.

We model the persisting immunity to influenza at the start of each season individually, with an initial proportion susceptible. This allows for a different immunity profile at the start of each season, without needing to account for waning of immunity, dynamics across virus subtypes and seasonality specifically. [25]. Due to infections in previous influenza seasons and potential vaccination, susceptibility to influenza is assumed to decline exponentially at rate η with age (see section 6 in S1 Text). This is modelled as non-leaky cross-protection, due to the combination of different exposures. For RSV we assume the only immunity at the beginning of the season is the age-specific reduction in susceptibility (leaky immunity), as immunity to reinfection typically lasts less than a year. Due to the short-time period modelled (max 66 weeks), we do not include births, deaths or ageing, but instead hold the age-group specific population sizes constant at the levels of 2010. Parameter definitions and values are shown in Table 1.

To capture the annual influenza and RSV epidemics despite regular changes, particularly in the timing of influenza circulation, we defined the annual start of the season as the minimum number of combined RSV and Influenza cases (on a 4 week rolling mean) between the 1st of November and the 1st of May each season. If one or more weeks had the same rolling average, we took the first occurrence within the time window.

Likelihood

We fitted the model to the age-stratified weekly number of ARI cases with nasopharyngeal carriage of either influenza or RSV using a negative binomial likelihood. To fit the allocation of those cases into influenza, RSV or dual infections we added a multinomial component to the likelihood, resulting in an overall log likelihood of:

Where x are the reported infections, θ are the parameters, j are the two age groups 0–1 and 2–4, and i are the weeks, with n being the total number of weeks. NB is the likelihood of the observed number of cases being a random draw from a negative binomial distribution with the total number of modelled infections as the mean, μ, and the fitted overdispersion parameter, k. MN is the multinomial likelihood, with pRSV, pFlu and pDual being the respective probabilities of the infection with influenza, RSV or both, calculated from the ratio of model reported cases.

Inference

We used parallel tempering to fit the model. This method involves running multiple markov chains simultaneously, with different ‘temperatures’ that place a weighting on the likelihood. Swaps between the chains are then proposed every x (in this case 5) iterations, and accepted with acceptance ratio:

Where

For full details of the method see Vousden et al (2016)[36] We ran the parallel tempering algorithm with 12 chains and 450,000 iterations. The initial 250,000 iterations were discarded as burn-in. Accepted samples from the first chain were then thinned to 1 in 10 for analysis, resulting in a final sample size of 20’000.

Single pathogen simulations

To assess the maximal indirect heterologous effect an intervention against either pathogen (e.g. widespread vaccination)could have in this environment, we assumed that complete absence of one of the viruses. We then calculated the relative change in the number of cases in the pathogen not targeted by the intervention. Estimates are based on model simulations using 1000 posterior parameter samples of the fitted model.

Sensitivity analyses

We assessed the sensitivity of our estimates of the strength of interaction of influenza and RSV to the prior on the interaction parameter and to the reporting rates of dual infections. We reran the model with a prior on the interaction parameter for strong interaction, using a normal distribution with mean 0.8 and standard deviation of 0.15. In addition, we ran a version of the model which did not allow an increased reporting rate for dual infections, as it has been reported that in this setting there is no increased severity of dual infections among hospitalised children [35]. Instead, it was assumed that the reporting rate for dual infections was the same as that for RSV-only infections.

Software

All analysis except for the fitting was done in R version 4.0.0. The fitting was done on R 3.4 on AWS ec2 machines. Code is available at https://github.com/NaomiWaterlow/NhaTrang_flu_rsv_interaction

Results

Descriptive analysis

A total of 788 influenza and 1687 RSV hospitalised paediatric ARI cases were reported between 5th February 2007 and 4th December 2017; 78 (9% of influenza cases and 4% of RSV cases) of these were dual infections (Fig 1A and 1B). The mean age of at admission was 22 months and 16 months for influenza and RSV cases respectively. RSV notifications showed strong consistent seasonality across years, peaking usually in the 34th week of the year, whereas influenza showed less seasonality, but typically occurs after Tết Nguyên Đán holidays and before the RSV epidemic (Fig 1C).

Fig 1

Data.

A) Weekly reported infections of children under 5 years old infected with influenza and RSV over the study period. B) Total number of cases reported over the entire study period by age group and virus. C) Percentage of reported cases by week of the year for RSV and Influenza. The thick lines show the combined total reported across all years, the semi-transparent lines show the 4-week moving average per year.

There was a small, not statistically significant, negative correlation between weekly influenza and RSV incidence, with the Pearson correlation coefficient -0.074 (95% CrI:-0.160 to 0.009). (Figure A in S1 Text)

We estimated that in order to observe the weekly reported number of dual infections when assuming independence of influenza and RSV infection, the annual RSV attack rate needed to be 720% (95% CrI: 560–1000) in ages 0-1y and 430% (95% CrI: 270–980) in ages 2-4y. The high attack rate suggests that in fact influenza and RSV infections are not independent but that co-infection is likely to substantially enhance the propensity for hospitalisation with ARI in this setting.

Model inference

The model was able to fit the case data for influenza and RSV well (Fig 2, see section 7 in S1 Text for further fitting and convergence details). The posterior estimates for the relative reduction in heterologous acquisition rates during and following Influenza or RSV infection was bimodal, with one mode at 0.004 (95%CrI 0.000–0.046), indicating hardly any competition for infection, and one mode at 0.41 (95%CrI 0.36–0.54), indicating moderate competition, assuming a cut-off between modes at 0.2 (Fig 3). The posterior for the duration of interaction also had multiple modes, with the mode corresponding to moderate competition at 10.0 days (95%CrI 7.1–12.8 days).

Fig 2

Model Fit: Black lines are the data, coloured lines are the 95% CrI posterior predictive interval.

Panels show the fit by age group and Virus.

Fig 3

Posterior estimates for parameters sigma and rho, and the corresponding likelihood values.

Colour is split by sigma value of 0.2. B) Goodness of Fit: Observed cases by season against Modelled cases by season by virus and age group. The black line indicates the same value.

The main differences between modes for other parameters were in the detection rate of influenza, which ranged from 13 to 21% of infections reported (section 8 in S1 Text) and the increased reporting for dual infections. We estimate that in the moderate competition mode the observation of influenza and RSV coinfection among ARI cases was 8.2 (95%CrI 6.9–9.9) times more likely than would have been expected by chance in ages 2–4 and 16.6 (95%CrI 14.5–18.4) in ages 0–1. This compares to the no competition scenario where the observation of influenza and RSV coinfection among ARI cases was 3.6 (95%CrI 2.5–5.8) and 7.2 (95%CrI 5.0–11.4) times more likely than would have been expected by chance in ages 2–4 and 0–1 respectively.

To assess the relevance of RSV and influenza interaction on population level in this setting we simulated single pathogen versions of the parameterised model. In the case of no competition, absence of influenza (e.g. through widespread vaccination) reduced RSV hospitalisations by 4.1% (95%CrI 3.3–7.1%) due to a lack of co-infections with higher propensity for severe disease and absence of RSV reduced influenza hospitalisations by 7.2% (95%CrI 4.4–7.2%) in the study period. In the moderate competition mode, absence of influenza reduced RSV hospitalisations by 5.7% (95%CrI 4.9–6.5%). In the absence of RSV 1.8% (95%CrI -0.7–7.2%) more hospitalised cases for influenza occurred (Fig 4).

Fig 4

Vaccination scenarios: Number of cases modelled over all seasons, with different vaccination assumptions.

Dots represent the median and lines the 95% CrI.

Sensitivity

As a sensitivity analysis, we assumed that co-infection will not modulate the propensity to be admitted to the local hospital with ARI symptoms (see section 9 in S1 Text), which pushed the posterior to the no competition mode. Further, we reran the model with a prior for strong competition through cross-protection, which pushed the posterior to the moderate interaction mode (see section 10 in S1 Text).

Discussion

We use data from more than 10 years of enhanced surveillance in Nha Trang, Vietnam to estimate the interaction of influenza and RSV epidemiology. We find that the observed data is consistent with infection reducing heterologous acquisition either by 41% (95%CrI 36% - 54%) for 10.0 days (95%CrI 7.1–12.8 days) after infection or hardly at all. We estimate that influenza-RSV co-infection increases the propensity of an infection to be reported through the ARI hospital surveillance by between 2.5 and 18.4 times. We go on to show that control of one virus in this setting may have little impact on the circulation of the other but can have an added benefit in reducing hospitalisations with co-infections.

A key strength of this dataset is the inclusion of cases infected with both influenza and RSV. Surprisingly though, some dual infections are reported at times when the two viruses seldom found in the hospital setting individually which may be a result of stochastic effect owing to the low number of observed dual infections and has limited the strength of inference from them. While many papers reporting co-infections do not include timings of the co-infections [37-39], these off-peak co-infections were not observed in Texas [40]. In addition, in Nha Trang influenza circulates continuously at low prevalence with small epidemics, which would result in constant low-level cross-protection, rather than a short-term more concentrated interaction after a large epidemic, such as in the UK [41]. This low-level cross-protection could have been absorbed into the transmission rate of RSV, explaining the low estimate of the RSV R (1.24 compared to a range of 1.2–9.1 in published papers [26]). These location-specific features will need to be considered when generalising the findings. However, in general, the setting of this study is a strength of the paper, as many of the countries where respiratory viruses circulate at low levels year-round are the same countries with limited influenza vaccine uptake. This study also has a high participation rate, however we do not have information on participants who declined to take part in the study, and this may bias the results.

Much of the evidence for cross-protection is on an individual, biological level [5,6,20]. However the impact of this individual level cross-protection on population level has been unclear, due to relatively small infection prevalence at any point in time (we estimate the seasonal attack rate for influenza as 0.011–0.15, and that peak prevalence never exceeded 0.8%), and thus a low propensity for co-infection with RSV. This may be exaggerated by clustering factors such as household transmission, reducing the opportunity for cross-infection. As an example in Kilifi, Kenya, household transmission of RSV contributed about 50% of all RSV transmission in households with young children [42]. Our model assumes a well-mixed population, so does not account for any population clustering beyond the age-specific contact matrices. In addition, we assume that risk of infection is age-dependent, but otherwise homogeneous. However, increased risk of influenza infection may be correlated with increased risk of RSV infection, due to demographic factors such as poor hygiene and household clustering. This may overestimate the effect of dual infections on reporting.

In our model we assumed the cross-protection between the two viruses to be bi-directional, having the same impact irrelevant of which virus caused the first infection. This assumption is based on the mechanism of cross-protection being the activation of the innate immune system into a general antiviral state. However, in the study site the influenza epidemic occurs before the RSV epidemic, therefore our estimate of the strength of cross-protection on susceptibility to the second virus is mainly an estimate of the impact of influenza on RSV and does not necessarily capture any dynamics in the other direction. In addition, our model is not able to capture delays in the timing of cross-protection, which could potentially occur when considering other mechanisms of cross-protection.

Evidence of cross-protection between influenza and RSV also comes from shifts in epidemic peaks, particularly after the 2009 influenza pandemic [10-13]. However these studies are observational, and cannot test mechanisms. As the SARS-CoV-2 pandemic has demonstrated, behavioural responses can have huge impacts on viral circulation, with many geographies seeing shifts in epidemic peaks for usually consistent viruses, such as RSV [43,44], due to limitations on social contacts. Fear generated from high infection rates can also drastically alter individuals behaviour [45], even without wide-spread implementations of restrictions.

Our model does not take into account different subtypes of influenza or RSV, due to the added complexity, additional parameters required and the lack of subtype specific data. We therefore assume that any cross-protection between influenza and RSV does not vary by subtype. We account for different immunity levels to circulating influenza subtypes by fitting a susceptibility parameter at the start of each season. This is necessary because we fit to each season, rather than including immunity waning and fitting over the time period combined. In addition, the dual infections appear to cluster in certain years, and a different explanation of this could be interaction between different viral subtypes each year. While the start weeks of our season are fixed manually, we account for any impacts of this by fitting the proportion infected at the start of each season for each virus. While most of the posterior estimates are reasonable, the reporting rate for influenza infections is high, between 13 and 21%, compared to estimated 12% from data in Yoshida et al. (2013). However, many milder cases (including outpatients) are included in the reports as they may seek healthcare at the hospital, thereby increasing the expected reporting rate in this context. The posterior for the detection rate of influenza is one of the few parameters that, like the strength of cross-protection parameter, is bimodal. However these were not greatly different, with the medians of the two priors only differing by 3.6% and therefore not substantially different to help us distinguish between the two modes. Overall therefore, our model estimates fit the data well, as well as known aspects of influenza and RSV transmission, such as high influenza attack rates in children [46,47], and higher RSV severity in the youngest children [48].

Conclusions

In summary, we use a novel modelling framework to interrogate a unique case time-series for single and dual infection from Nha Trang, Vietnam. We find that influenza and RSV co-infection substantially increases hospitalisation rates in children. In addition we show that the data supports either no or moderate individual-level cross protection against infection but either way with relatively little population level impact. This alleviates some concerns of heterologous effects of RSV or influenza vaccination, however, particularly in settings with more pronounced and overlapping RSV and influenza seasons the impact of vaccination on the other pathogen’s epidemiology may be more noticeable.

Further technical details and sensitivity analyses.

(DOCX)

Click here for additional data file.

18 Jan 2022

Dear Mrs Waterlow,

Thank you very much for submitting your manuscript "Evidence for Influenza and RSV interaction from 10 years of enhanced surveillance in Nha Trang, Vietnam, a modelling study." for consideration at PLOS Computational Biology.

As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. In light of the reviews (below this email), we would like to invite the resubmission of a significantly-revised version that takes into account the reviewers' comments.

The reviewers raise important questions and concerns about the model structure and complexity and the description of the model, which need to be addressed.

We cannot make any decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript is also likely to be sent to reviewers for further evaluation.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Please prepare and submit your revised manuscript within 60 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email. Please note that revised manuscripts received after the 60-day due date may require evaluation and peer review similar to newly submitted manuscripts.

Thank you again for your submission. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Daniel M Weinberger

Guest Editor

PLOS Computational Biology

Virginia Pitzer

Deputy Editor-in-Chief

PLOS Computational Biology

***********************

The reviewers are generally enthusiastic about the analysis but raise important questions and concerns about the model structure and complexity, which need to be addressed.

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: Overall, this is a well-done and clearly-written analysis of interactions between Influenza and RSV using detailed surveillance data from Vietnam. The authors developed an age-stratified transmission model for RSV and influenza, which was fitted to children aged <5 years with either one or two respiratory viruses infections to determine the strength of cross-protection and the effects of co-infections on disease severity. The authors concluded that co-infections increase the probability of being hospitalized and had a bimodal conclusion on cross-protection. The methods employed are well-justified and clearly articulated, and the conclusions appear to be broadly consistent with the results presented. However, I do have a few major comments regarding the methodology.

[1] The authors fitted 15 parameters for each season individually (line 159 and table 1). The maximum time period for each season in the model is 66 weeks and it seems that RSV infections are mostly concentrated between November and May (line 172-line 173 and Figure 1C), which are about 26 weeks. I am concerned that the sparsity of the data may cause issues of parameter identifiability. I find it hard to be convinced that the whole model converged well giving that several parameters showed bimodal posterior distributions. I would recommend the authors use the entire 10 year period with yearly varying parameters of influenza and use weakly informative priors based on published literature, especially for those 0-inf priors.

[2] In the Supplement, Model equations section, the outflow from line 68 -(1-sigma)*lambda_INFi*PS_i does not equal to the inflow in line 72 lambda*PS_i. Also, if the epsilon means introduction rate from external sources, it should not show up as outflows from the susceptible population. It would also help readers to understand the equations if the authors can specify the meaning of each compartment like what is done in main text 138-139 and the captions of Figure 2C.

Minor comments:

Introduction

[3] It may be better to clarify what is the “interaction” that this paper refers to. In lines 88 and 89, “to estimate the strength of cross-protection and the effects of co-infections on disease severity” may be more precise as it corresponds to the conclusions.

Methods

[4] Line 98: Do you have information on the 6% that were not enrolled in the study? Are their demographic background (remote living? who may have less chance of interactions with those living in the city) and contact patterns (such as household size) the same as those enrolled in the study?

[5] Line 100: Can you provide the sensitivity and specificity of the multiplex-PCR test?

[6] Line 198: Please provide information on how many interactions were left for the parameter estimates.

Reviewer #2: PCOMPBIOL-D-21-01994

The authors aim to better quantify the interaction between influenza virus and RSV, using data on pediatric infection from Nha Trang in Vietnam, and using a transmission modeling approach.

I appreciate the challenging nature of this type of study. But in this current form, I find the study lacking in several ways, and I have tried to describe these below. Perhaps some of these shortcomings can be addressed, in which case, it could be a worthwhile contribution to this field.

1. Important parts of the transmission model (and data) are not presented or adequately described.

It is unclear what assumptions the authors make about the dynamics of and transmission from individuals >5 years? It is stated in the methods that older age groups are modeled (which is good), but the data used to calibrate these, and the results (ie, incidence of flu and RSV from model simulations) are lacking.

Given that >5 population will be substantially larger, and may play an overriding role in driving the dynamics of pediatric infections, it should be described with more clarity and supported with data.

Similarly, it is unclear if and how seasonality is modeled. Given seasonal nature of transmission, especially of RSV, I think it would be important to carefully include this.

2. The model for interaction between influenza and RSV, is not conceptualized or described in detail.

In the background, authors describe that 1. increased severity of co-infections and 2. cross protection between two viral infections as the two pathways of interactions that the authors are interested in estimating.

However, it is unclear how is severity of co-infections are modeled; and what data are used to measure severe forms of co-infections.

For example, cross protection is modeled in a specific way, and it is unclear to me that the assumptions are justified. For example, it is assumed to be symmetric -- that protection from an infection of influenza to RSV would be the same in the opposite direction?

The evidence that authors cite in the background are mostly protection from prior influenza infection on RSV. At the very least, the modeling framework should be flexible to allow for protective interaction in one direction.

The fact that they find bimodality in their results may also be pointing to this.

Relatedly, while the duration of the cross protection is explored, the onset of the cross-protection is assumed to begin instantaneously. It would be importantly explore the timing, especially given that some of the studies they site report protection after several weeks.

3. I find it hard to make sense of the findings -- 41% cross-protection for 10 days or nothing. I can make sense of a finding where the data and the model are not able to find statistically significant evidence of interaction -- that no interaction and some interactions are equally likely.

But from what the authors are presenting, it appears to argue for two very distinct, yet very clearly defined interactions, which seems to be contradictory. I think this aspect needs to be explored in greater detail.

A few options may be:

(i) exploring more flexible forms of interaction (eg, allowing for asymmetry or delay in onset, as I have argued in 2)

(ii) showing simulations of two explanations -- whether one model captures certain features of the data better.

(iii) showing profile likelihoods for rho, to clearly show how evidence for the estimates compare.

Minor points/comments:

Abstract:

Line 24: Unclear what "heterologous ecological changes" means. Suggest simplifying the language.

Lines 31-33: The statement seems confusing or contradictory. Is this meant to suggest that you data cannot distinguish between

Lines 33-35: I do not understand what the two respective scenarios with 7.2 and 16.6 fold increases are -- perhaps something is missing?

Background:

One thing to point out in describing the two potential types of interactions, increased severity of co-infections and protective effect of influenza on RSV, is also the timing of those.

Clarifying in these studies the evidence for concurrent infections vs consecutive ones, would be helpful to the readers.

Perhaps useful to stipulate the specific types of cross-protection you are describing, particularly those in which two viral agents are not related antigenically.

Methods

It would useful to know the population denominators for Nha Trang.

It is unclear how seasonality is modeled, if it is at all?

Figure 1: Age groups do not include 4-5 years?

Figure 2: Part of this figure is illegible, especially panel C, model diagram, both due to smal text size and the quality of the figure.

Reviewer #3: Overall I think the authors present a strong, well-written, and interesting paper that will help improve our understanding of the interactions between various respiratory pathogens, and the implications for public health. I should make clear that I am not an expert in mathematical modeling, though what the authors describe having done seems reasonable to me. I do have a few minor comments/suggestions for the authors that I think will help improve the overall clarity and readability of the manuscript.

Comments:

Pg 7 (lines 62-63): This study was only in adults. The evidence in children has been far more mixed. Edit: I see you addressed this later on!

Pgs 10-11 (lines 144-145) and pg 11 (lines 162-163): Upon first reading these sentences they appear to contradict each other. After looking at it a bit more closely it becomes clear that they do not, but I would recommend revising these sentences to make this more apparent. Your language in the discussion (pg 22, lines: 343-345) was particularly helpful in understanding the reasoning behind your approach, but addressing this earlier in the methods of the paper would be helpful.

Pg 15 (lines: 209-212): The reasoning behind the second sensitivity analysis is a bit confusing as it is currently written. I would suggest the authors revise these sentences a bit to make this clearer.

Pg 19 figure 2A: It appears that the dual infections tended to cluster in certain years which might indicate the importance of subtype specific interactions. While I appreciate the substantial complexity that influenza and RSV subtypes would add to the model, if you have data on the influenza and/or RSV subtypes that tended to predominate in years where dual infections were more/less frequently observed this would be quite helpful to include in the descriptive analysis.

Pgs 20-21 (lines: 310-313): The setting of this analysis I think is a strength as many of the countries where respiratory viruses circulate at low levels year-round are the same countries with limited influenza vaccine uptake.

**********

Have the authors made all data and (if applicable) computational code underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data and code underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data and code should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data or code —e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No: Please refer to general comments

Reviewer #3: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #2: No

Reviewer #3: No

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, . PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at .

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example in PLOS Biology see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

9 Mar 2022

Submitted filename: NhaTrang Reviewer response.pdf

Click here for additional data file.

20 May 2022

Dear Mrs Waterlow,

We are pleased to inform you that your manuscript 'Evidence for Influenza and RSV interaction from 10 years of enhanced surveillance in Nha Trang, Vietnam, a modelling study.' has been provisionally accepted for publication in PLOS Computational Biology.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

You may also wish to address the small (optional) comment from the reviewer below.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us now if you or your institution is planning to press release the article. All press must be co-ordinated with PLOS.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Computational Biology.

Best regards,

Daniel M Weinberger

Guest Editor

PLOS Computational Biology

Virginia Pitzer

Deputy Editor-in-Chief

PLOS Computational Biology

***********************************************************

Thanks for your re-submission and apologies for the long review--it was difficult to wrangle the reviewers to have a look at it. I am recommending acceptance; Reviewer 1 does have 1 comment that could be addressed if you would like to address it with a minor revision. Comment from Reviewer 1: "One thing that can be improved is the discussion of simulation as shown in reviewer 2’s comment. The author can discuss how well their model structure is able to capture the parameter space based on their previously published simulation study entitled Competition between RSV and influenza: Limits of modelling inference from surveillance data."

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: The authors addressed all the comments pretty well. I do not have any further comments except for one, which is also optional.

One thing that can be improved is the discussion of simulation as shown in reviewer 2’s comment. The author can discuss how well their model structure is able to capture the parameter space based on their previously published simulation study entitled Competition between RSV and influenza: Limits of modelling inference from surveillance data.

**********

Have the authors made all data and (if applicable) computational code underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data and code underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data and code should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data or code —e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: Yes: Zhe Zheng

21 Jun 2022

PCOMPBIOL-D-21-01994R1

Evidence for Influenza and RSV interaction from 10 years of enhanced surveillance in Nha Trang, Vietnam, a modelling study.

Dear Dr Waterlow,

I am pleased to inform you that your manuscript has been formally accepted for publication in PLOS Computational Biology. Your manuscript is now with our production department and you will be notified of the publication date in due course.

The corresponding author will soon be receiving a typeset proof for review, to ensure errors have not been introduced during production. Please review the PDF proof of your manuscript carefully, as this is the last chance to correct any errors. Please note that major changes, or those which affect the scientific understanding of the work, will likely cause delays to the publication date of your manuscript.

Soon after your final files are uploaded, unless you have opted out, the early version of your manuscript will be published online. The date of the early version will be your article's publication date. The final article will be published to the same URL, and all versions of the paper will be accessible to readers.

Thank you again for supporting PLOS Computational Biology and open-access publishing. We are looking forward to publishing your work!

With kind regards,

Marianna Bach

PLOS Computational Biology | Carlyle House, Carlyle Road, Cambridge CB4 3DN | United Kingdom ploscompbiol@plos.org | Phone +44 (0) 1223-442824 | ploscompbiol.org | @PLOSCompBiol