Together with seasonal influenza, respiratory syncytial virus (RSV) is a

leading cause of respiratory infections and is responsible for a significant

socioeconomic burden in all age groups, especially at the extremes of age [McLaughlin,

2022; Rafferty, 2022]. Indeed, a recent modelling study [Osei-Yeboah, 2023] has

estimated that on average, a total of 158,229 RSV-associated hospitalizations among

European adults occur annually and 92% of these concern older adults aged ≥ 65

years. Contrary to young children, in the last 20 years mortality attributable

to RSV increased in both working-age and older adults [Du, 2023].

Several systematic reviews [Shi, 2020; Tin Tin Htar, 2020; Nguyen-Van-Tam,

2022; Shi, 2022; Maggi, 2022; Li, 2023] have investigated global epidemiology

and burden of RSV in (older) adults. These reviews have confirmed significant

RSV attack rates, hospitalization, mortality and case-fatality rates [Shi,

2020; Tin Tin Htar, 2020; Nguyen-Van-Tam, 2022; Shi, 2022; Maggi, 2022; Li,

2023], which seem similar to seasonal influenza [Maggi, 2022], and noted a

substantial case under-ascertainment [Li, 2023]. Moreover, there was a large

between-country variation in burden estimates [Tin Tin Htar, 2020], which is

driven by numerous factors, from climatic conditions [Haynes, 2013] to case

definitions and efficiency of the surveillance system in place [Staadegaard,

2021]. Notably, the available systematic reviews [Shi, 2020; Tin Tin Htar,

2020; Nguyen-Van-Tam, 2022; Shi, 2022; Maggi, 2022; Li, 2023] have identified

only up to six primary studies conducted in Italy.

Two vaccines have been recently authorized to prevent lower respiratory

tract disease (LRTD) caused by RSV in adults aged ≥ 60 years [Kotton, 2023].

This age indication will be likely extended to younger adults in the upcoming

years. In the United States (US), a single dose of RSV vaccine is recommended

to adults ≥ 60 years, as a part of shared clinical decision-making between

patient and healthcare provider [CDC, 2023]. In the United Kingdom (UK), RSV

vaccination is recommended for older adults aged ≥ 75 years, being the most

cost-effective option [United Kingdom Department of Health & Social Care,

2023]. By contrast, as of November 2023, no recommendations have been issued in

Italy.

Understanding country-specific burden of disease (BoD) is a key driver

for policy decisions on the introduction of new vaccines [Levine, 1997]. A

systematic appraisal of the burden of RSV enables policy makers, health

professionals and other relevant stakeholders to make informed decisions

regarding the recently available vaccines. In this regard, systematic reviews

on the frequency of different BoD indicators are important in the description

of the spatiotemporal distribution and the variation between population

subgroups potentially targeted by the novel preventive measures [Munn 2018;

Barker, 2021]. Country-specific BoD indicators are also essential for all types

of pharmacoeconomic models.

In Italy, a recent systematic review [Boccalini, 2023] has assessed RSV

BoD in pediatric outpatients. By including six studies, the authors found that

18–41% of children with respiratory infections were positive for RSV.

Conversely, no review systematically appraised the burden of RSV in Italian

adults. Indeed, RSV epidemiology is highly age-dependent, which hinders

transferability of pediatric estimates to other target population groups.

Furthermore, as we mentioned earlier, the available global reviews on (older)

adults [Shi, 2020; Tin Tin Htar, 2020; Nguyen-Van-Tam, 2022; Shi, 2022; Maggi,

2022; Li, 2023] were able to identify only a limited number of Italian studies.

In this systematic review, we aim to comprehensively collect and analyze

available data on the BoD of RSV in Italian adults in both primary care and

hospital settings with the ultimate goal of informing and supporting National

and local decision makers on the planification and implementation of

vaccination strategies.

Methods

Reporting standards s

PRISMA (preferred reporting items for systematic reviews and

meta-analyses) statement [Page, 2021] will be adopted as a reporting standard.

Methodological guidance for systematic reviews of observational epidemiological

studies reporting prevalence and cumulative incidence data developed by the

Joanna Briggs Institute (JBI) [Zachary, 2015] will be also consulted.

Eligibility criteria a

All types of observational studies (e.g., surveillance, cross-sectional,

cohort, case-series) and published in any modality (i.e., peer-reviewed

article, preprint, conference abstract, official report and other forms of grey

literature) will be eligible. The CoCoPop (condition, context, and population)

approach [Munn, 2015] will be used to formulate the inclusion criteria. In

particular, the condition of interest will be RSV infection detected by any

laboratory technique, including reverse-transcription polymerase chain reaction

(RT-PCR), cell culture, immunofluorescence assay (IFA) and rapid antigen

detection test (RADT). Moreover, RSV-specific International Statistical Classification

of Diseases and Related Health Problems (ICD) diagnosis codes will be also

considered a good proxy for the true RSV infection. Indeed, specificity of the

RSV-specific codes is 99.6–99.8% [Pisesky, 2016; Cai, 2020].

For the context, we will consider studies conducted in Italy, in any

setting (outpatient, inpatient or mixed) and calendar period. For this latter,

it is important to distinguish between year-around studies and those conducted

during a typical RSV season, in which the probability of RSV detection is much

higher [Rozenbaum, 2023]. We will define RSV epidemic season as a period

between October and April [Hamid, 2023].

Population will consist of adults defined as individuals aged  ≥ 14/18 years. When possible, we will

consider estimates for different age subgroups (e.g., working-age and older

adults) reported by single studies. Any clinical entity [e.g., influenza-like

illness (ILI), acute respiratory infection (ARI), severe ARI (SARI)] that

triggers swab collection will be eligible.

The following will be the exclusion criteria: (i) modeling,

pharmacoeconomic and similar studies with no original data; (ii) records with

insufficient data on RSV; (ii) studies on general population with no separate

data for adults; (iii) multi-country studies with no separate information for

Italy; (iv) redundant publications.

Study endpoints s

RSV attack rate (cumulative incidence) will be defined as the occurrence

of laboratory-confirmed RSV detections in a population (symptomatic/medically

attended, asymptomatic or both) in a specific period (e.g., RSV season). RSV

prevalence (overall, by subtype and genotype) will be defined as proportion of

RSV detections to the total number of subjects tested. Prevalence of

co-detections will be described as number of samples tested positive for both

RSV and any other respiratory pathogen to the total number of samples tested

positive for RSV. Case complication rate will be defined as proportion of

subjects tested positive for RSV and who developed at least one complication,

such as pneumonia, exacerbation of chronic obstructive pulmonary disease,

asthma, or congestive heart failure. As for drug use indicators, we will consider

frequency of drug prescriptions among subjects who was diagnosed with RSV. For

what concerns inpatient outcomes, crude hospitalization (per 100,000), case

hospitalization (i.e., proportion of RSV-positive individuals who were hospitalized)

rates, length of stay, frequency of admission to intensive care units will be

of interest. Analogously, crude, in-hospital, 30-day mortality and

case-fatality rates will be eligible. For these indicators, we will consider

only RSV-specific codes for RSV pneumonia (ICD-9-CM: 480.1; ICD-10-CM: J12.1),

acute bronchiolitis due to RSV (ICD-9-CM: 466.11; ICD-10-CM: J21.0), acute

bronchitis due to RSV (ICD-10-CM: J20.5), and RSV as the cause of diseases

classified elsewhere (ICD-9-CM: 079.6; ICD-10-CM: B97.4).

When possible, all study endpoints will be described overall, by age

group (working-age and older adults) and RSV season. Additional not

prespecified endpoints will be also eligible for exploratory purposes.

Search strategy y

The automatic search will be performed in the

following: PubMed/Medline, Biological Abstracts, Global Health through Ovid and

Scopus. In order to increase sensitivity, the number of selection criteria will

be minimized to the condition and population of interest and no filters or

other restrictions (e.g., language or publication year) will be applied. The

search following script, which will be adapted to each citation database, will

be used: (“Respiratory Syncytial Viruses”[MeSH Terms] OR "Respiratory

Syncytial Virus, Human"[MeSH Terms] OR “Respiratory Syncytial Virus

Infections” [MeSH Terms] OR “respiratory syncytial” OR “RSV”) AND (“Adult”

[MeSH Terms] OR “Aging”[MeSH Terms] OR “Men”[MeSH Terms] OR “Women”[MeSH Terms]

OR “Retirement”[MeSH Terms] OR “Long-term care”[MeSH Terms] OR “Nursing

care”[MeSH Terms] OR “Palliative care”[MeSH Terms] OR pension* OR retire* OR

adult* OR aged OR elderly OR senior* OR geriatric* OR nursing home*) AND

(“Italy”[MeSH Terms] OR Italy OR Italian*).

We will then perform a manual search through several

two modalities. First, the reference lists of the available systematic reviews

will be checked. Second, a backward cross-reference checking of the included

studies will be carried out. Third, a forward citation search by using Google

Scholar (http://scholar.google.com/) will be performed, as this search engine

is better suited for identifying grey literature sources [Haddaway, 2015].

Fourth, periodic reports of the Italian surveillance reports on influenza and

other respiratory viruses (http://respivirnet.iss.it) will be examined.

Finally, we will screen abstract books and proceedings of some relevant

conferences, including ECCMID (European Society of Clinical Microbiology and

Infectious Diseases), ESWI (European Scientific Working Group on Influenza) and

ReSViNET.

Study selection n

Results of the automatic search will be merged into a

single spreadsheet and duplicates will be removed in a semi-automatic modality.

The resulting list of unique records will undergo the process of screening by

assessing titles and/or abstracts and clearly irrelevant citations will be

discarded. Full texts of potentially pertinent publications records will be

then downloaded and assessed for the eligibility criteria described earlier.   Selection will be finalized by performing

the manual search, as described above. The entire process of study selection will

be performed by two reviewers, each working independently, and eventual

disagreements will be solved by consensus.

Data extraction and coding g

The following data will be extracted into a

spreadsheet: (i) full citation record; (ii) (region); (iii) period;

(iv) design; (v) setting; (vi) population; (vii) sample size; (viii) funding

source; (ix) definition of the clinical entity that triggered swab collection;

(x) methods used for RSV case ascertainment; (xi) numerators and denominators

used to compute the endpoints of interest described above. On the basis of

period, studies will be dichotomized on whether they overlapped with the

COVID-19 pandemic, which had a significant impact on circulation of RSV and

many other respiratory viruses. In particular, the northern hemisphere winter

season 2020/21 was characterized by the absence of very limited circulation of

RSV [Hamid, 2023; Boccalini, 2023; Gomez, 2021]. Estimates for that season will

be extracted and reported, but not included in the quantitative synthesis.

Starting from the 2021/22 season, RSV returned to the epidemiological scene

[Hamid, 2023] and therefore estimates from 2021/22 will be fully considered.

Multi-season studies that reported separate seasonal estimates will be

considered as distinct estimates [Bergeri, 2022]. On the basis of sample size,

studies will be median split. Regions will be categorized into three macro-areas

of North (Aosta Valley, Liguria, Lombardy, Piedmont, Emilia-Romagna,

Friuli-Venezia Giulia, Trentino-South Tyrol, and Veneto), Center (Lazio,

Marche, Tuscany, and Umbria), and South (Abruzzo, Apulia, Basilicata, Calabria,

Campania, Molise, Sicily, and Sardinia).

Missing data on relevant numerators and/or

denominators will be handled as follows. First, the corresponding author will

be contacted by email for clarification. In case of no reply, these data will

be imputed from the available percentages and/or by extracting data from

figures using WebPlotDigitizer v.4.6 (http://automeris.io/WebPlotDigitizer).

Data will be extracted by AD and then validated by GEC.

Critical appraisal l

The JBI checklist for prevalence/incidence studies

[Zachary, 2015] will be used to assess quality of the included studies.

Critical appraisal will be performed independently by two reviewers and

eventual conflicts will be solved by consensus.

Data synthesis s

Tabulated data will be first appraised qualitatively and

by visualizing forest plots. For quantitative synthesis, a proportional

meta-analysis will be undertaken according to the available recommendations

[Zachary, 2015; Barker, 2021]. As heterogeneity is expected to be high, random-effects

models with double arcsine transformation to stabilize variances will be used.

Pooled estimates will be expressed as proportions with both 95% Clopper-Pearson

exact confidence intervals (CIs) and 95% prediction intervals (PIs).

Heterogeneity will be quantified by means I ² the I2statistic. Notably I ²high I2 values do not necessarily mean

that data are inconsistent since true heterogeneity is expected in prevalence

estimates due to spatiotemporal differences [Barker, 2021]. As recommended,

publication bias will be not formally assessed (i.e., by funnel plots or

statistical tests), since there is no consensus about what a positive result in

a meta-analysis of proportions is [Barker, 2021].

To investigate the sources of heterogeneity across

studies, both subgroup and meta-regression analyses will be performed. In

particular, the subgroup analysis will be performed by age group (working-age

and older adults), setting (outpatient, inpatient and mixed) and study period

in relation to the COVID-19 pandemic (before and after the 2020/21 season). The

meta-regression modeling will be then conducted to examine the influence of

study characteristics on the RSV-related endpoints. This latter will be

performed only when at least 10 estimates are available [The Cochrane

Collaboration, 2011].

Meta-analysis will be performed in R (R Foundation for

Statistical Computing; Vienna, Austria) package “Meta” v. 6.5-0.