A systematic literature review of the economic and healthcare resource burden of cytomegalovirus

Abstract

Objective

Cytomegalovirus (CMV) can infect individuals at any age, including infants, who may contract it from infected mothers (congenital CMV [cCMV]). Whereas CMV infection is typically asymptomatic or causes mild illness in healthy individuals, infection can result in severe outcomes in immunocompromised individuals and in infants with cCMV. This systematic review aims to characterize the economic impact of CMV and cCMV infections.

Methods

Medline, Embase, and LILACS databases were searched for publications reporting the economic impact of cCMV and CMV infections across all age groups. Manuscripts published between 2010 and 2020 from Australia, Latin America, Canada, Europe, Israel, Japan, the United States, and global (international, worldwide) studies were included; congress materials were excluded. Outcomes of interest included cCMV- and CMV-attributable direct costs/charges, resource utilization, and indirect/societal costs.

Results

Of 751 records identified, 518 were excluded based on duplication, population, outcome, study design, or country. Overall, 55 articles were eligible for full-text review; 25 were further excluded due to population, outcome, study design, or congress abstract. Two publications were additionally identified, resulting in economic impact data compiled from 32 publications. Of these, 24 publications reported cost studies of cCMV or CMV, including evaluation of direct costs/charges, healthcare resource utilization, and indirect/societal costs, and 7 publications reported economic evaluations of interventions. The populations, methods and outcomes used across these studies varied widely.

Conclusions

CMV and cCMV infections impose a considerable economic impact on different countries, populations, and outcomes. There are substantial evidence gaps where further research is warranted.

PLAIN LANGUAGE SUMMARY

A plain language summary of this manuscript can be found here.

Keywords:

• Cytomegalovirus
• CMV
• congenital cytomegalovirus
• economic burden
• cCMV
• healthcare resource utilization

This article is related to:

A systematic literature review on the humanistic burden of cytomegalovirus

Introduction

Human cytomegalovirus (CMV), a member of the Herpesviridae family of viruses, is a common viral pathogen that can infect individuals of all ages^1,2. An estimated 83% of the world’s population has been infected with CMV³. In the United States, approximately 1 in 3 children will be infected with CMV by the age of 5 years, and more than half of adults will be infected by age 40 years².

CMV infection in healthy individuals is typically asymptomatic or results in mild illness; however, infection in immunocompromised individuals is more likely to lead to serious disease^2,4,5. Congenital CMV (cCMV) infection can result from vertical transmission during pregnancy in women who have had primary or reactivation of latent CMV infection, or have been reinfected while pregnant². cCMV infection occurs in approximately 0.5% to 1% of live births in the United States and Europe and is a leading viral cause of congenital birth defects worldwide, including neurodevelopmental delay and sensorineural hearing loss^5–8. Approximately 10% of infants with cCMV infection will exhibit symptoms at birth, such as rash, jaundice, retinitis, low birth weight, and microcephaly; 40% to 60% of these infants will have long-term complications such as hearing and vision loss, intellectual disability, or seizures^2,9. Among infants born with cCMV infection who are asymptomatic at birth, 10% to 15% will develop permanent sequelae such as sensorineural hearing loss (SNHL) by the time they begin school^7,9,10. Antiviral medications, such as oral valganciclovir, are approved for the prevention and/or treatment of CMV infection certain populations including newborns with moderate to severe symptomatic cCMV infection¹¹, solid organ transplant recipients¹², hematopoietic cell transplant recipients¹³, and those living with HIV¹⁴. Although no approved vaccines are currently available to prevent infection, a CMV vaccine has been designated as a high priority by the Institute of Medicine, given the potential significant public health impact¹⁵.

When assessing the value of preventive strategies, it is important to understand the economic impact and financial burden of CMV and cCMV infections on individuals and healthcare systems; however, it is unclear what data are available on the global economic impact of CMV and cCMV infections in different countries, in different populations, and for different economic outcomes. A comprehensive understanding of the economic impact of CMV and cCMV infections can provide guidance to decision makers developing and implementing strategies for prevention, including the use of vaccinations. Herein, we conducted a systematic literature review to describe data from 2010 to 2020 on the global economic impact of CMV and cCMV infections.

Methods

Literature search and information sources

The PICO (+) framework (Population, Intervention, Comparator, and Outcome with (+) indicating additional categories added to describe the research) was used to determine the scope of this review. A literature search on the economic impact of CMV infection was conducted on 24 December 2020, using the Medline, Embase, and the Latin American and Caribbean Health Sciences Literature (LILACS) databases. Searches were limited to the time frame of 2010–2020 for journal articles and 2017–2020 for congress abstracts, and restricted to English language publications from the following countries and regions: Australia, Latin America, Canada, Europe, Israel, Japan, the United States, and global (international, worldwide). A full list of free-text and controlled vocabulary terms utilized to conduct the searches is provided in Supplementary Tables S1–S3.

Study selection and data extraction

Titles of articles and abstracts identified in the literature search were assessed for relevance against predefined eligibility criteria in an initial screening, and then full texts were reviewed to further assess eligibility. Double independent record selection was performed during screenings of titles of articles and abstracts, as well as during full-text reviews. Discrepancies regarding inclusion or exclusion of records were resolved through discussions between reviewers or reconciled by a third reviewer. Following screening, data from the selected publications were abstracted into an evidence matrix.

Data included in the evidence matrix included study identification variables (i.e. reference identifier, author and year, and full reference), study characteristics (i.e. study design, study period, country, data type, and other information), and population characteristics (i.e. general information, sample sizes, subgroup analyses, age distribution, age range, sex, race/ethnicity, social status, educational level, and other information). Extracted economic outcomes of interest were direct costs/charges (i.e. direct costs/charges related to CMV infection and other direct costs); indirect costs/charges (i.e. productivity losses for patients and caregivers, and other); and resource utilization, including drug utilization, hospitalizations (i.e. hospital admission, readmission, discharge, length of stay, frequency of hospitalization, and other), emergency department visits, outpatient/ambulatory visits (i.e. outpatient/ambulatory visits, language/speech therapy, frequency, and other types of outpatient visits), and other resource utilization (i.e. laboratory tests, examinations, equipment related to CMV infection, palliative care or terminal care, and other).

Eligibility criteria

Each record identified during the initial screening and full-text review was assessed for relevance against predefined eligibility criteria for study types, populations, and outcomes. Populations included in this review were the general population at all ages; toddlers and children up to 5 years of age; adolescents and adults 9 through 40 years of age; mothers and babies infected with HIV; and other specific subpopulations or immunocompromised groups. These subpopulations included patients with autoimmune diseases (i.e. rheumatoid arthritis, systemic lupus erythematosus, Crohn’s disease, multiple sclerosis, Graves’ disease, Still’s disease, and polymyalgia rheumatica), cCMV-related inflammatory bowel disease (IBD), cancer, childhood leukemia, glioblastoma multiforme, cCMV-related subacute sclerosing panencephalitis and other neurologic disorders, hearing loss, ventilator-associated pneumonia (for babies with CMV in intensive care units [ICUs]), cardiovascular disease (i.e. atherosclerosis, coronary artery disease, hypertension), mononucleosis-like syndrome, and cCMV-related viral encephalitis. Sub-populations excluded from this study included blood and solid organ donors, transplant recipients, other high-risk and immunosuppressed groups (including people with HIV who were neither pregnant nor babies), and patients with the following diagnoses: diabetes mellitus, hepatitis or renal disorders, peripheral artery disease, systemic sclerosis, primary immunodeficiency such as autosomal recessive hyperimmunoglobulin E syndrome, Parkinson’s disease, Alzheimer’s disease, ulcerative colitis (UC; considered for inclusion only when presented as an irritable bowel syndrome subpopulation) or necrotizing enterocolitis, respiratory disease, dermatologic conditions, or Down syndrome.

Study types included in this review were epidemiologic studies, seroepidemiologic studies, cross-sectional studies, longitudinal studies, surveillance studies, registries, mathematical models, economic studies, claims analyses, systematic reviews, and meta-analyses; clinical trials were excluded. Inclusion criteria for the economic impact outcomes of CMV and cCMV infections were direct costs or charges related to disease, including treatment of related comorbidities, and lifetime cost; indirect/societal costs or charges, including loss of productivity and childcare outcomes; and resource utilization. Following review, congress materials such as abstracts and presentations were excluded to enhance the quality of the data.

Results

Study selection

A total of 751 citations were identified through the systematic literature search (Figure 1). After removing duplicates, 573 references were screened. Of these, 518 references were excluded because they did not meet the inclusion criteria for the study population (n = 269), economic outcome (n = 210), study design (n = 28), or country (n = 11). After the initial screening, 55 citations were eligible for full-text review. Of these, 25 were excluded because they did not satisfy the study population (n = 9), economic outcome (n = 9), or study design (n = 2), or because they were congress abstracts (n = 5). Two additional publications were identified and added following analysis of a systematic literature review published after the predefined search timeframe^16,17. A total of 32 publications were included in the data extraction stage and are included in this review.

Figure 1. PRISMA flow diagram of the screening process and study selection for the assessment of the economic impact of CMV.

Cost studies

The characteristics of studies (n = 24) reporting CMV- and cCMV-related direct healthcare costs/charges, resource utilization, and indirect/societal costs to patients, caregivers, and the general population are summarized below and in Table 1, and extracted results are shown in Supplementary Table S4.

Table 1. Summary of included studies (n = 24) reporting direct healthcare costs/charges, resource utilization and indirect costs related to CMV or cCMV.

Reference	Type of analysis/outcome	Disease/case type	Study design/methods	Study period	Study population and comparators	Top-line results
United States
Meyers et al.^18	Direct healthcare costs Resource utilization (hospitalization)	cCMV	Retrospective claims database analysis with matched control cohort	2011–2016	Birth analysis of infants with symptomatic cCMV (n = 397 [vaginal delivery (n = 167), cesarean delivery (n = 230]) and matched controls.	Mean difference in total healthcare costs in the first year of life for infants with versus without cCMV infection was $58,806 cCMV infection was associated with ∼70% and ∼32% increase in length of hospital stay among newborns with vaginal and cesarean delivery, respectively
Inagaki et al.^19	Healthcare charges Resource utilization (hospitalization)	cCMV	Retrospective analysis of serial cross-sectional data by birth year	2000–2012	Infants with symptomatic cCMV (n = 1349) by birth year: 2000 (n = 221), 2003 (n = 222), 2006 (n = 235), 2009 (n = 327), 2012 (n = 344)	Total charges per case for infants with cCMV infection were $45,771 for those born in 2000 and $86,514 for those born in 2012 Median duration of hospital stay for infants with cCMV infection was similar in 2000 (15 days) and in 2012 (18 days), and longer than for infants without cCMV infection (2 days)
Grossberg et al.^20	Healthcare charges Resource utilization (hospitalization)	CMV and UC	Retrospective database analysis	2006–2012	Adults aged 18–85 with both UC and CMV (n = 145) or with UC and no CMV (n = 32,290)	Mean length of hospital stay for patients with UC and CMV infection was longer (16.31 days) versus that for patients with UC and no CMV infection (5.52 days) Mean total hospital charges for patients with UC and CMV infection were higher ($111,835.50) versus that for patients with UC and no CMV infection ($39,895.50)
Lucas et al.^21	Direct healthcare costs Indirect costs	cCMV	Conceptual framework based on literature review and validation Analysis of direct, indirect, and intangible costs	2019	cCMV with vs without disability, developmental delay, or impairment in three life stages namely, infants, children and adolescents, and adults	Costs varied depending on the life stage, diagnosis, and the type and degree of developmental disability
Lanzieri et al.^22	Resource utilization (hospitalization)	cCMV or CMV	Retrospective database analysis of prevalence and clinical characteristics of infants	2005–2010	Infants in NICU with cCMV/CMV infection (VLBW [n = 87] and birthweight >1500 g [n = 69]) Infants in NICU without cCMV/CMV (VLBW (n = 33,424) or birthweight >1500 g (n = 57,614)	Median length of hospital stay was longer for surviving VLBW (<1500 g) infants with cCMV/CMV infection (99 days) versus those without (60 days) Median length of hospital stay was longer for surviving infants with birthweight >1500 g with cCMV/CMV infection (48 days) versus those without (14 days)
Tran et al.^23	Resource utilization (hospitalization)	cCMV or CMV	Retrospective database analysis	2005–2016	Infants in NICU with cCMV/CMV (VLBW [n = 174] or birthweight [>1500 g; n = 145]) Infants in NICU without cCMV/CMV (VLBW [n = 64,101] or birthweight >1500 g [n = 116,416])	Median length of hospital stay was longer for VLBW (<1500 g) infants with cCMV/CMV infection (98 days) versus those without (46 days) Median length of hospital stay was longer infants with birthweight >1500 g with cCMV/CMV infection (98 days) versus those without (14 days)
The Netherlands
Korndewal et al.^24	Direct healthcare costs Resource utilization (hospitalization)	cCMV	Retrospective cohort study	2012–2013	Children with cCMV (n = 133) and without cCMV (n = 274)	Total healthcare costs in the first six years of life amount to €6113 per child with cCMV and €3546 per child without cCMV Mean total costs per child related to long-term impairments, such as hearing loss were €17,205 for children with cCMV and €8332 for children without cCMV
United Kingdom
Retzler et al.^25	Direct healthcare costs Indirect (nonmedical) costs	cCMV	Static cost model	2016	Individuals of all ages with cCMV	Estimated mean total costs of cCMV infection: ₤732 million Approximately 40% of total cost from direct costs incurred by the government for health, social, and educational services with the remaining 60% attributed to indirect costs, including loss of productivity to society
Ahmed et al.^26	Resource utilization (hospitalization, colectomy rates)	cCMV	Retrospective chart review of from the authors’ institution from 2005 to 2011	2005–2011	Adult patients with IBD and a diagnosis of CMV infection (n = 35)	Six of 11 patients with IBD and CMV colitis experienced a recurrence of symptoms within 4 weeks of stopping a 1-week course of IV ganciclovir treatment and were readmitted to the hospital; 16 patients treated with a 2-week course of IV ganciclovir showed more favorable outcomes including a shortened length of hospital stay than patients treated with a 1-week course of IV ganciclovir or a 2-week course of oral valganciclovir
Germany
Walter et al.^17	Resource utilization Direct healthcare costs Indirect costs	cCMV	Incidence-based, cost of illness model	2008	Children with cCMV (n = 1431); grouped by the presence or absence of symptoms at birth and by whether or not they experienced remote damages	Total direct costs for children with symptomatic cCMV infection at birth were €68,379 and €99,613 per child with and without remote damages, respectively 92% of total costs related to cCMV in Germany are indirect costs
France
Delvincourt et al.^27	Resource utilization (length of hospital stay, colectomy rates; drug use)	CMV	Case-control and retrospective analyses evaluating consequences of latent CMV infection reactivation in flaring IBD patients	2003–2012	CMV-positive (+) UC patients (n = 26) and matched controls; IBD patients with CMV reactivation (n = 105 including 80 with positive blood PCR receiving antiviral treatment (n = 49) or not receiving antiviral treatment (n = 31)	There was no significant difference in the length of hospital stay between CMV + and CMV-negative patients with IBD (8.7 vs. 9.7 days) and 3-month colectomy rates (15.4% vs. 23.1 %). The mean length of hospital stay was longer for patients with CMV + who received antivirals than those who did not (16.3 days versus 8.3 days), suggesting higher severity in the treated population
Touhami et al.^28	Resource utilization (% of patients receiving surgery; drug utilization)	CMV	Retrospective chart review of the treatment of patients with CMV AU	2005–2015	Overall study population (n = 35); CMV AU patients (n = 35; with PSS [n = 24] or chronic non-specific AU [n = 11])	Glaucoma or phacoemulsification surgery was performed in 25.7% and 31.4% of all CMV patients, 20.8% and 25.0% of CMV patients with PSS, and 36.3% and 45.5%% of CMV patients with chronic AU, respectively
Slovenia
Saletinger et al.^29	Resource utilization (hospitalization and treatment with oxygen)	CMV	Prospective analysis reporting length of hospital stay and treatment with oxygen for patients with reactivated CMV versus patients without CMV reactivation	2011–2013	Adult patients with community-acquired pneumonia (n = 80): CMV DNA positive (n = 6); CMV DNA negative (n = 74)	Among immunocompetent adult patients with community-acquired pneumonia, not treated in ICU, there was no significant difference in median length of hospital stay between those who tested positive for CMV DNA versus those who tested negative for CMV DNA (12 days versus 10 days, respectively)
Spain
Vergara et al.^30	Resource utilization (hospitalization/ICU)	CMV	Prospective analysis of length of stays in the hospital and ICU (days) for adult patients with CMV infection	2013–2015	ICU patients with CMV as detected by bronchioalveolar lavage: CMV positive (n = 35) CMV negative (n = 98)	Length of hospital and ICU stays were longer for ICU patients with CMV infection (46 and 31.5 days, respectively) than those without (35 and 19.5, respectively)
Poland
Jedlińska-Pijanowska et al.^31	Resource utilization (hospitalization)	cCMV	Retrospective study	2012–2017	Infants with symptomatic cCMV treated with ganciclovir-based antiviral therapy (n = 60) or oral VGCV-based therapy (n = 38)	Mean length of hospitalization depended on the type of antiviral therapy administered. In 2012, the mean length of hospital stay was 38, 28, and 6 days for ganciclovir monotherapy, sequential ganciclovir plus VGCV, and oral VGCV monotherapy, respectively. In 2017, the mean length of hospital stay was 29 and 9 days for sequential and VGCV therapy, respectively
Italy
Maconi et al.^32	Resource utilization (length of hospital stay, colectomy rates)	CMV	Retrospective database analysis included data on antiviral treatments, lengths of hospital stay, and colectomy rates	1998–2012	Active IBD patients with CMV (n = 38); patients who received antiviral treatment (n = 13) and patients who did not receive antiviral treatment (n = 25)	In a cohort of 38 patients with IBD flare and CMV infection, the average length of stay was 14 days (range: 7–35 days); hospital stay duration was comparable between treated and untreated patients
Japan
Aoki et al.^33	Direct healthcare costs Indirect costs	cCMV	Literature-based model of economic and lifetime health burdens of cCMV infection in Japan based on annual birth cohort in 2019 using demographic, epidemiologic, health value, and economic indicators	2019	2019 Japanese birth cohort	The impact of cCMV infection for the Japanese 2019 birth cohort was approximately ¥27.6 billion, of which 4% comprised direct healthcare costs/charges
Maeda et al.^34	Total costs of hospitalization Resource utilization (hospitalization)	CMV	Retrospective database analysis reporting on the impact of opportunistic infections on in-hospital mortality, lengths of hospital stay, and total costs	2007–2010	Adults with T-cell leukemia and CMV as an opportunistic infection (n = 109)	Among adult patients with T-cell leukemia between 2007–2010, median total charges associated with CMV infections were $20,549 per instance
Lin et al.^35	Resource utilization (% of admissions to NICU)	cCMV	Longitudinal observational analysis using claims data, reporting incidence of adverse CMV sequelae in cCMV patients	2010–2017	Overall study population (n = 207,547); Infants with cCMV (n = 53); Infants with cCMV and related symptoms (n = 36); Individuals with cCMV without related symptoms (n = 17)	Overall, 47% of patients with cCMV infection were admitted to the neonatal ICU compared with 18% of infants without symptoms of cCMV infection at birth; however, these differences did not reach statistical significance
Brazil
Niz Xavier et al.^36	Resource utilization (% of newborns receiving mechanical ventilation)	cCMV	Cross-sectional analysis on the frequency of cCMV infection in newborns admitted to the NICU	2010–2012	Overall study population (n = 520); Newborns who tested positive for cCMV (n = 13)	23% (n = 3/13) of newborns who tested positive for cCMV received mechanical ventilation
Yamamoto et al.^37	Resource utilization (OAE and ABR examinations)	cCMV	Prospective analysis of SNHL in newborns in 2 public hospitals	2003–2009	Overall study population (n = 12,295); Infants with cCMV (n = 121)	84% of neonates with cCMV underwent at least one ABR assessment (median age of 12 months) Seven of 10 infants with hearing loss were tested by OAE by 1 month of age, of whom 6 failed OAE and subsequent ABR
Marin et al.^38	Direct healthcare costs (antiviral use)	cCMV	Prospective analysis of the prevalence and clinical presentation of cCMV infection among a low-income population	2010–2012	Overall study population (n = 2100); Infants with cCMV (n = 25; 24 asymptomatic and 1 with symptoms)	For newborns with asymptomatic cCMV, 0% were treated with antiviral drugs until the second year of life
Global
Lachance et al.^39	Resource utilization (hospitalization, mechanical ventilation, renal replacement therapy)	CMV	Systematic literature review of association between CMV reactivation and survival, and resource utilization in immunocompetent critically ill patients	1990–2016	Immunocompetent critically ill patients with CMV admitted to ICU (n = 973) and/or receiving mechanical ventilation (n = 683).	The mean increase in length of ICU stays and mechanical ventilation for patients with CMV reactivation versus without CMV reactivation were 8.18 days and 6.6 days, respectively
Li et al.^40	Resource utilization (hospitalization)	CMV	Systematic literature review analyzed CMV infection and requirement for mechanical ventilation and duration of ICU stay in immunocompetent patients	1990–2016	Immunocompetent patients admitted to ICUs with/without CMV infection (n = 2398)	The mean increases in length of stay in ICU and duration of mechanical ventilation for patients with versus without CMV infection were 12 days (n = 875) and 9 days (n = 949), respectively

Abbreviations. ABR, auditory brainstem response; AU, anterior uveitis; CMV, cytomegalovirus; cCMV, congenital CMV; IBD, inflammatory bowel disease; ICU, intensive care unit; IV, intravenous; NICU, neonatal intensive care unit; OAE, otoacoustic emission; PSS, Posner-Schlossman syndrome; UC, ulcerative colitis; VLBW, very low birth weight.

United States

Six studies of cCMV and CMV infections in infants, children, adolescents, and adults in the United States were included (Table 1)^18–23. Of these, four studies evaluated direct healthcare costs or healthcare charges associated with drug utilization; visits to inpatient, outpatient, and emergency facilities; equipment; laboratory tests; and/or physical and occupational therapy related to cCMV (three studies) or CMV (one study) infection^18–21. These reports consistently found additional/higher healthcare costs associated with cCMV infection^18,19,21 and in immunocompromised patients with UC and CMV infection²⁰. These results align with the reported increase in the length of hospital stay in five of the six US-based cost studies^{18–20,22,23}. Four of the studies evaluated length of hospital stay in infants with cCMV infection^18,19,22,23 and one study assessed hospitalization in immunocompromised adults with IBD and CMV infection²⁰. Overall, the studies suggest that CMV infections in infants and immunocompromised adults are associated with increased length of hospital stay and increased healthcare cost.

Europe

Ten publications reporting cost studies of cCMV or CMV infection among infants, children, adolescents, and adults from eight European countries were identified^17,24–32. Countries with available data include France (n = 2), the United Kingdom (n = 2), Germany (n = 1), Italy (n = 1), Poland (n = 1), Slovenia (n = 1), Spain (n = 1), and the Netherlands (n = 1).

Direct healthcare costs of cCMV or CMV infection were assessed in three of the studies^17,24,25. The studies from Europe were more diverse in observation periods and scope, reporting data for costs up to six years after birth, direct costs incurred by the government for health, social and educational services and costs for children with cCMV, stratified by long-term impairments and remote damages^17,24,25. Mean total costs incurred due to long-term impairments, such as hearing loss, were more than double in children with cCMV infection than in those without in the Netherlands²⁴. Direct healthcare costs/charges due to cCMV infection were estimated at 324 million British pound sterling (GBP) in the United Kingdom in 2016²⁵, and 70.5 million EUR per annum for all children affected by cCMV in Germany¹⁷. Notably, direct costs comprised a minor proportion of the total cost in both the United Kingdom (40%)²⁵ and Germany (8%)¹⁷.

Resource utilization: hospitalization and other

Seven studies reported length of stay in hospital associated with CMV or cCMV infection in Europe^{24,26,27,29–32}. Results were diverse for adult patients, with length of hospital stays in Slovenia and France not significantly different between CMV-positive and CMV-negative patients while in Spain, average length of stay in hospital was longer for patients with CMV infection^27,29,30. Although this variation in length of hospital stay is notable, it is important to consider that the populations reported in these studies were heterogenous with respect to the presence of comorbidities and the use of treatment regimens. Length of stay in hospital were similar for infants with and without cCMV²⁴; however, among infants with cCMV infection, length of stay varied between depending on the type of antiviral therapy administered (6 days with oral valganciclovir [VGCG] and 38 days for IV ganciclovir [GC] monotherapy)³¹. In adult patients with inflammatory bowel disease and CMV colitis, an intravenous regimen of GC was associated with a superior clinical response, including a shorter hospital stay compared with oral VGCV²⁶. Despite the heterogeneity of these populations, these reports suggest an overall trend of increased length of hospital stay in patients with CMV infection.

Indirect costs

In contrast to the absence of data on indirect costs in the United States, two publications reported these costs in Europe^17,25. In Germany, indirect costs contribute 92% of the total costs associated cCMV infection; caregivers to children with cCMV infection experience changes in employment status and careers while costs of education were as high as 44,835 EUR for children with cCMV-related disabilities¹⁷. By contrast, 60% of cCMV-associated costs in the United Kingdom were estimated to be indirect costs, amounting to 408 million GBP incurred by families and society as a result of productivity losses and costs of SNHL, cerebral palsy, epilepsy, and autism spectrum disorders²⁵. Taken together, these reports suggest that the indirect costs constitute a major proportion of the overall economic burden associated with CMV infections.

Japan

Three studies reported on CMV- and cCMV-related direct healthcare costs/charges, resource utilization, and/or indirect costs in Japan^33–35.

Direct and indirect healthcare costs/charges

Two publications providing data on direct healthcare costs/charges^33,34. One of these publications suggests that the economic impact of CMV infection in adult patients with T-cell leukemia in Japan is substantial³⁴. The second study estimated the overall costs of cCMV infection in the 2019 Japanese birth cohort at approximately 27.6 billion Japanese Yen (JPY), of which only 4% represented the direct healthcare costs and those associated with hospitalization³³.

Resource utilization: hospitalization and other

Two publications reported data on cCMV- or CMV-related hospitalizations in Japan^34,35. Although 47% of patients with cCMV-related symptoms at birth were admitted to the neonatal ICU (NICU), this proportion was not significantly higher than that for infants without cCMV symptoms at birth (18%)³⁵. This longitudinal study of medical claims also reported that the proportion of neonates, <6 months of age, receiving antiviral treatment in Japan (26%) is similar to that reported for the United States. Similar with publications from the United States and Europe reporting longer hospital stays among patients with comorbid CMV infection, stays in hospital among immunocompromised adults with T-cell leukemia were longer for those with CMV infection compared with other opportunistic infections³⁴.

Indirect costs

As observed in the case of the United Kingdom and Germany, indirect costs represented a major proportion of the total cost burden of CMV, approximately 96% of costs incurred due to cCMV infection in Japan in 2019 comprised indirect costs, such as education and productivity losses for both patients and parents³³.

Brazil

Three publications from Brazil reported direct healthcare costs/charges and resource utilization associated with CMV and cCMV infection^36–38. One of these studies reported the number of infants with cCMV infection who received mechanical ventilation (23%) following admission to a neonatal ICU³⁶. Another reported on the use of auditory brainstem response assessments during early childhood (84%) and the observed frequency of SNHL³⁷.

Global

Two systematic literature reviews provided data on resource utilization among adults with CMV infection in multiple countries, one of which reported a mean increase of eight days in length of ICU stay and an increased duration of mechanical ventilation among immunocompetent critically ill patients with CMV reactivation versus patients without CMV reactivation (this review did not report countries of included studies)³⁹. Similarly, a second review of resource utilization across various countries, including the United States, Europe, Egypt, and New Zealand, reported mean increases of 12 days for length of stay in ICU and 9 days for the duration of mechanical ventilation between patients with and without CMV infection⁴⁰.

Economic evaluations of interventions

The seven economic evaluation studies that reported on interventions for CMV and cCMV infection are presented in Table 2, and a summary of the extracted results is below and in Supplementary Table S5. As described previously⁴⁷, economic evaluations were of two types, cost-effectiveness and cost-benefit analyses that predict the effect of preventive or therapeutic interventions on health outcomes or costs.

Table 2. Summary of included economic evaluations of interventions related to CMV or cCMV (n = 7).

Reference	Type of analysis/outcomes	Disease/case type	Study design/methods	Study period	Study population and comparators	Top-line results
United States
Bergevin et al.^41	Cost-benefit	cCMV	Economic estimation of administrative and medical costs of cCMV screening and differential cost of treatment of cCMV positive infants	2014–2015	Infants who failed two newborn screening tests	The study suggests net benefits and societal savings from early identification and treatment of cCMV in infants under 3 weeks of age who failed newborn hearing screening (based on the cost of cochlear implants). The authors acknowledge that these estimates may be overly generous, as only a minority of eligible children would receive cochlear implants, this is likely offset by the exclusion of familial, societal, and educational benefits from this analysis
Albright et al.^16	Cost-effectiveness	cCMV and primary maternal CMV	Decision model comparing cost-effectiveness of universal versus risk-based screening using prenatal serum to detect cCMV	2018	Pregnant women	In estimating the costs and benefits of either universal prenatal serum screening or routine, risk-based screening of pregnant women for CMV infection (assumed CMV testing conducted prior to 20-week gestation or following observation of ultrasound abnormalities, respectively), the model predicted universal screening would be cost-effective in 62.8% of all modeled scenarios. The cost-effectiveness results for universal prenatal screening was significantly influenced by the incidence of primary CMV infection during pregnancy, with universal screening remaining cost-effective as long as the incidence of CMV was more than 0.89%
Gantt et al.^42	Cost-effectiveness	cCMV	Decision model comparing cost-effectiveness of universal and targeted newborn cCMV screening versus no cCMV screening	2014–2016	General newborn population	Estimated mean incremental costs of identifying 1 cCMV infection in newborns in the United States ranged from $2000 to $10,000 and $566 to $2832 for universal and targeted screening, respectively. Based on assumptions of antiviral treatment benefit in reducing hearing loss, including receipt of a cochlear implant, for symptomatic infants, modeling analyses estimated moderate direct costs of $10.86 (universal screening) and net savings of $0.90 (targeted screening) per newborn screened. Net savings of $21.34 (universal screening) and $10.66 (targeted screening) per newborn screened were estimated when costs related to loss-of-productivity were included
Dempsey et al.^43	Cost-effectiveness	cCMV and CMV	Decision model comparing costs, potential clinical impacts, and cost-effectiveness of the current strategy of no CMV vaccination versus a strategy where all adolescent females are vaccinated against CMV prior to their first pregnancy.	2010	Adolescent females	The incremental cost savings resulting from vaccination of adolescent females against CMV before their first pregnancies (three-dose schedule) was estimated at $32.3 million per 100,000 females and would substantially reduce the number of infants affected by long-term sequelae of CMV infection. Vaccination avoided 8 infant deaths, as well as reductions in the number of infants affected with hearing loss, vision loss, and mental retardation compared to no vaccination (14/100,000 versus 66/100,000, 1/100,000 versus 6/100,000, 5/100,000 versus 23/100,000, respectively. Universal vaccination of adolescent females to protect their future children against cCMV was cost-effective under a wide variety of conditions, assuming that vaccine efficacy against disease would need to be at least 61%
France
N’Diaye et al.^44	Cost-effectiveness	CMV	Decision model evaluating the cost-effectiveness of various strategies to vaccinate adolescent females, all women, or CMV seronegative women only versus no vaccination (current practice)	2014	Adolescent females	Systematic vaccination (three doses) of all eligible girls was found to be most cost-effective strategy for the prevention of cCMV infection in infants, followed by screening of all women and vaccinating only those who were seronegative for CMV infection
United Kingdom
Williams et al.^45	Cost-benefit	cCMV	Decision model cost analysis of targeted screening and treatment for cCMV-related hearing loss in otherwise asymptomatic infants	2010–2011	Infants with asymptomatic cCMV who were referred for further assessment after initial hearing screening tests	A cost analysis of targeted screening and subsequent treatment for cCMV-related SNHL in otherwise asymptomatic infants reported total costs of identifying, treating, and following up on identified cases were ₤6683 per case of SNHL and ₤113,615 per 100,000 infants undergoing newborn hearing screening
Australia
Beswick et al.^46	Cost-effectiveness	cCMV	Decision tree model evaluating cost-effectiveness of introducing targeted salivary cCMV testing into a newborn hearing screening program	2014–2016	Infants who did not pass their first and second newborn hearing screen (n = 283)	The differences in costs between implementing newborn hearing screening and not screening were negligible

Abbreviations. CMV, cytomegalovirus; cCMV, congenital CMV; SNHL, sensorineural hearing loss.

United States

Of the four economic evaluations from the United States, one cost-benefit analysis predicted that net benefits and societal savings would result from early screening and treatment of cCMV infection in infants who fail newborn hearing screening assessments, based on the assumption that antiviral therapy would reduce the need for cochlear implants⁴¹. The other three studies estimated the cost-effectiveness of maternal CMV screening, neonatal cCMV screening, and female vaccination before first pregnancies using decision models to compare costs versus outcomes associated with these interventions^16,42,43. The cost-effectiveness of universal prenatal serum screening outweighed that of maternal risk-based screening for CMV infection during pregnancy and was cost-effective in 63% of modeled scenarios according to a 2019 study¹⁶. However, the cost-effectiveness of universal prenatal screening was significantly influenced by the assumed incidence of primary CMV infection during pregnancy. Costs of identifying one cCMV infection were estimated to be substantially higher for universal screening than targeted screening of newborns⁴². This study also modeled direct costs related to antiviral treatment for symptomatic newborns and estimated moderate costs of 11 USD (universal screening) to net savings of 1 USD (targeted screening) per screened newborn⁴². Modeling a scenario in which a CMV vaccine was available, savings were estimated at approximately 32 million USD per 100,000 adolescent females vaccinated against CMV infection before their first pregnancies; in this scenario, the number of infants affected by long-term sequelae was also reduced as the vaccination strategy avoided eight infant deaths as well as reductions in the number of infants affected with hearing loss, vision loss, and mental retardation compared to no vaccination (14/100,000 versus 66/100,000, 1/100,000 versus 6/100,000, 5/100,000 versus 23/100,000, respectively)⁴³.

Europe

Two studies used decision modeling to evaluate interventions for CMV and cCMV infection in Europe^44,45. Various strategies were modeled for CMV vaccination in France; vaccination of all eligible adolescent females, as in the cost-effectiveness evaluation from United States, or of all CMV seronegative women were among the most beneficial strategies against cCMV infection in infants⁴⁴. Similar to the estimated benefits of broader newborn cCMV screening in the United States, costs incurred as a result of targeted screening of infants who did not pass their universal newborn hearing screening assessment in the United Kingdom were low (14.000 GBP per case) and could potentially result in significant cost benefit over the lifetime of the individual⁴⁵.

Australia

A single cost-effectiveness analysis of cCMV-related costs in Australia used decision tree modeling to evaluate the costs of targeted salivary cCMV testing into a newborn hearing screening program⁴⁶. In contrast to the potential savings arising from the introduction of targeted newborn cCMV screening in the United Kingdom, this economic evaluation projected negligible cost differences as a result of implementing targeted cCMV screening for infants who do not pass newborn hearing screening, with authors concluding that the inclusion of cCMV screening would be practically and financially achievable.

Other studies

One other publication identified in our systematic literature review was neither a cost study, nor an economic evaluation of interventions (Table 3)⁴⁸. This observational study reported an increase in annual discharge rates of infants with cCMV infection from 4.6 per 100,000 infant population during 1979–1985 to 22.3 per 100,000 in 2016 in the United Kingdom.

Table 3. Summary of other studies reporting the burden of cCMV (n = 1).

Reference	Type of analysis/outcomes	Disease/case type	Study design/methods	Study period	Study population and comparators	Top-line results
United Kingdom
Kadambari et al.^48	Hospital discharge records with viral diagnosis	cCMV	Population-based observational analysis of incidence of cCMV, HSV, VZV, and rubella with perinatal and maternal factors	1968–2016	Infants aged <1 month diagnosed with cCMV, HSV, VZV, and rubella from 1968 to 2016	The annual discharge rates of infants with cCMV infection increased from 4.6 patients per 100,000 during 1979–1985 to 22.3 patients per 100,000 in 2016

Abbreviations. cCMV, congenital cytomegalovirus; HSV, herpes simplex virus; VZV, varicella zoster virus.

Discussion

The goal of this systematic literature review was to summarize the economic impact of CMV and cCMV infections across a broad selection of outcomes, populations, and countries, as well as to identify any gaps in the current body of literature. The heterogeneity of the outcomes reported in the available data limited the ability to succinctly collate and draw conclusions on economic impact. Of the 24 cost studies identified, 15 studies reported outcomes for cCMV, and 10 on CMV. Limited analyses of direct healthcare costs were available for cCMV (n = 6) and CMV (n = 3). Data on hospitalization and resource utilization were available from 11 reports in cCMV, and in all 10 reports on acquired CMV infection. Increased costs and resource utilization related to medical care were reported for other specific conditions in people with CMV, such as IBD. Of the seven economic evaluation studies identified, the cost-effectiveness or cost-benefit of a wide range of preventative measures or therapies for cCMV or acquired CMV infection, including prenatal screening, newborn screening, and adolescent vaccination, were assessed. There was less heterogeneity with respect to economic perspective, with most publications evaluating costs and or charges from a healthcare perspective (n = 30 publications), societal impact (n = 18), and the least from a patient/family centric perspective (n = 7). Economic impact data compiled for this systematic literature review stemmed from 32 publications from the United States, Europe, Brazil, Australia, Japan, and global (international, worldwide); however, most studies were conducted in the United States (n = 10).

The heterogeneity of outcomes and study populations, and in many cases, the small sample sizes, analyzed in the publications included in this review, highlights the need for more comprehensive and carefully designed studies to fully understand the economic impact of cCMV and acquired CMV infection both on a global and region scale. Nonetheless, taken together, these data suggest that CMV and cCMV infections have a substantial economic impact on healthcare systems, patients, and society as a whole. For example, multiple studies show that the mean all-cause healthcare costs per infant are higher for those with cCMV infection compared with those without cCMV infection, and mean total costs per child are higher for cCMV positive children with symptomatic cCMV infection at birth compared to children with asymptomatic cCMV infection at birth^18,24. Mean lengths of hospital stays for infants with cCMV infection were longer than for infants without cCMV infection, with increased lengths of stays for newborns with low birth weights and other cCMV-related symptoms^22–24. Similar findings were noted for length of hospitalization and ICU stays in CMV-infected individuals in immunocompromised populations (leukemia or IBD)^20,30,34,39. In contrast to these studies, analyses from 2 other studies suggest that CMV infection did not have a significant impact on the lengths of hospital stays for patients treated for IBD or community-acquired pneumonia^27,29.

Three studies included in this review obtained data via diagnostic codes associated with cCMV from hospital and insurance claims databases in their respective countries^18,19,35. A recent review of cCMV identification in healthcare administrative databases found that although the prevalence of cCMV infection in developed countries is estimated to be 30 to 70 per 10,000 records, only 1 to 3 per 10,000 records in these countries are coded for cCMV⁴⁹. It is therefore likely that these per child total costs for cCMV diagnosis and treatment may not extend to the whole population of infants and children with cCMV infection⁴⁹. However, information from these records can still provide insight into the economic costs and charges of cCMV infection and associated neurodevelopmental sequelae⁴⁹.

The economic impact of CMV infection remains a key consideration when assessing the value of interventions and preventive strategies, such as vaccines. Five publications presented mathematical models that estimated either the direct and indirect costs or direct costs/charges of CMV infection in four countries^{17,21,25,33,45}. A total of seven studies provided cost-effectiveness or cost-benefit analyses for CMV vaccination and screening strategies: vaccinating adolescent girls in both the United States and France before their first pregnancies was predicted to be more cost-effective than not vaccinating them^43,44, as was providing universal screening of pregnant women in the United States (versus targeted screening)¹⁶and providing universal/targeted newborn screening (versus no screening) in the United States, United Kingdom, and Australia^41,42,45,46.

However, modeling of hypothetical interventions, as was the case with the majority of the economic evaluations reported here, requires a large number of assumptions to be made. Certain assumptions may have a large influence on the cost-effectiveness or cost-benefit results; for example, the cost-effectiveness of adolescent vaccination strategies were sensitive to variations in CMV seroprevalence and vaccine efficacy^42,43, whereas the cost-effectiveness of universal prenatal screening for CMV was sensitive to the estimated incidence of primary CMV in pregnancy¹⁶. In addition, model inputs regarding the benefit of antiviral treatment in reducing hearing loss as well as the costs associated with management of hearing loss, including prevention of cochlear implants, varied across studies that evaluated the cost-effectiveness or cost-benefit of newborn cCMV screening programs, making comparisons difficult^41,44–46. Further research is needed to further quantify the economic burden of CMV infection, particularly for indirect/societal costs.

A strength of this review is the inclusion of multiple populations such as infants, children, adolescents, adults, and other specific subpopulations (including patients with comorbidities that impair their immune competence such as IBD and T-cell leukemia). Examination of multiple subpopulations provides data on the disease not only in the general population but also in groups where the burden of CMV infection may be significant. Another strength of our review is the wide array of outcomes included, namely direct costs, indirect and societal costs, and resource utilization related to disease.

Limitations of this review include variations in the quality of studies reported and the lack of data from countries regarding certain economic outcomes. Additionally, the geographical search was based within the literature search; therefore, studies that were not geographically indexed (MeSH, Entree) or did not contain the country or territory in title or abstract could have been missed. This review also revealed a lack of consistent data on the economic burden of CMV infection worldwide, as well as heterogeneity among existing studies. We did not adjust cost data for inflation, as the heterogeneity among outcomes made direct comparisons across studies all but impossible. In addition, this review included only English-language publications. Exclusion of other languages may have precluded local or regional literature with pertinent information; however, two included systematic literature reviews did not limit their searches to English-only publications (one review included studies in French³⁹ and another review did not set language restrictions⁴⁰). This review mainly focused on developed countries and information on the economic burden of CMV and cCMV infections in developing regions or countries is not provided. Meta-analyses or other aggregate analyses were not performed, and the individual studies were not assessed or ranked according to methodology or data quality; these types of evaluations and were not feasible given the limited and heterogeneous nature of the data. However, presentation of the individual results descriptively enables evaluation of the available data for each country, population, subpopulation, and outcome to identify knowledge gaps and facilitate further research. Despite the limitations, our descriptive compilation of economic impact studies provides broad high-level perspective toward the current and future CMV and cCMV infection landscape, further complementing narrower critical reviews already published⁴⁹.

Conclusions

CMV and cCMV infections are globally prevalent and impose a substantial economic burden. This systematic literature review provides information on the economic impact of CMV and cCMV infections for different countries, populations, and outcomes, which can be used for further analysis and to identify gaps in data where further research may be beneficial. The majority of identified economic studies stemmed from the United States and Europe, with a minimal number of studies in Australia and Latin America. This review highlighted an overall dearth of research on indirect/societal costs, with data from only three publications identified. Additional country-level research is needed to further quantify the economic burden of CMV and cCMV infections. Finally, this comprehensive economic impact overview can provide guidance in developing, implementing, and assessing strategies to treat and prevent CMV and cCMV infections, including vaccination.

Transparency

Declaration of funding

This study was funded by Moderna, Inc.

Declaration of financial/other relationships

JDD, AN, and POB are employees of Moderna, Inc. and hold stock/stock options in the company. CT was an employee of Moderna, Inc. at the time of the study and held stock/stock options in the company. JM, MN, WL, MK, and ES are employees of Certara and were paid consultants for Moderna, Inc., for conduct of this research. EM is a consultant for Moderna, Inc. Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author contributions

MN, AN, CT and POB were involved in the study design. Data collection was performed by JM, MN, WL, MK and ES. Analysis and interpretation of the data were undertaken by JDD, EM, JM, WL, MK, ES, CT, and POB. JDD, EM, JM, MN, ES, CT, and POB were involved in the drafting and critical review of this manuscript. All authors give final approval of the published version and agree to be accountable for all aspects of the work.

Acknowledgements

Medical writing and editorial assistance were provided by Kate Russin, PhD, Jessica Nepomuceno, PhD, and Clare Lee, PhD, of MEDiSTRAVA in accordance with Good Publication Practice (GPP3) guidelines, funded by Moderna, Inc., and under the direction of the authors.

Data availability statement 

The data summarized from this review are from published articles and are publicly available.

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.