Authors: Luíza Silva Vernier, Carolina Pereira Fernandes, Pedro Pablo Skorin, Audrei Thayse Viegel de Ávila, Daniela Centenaro Levandowski
Categories: cost-effectiveness, hearing, neonatal screening, systematic review, Systematic Review
Source: International Archives of Otorhinolaryngology
Introduction Universal newborn hearing screening (UNHS) has been widely and strongly advocated as an early detection strategy for hearing loss in children. This intervention aims to prevent delays in speech and language development, which, in turn, has long-term beneficial effects on the social and emotional development and quality of life of individuals. However, the implementation of UNHS programs is circumstantial in different settings, for different reasons.
Objectives The present systematic review aimed to identify whether the implementation of UNHS programs are cost-effective, as well as their variations by localities.
Data Synthesis A search was conducted in seven PubMed (Medline), Scopus, Web of Science, Embase, CINAHL, Lilacs, and Cochrane Library. Studies that included a cost analysis of UNHS programs were eligible for inclusion. Studies on evaluations of preschool or school-based programs only were excluded, among others. A total of 1,291 records were found. Of these, 23 articles were analyzed in full. All articles identified the cost-effectiveness of the UNHS programs implemented. Regarding the UNHS protocols, a wide variation was observed in all tests used, period established between tests and retests, professionals responsible for screening, environment, and criteria for defining hearing loss, limiting the generalization of this information. All studies presented values related to the expenses with the program, but none of them showed statistical elements for the described analyzes or any theoretical basis for such.
Conclusion It is necessary to estimate local specific issues, as well as the accuracy of the chosen tests and the NHS protocols used, so that more accurate analyzes on cost-effectiveness are possible.
Keywords: cost-effectiveness, hearing, neonatal screening, systematic review
Hearing loss is the fourth largest factor for years lived with physical disability in a worldwide analysis. ^1^ The World Health Organization (WHO) estimates that, in 2050, ∼ 466 million people worldwide will have a disabling hearing loss (6.1% of the world population) and that almost 34 million (7%) will be children. ^2^ The prevalence of hearing impairment in neonates is ∼ 2 in every 1,000 live births; in ∼ 2/3 of these, the alterations are bilateral. ^3^
The impacts of hearing loss may extend throughout life. In child development, for example, it is possible to observe delays in language and/or speech, changes in school performance, personal-social maladjustments, and emotional disorders. ^4^ In adolescence and adulthood, there are limitations in social relationships, employment opportunities, and an early onset of cognitive decline.
In addition to the clinical effects of untreated hearing loss, its economic cost is substantial. The estimated annual global costs of untreated hearing loss in the healthcare sector alone exceed US 750-790 billion annually. ^5^ As such, prevention appears to be the most cost-effective way to lessen the high and growing impact of hearing loss. ^1^
Given this, we realize the importance of early identification and intervention in hearing impairment in infants. To this end, Universal Newborn Hearing Screening (UNHS) programs have been widely implemented. ^6^ The goal of these programs is to detect and rehabilitate all infants with hearing loss early, keeping false-positive result rates low to avoid unnecessary costs and decrease parental concern. The collection of data by these UNHS programs can support managerial decision making, as it allows monitoring and evaluation of the performance of the evaluated infants and programs. ^3^ Although these programs are established and standardized in most developed countries, expansion efforts to implement UNHS in other countries continue to exist.
To assess the benefits of implementing strategies or programs, a cost-effectiveness analysis should be conducted, allowing decision-makers to clearly understand the trade-offs in costs, harms, and benefits between alternatives, which should be combined into a single metric, the Incremental Cost-Effectiveness Ratio (ICER). This metric can also be used to inform decision-making when there are limited resources. Therefore, we understand that one of the relevant factors to determine the success of a UNHS program is the cost-effectiveness ratio ^7^ to identify the factors that affect its performance as a whole. ^8^
Public policy makers have a position regarding the attention given to cost-effective interventions in hearing health, aiming to reduce the consequences of hearing loss. ^9^ The analysis of these data can gather information for resource allocation and potentiate investment and prioritize interventions. ^1^ But many countries still do not include UNHS programs in their health agenda, partly because they are considered too expensive or because their value is questioned. ^10^ In countries where the implementation of UNHS programs takes place, there are variations in the approach and methods used, which can be attributed to a difference in available resources, financial or technological, but also to the lack of universal guidelines to be followed, to ensure a consistent approach in the implementation of these programs. ^11^
Therefore, the present systematic review aimed to identify, in the literature, whether the implementation of UNHS programs is cost-effective, as well as their variations by location. We also tried to identify the differences between the protocols used, as well as the quality indicators of UNHS programs.
The present review was registered in the International Prospective Register of Systematic Reviews (PROSPERO), under the number CRD42021257857, and is presented according to the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-analyzes - Extension for Scoping Reviews) guidelines. ^12^ ^13^
A protocol for the literature search was structured ( Supplementary Material Appendix 1 ).
The design of the studies had no restrictions. On the other hand, to be included in the present review, studies had to be original and present a formal economic analysis of UNHS programs, with descriptions of costs, cost analyzes, descriptions of cost outcomes or complete economic evaluations. Studies with evaluations of Hearing Screening programs for children in general, and not right after birth; that aimed to compare two or more protocols to perform UNHS; that presented the clipping of only one stage of UNHS; searches made based on simulated populations, and articles that limited the population to perform Neonatal Hearing Screening (NHS) were excluded.
Seven scientific databases were considered for the PubMed (Medline), Scopus, Web of Science, Embase, CINAHL, Lilacs, and Cochrane Library. These databases were independently and simultaneously consulted by two researchers in May 2021. There was no limitation regarding publication date, language and/or geographical location. All records were considered as eligible for inclusion if they had an abstract. After reading the titles and abstracts, the included articles were forwarded for analysis of the full text. From the exclusions made at this stage, the reference lists of all eligible studies were examined to include any additional studies relevant to the objectives of the present study, provided they met the eligibility criteria. ( Fig. 1 )
Fig. 1 PRISMA 2020 flow diagram for new systematic reviews that included searches of databases and registers only.
The descriptors were chosen through Medical Subject Heading (MeSH) and were integrated into the search strategy, being related to the theme of the Newborn Screening, Hearing and Cost-Effectiveness. The full search algorithms for each database can be found in Supplementary Material Appendix 2 . The overall search strategy was devised by the study authors, consisting ( Neonat * screening
screening *) AND ( Hear *) AND ( Costs and Cost Analysis OR Cost Control
Two authors of the study simultaneously, independently, and blindly performed the initial searches in the databases. After the elimination of duplicate records, all titles and abstracts of the articles were independently evaluated to verify the possibility of inclusion, according to the eligibility criteria. Differences regarding inclusion or exclusion were resolved with the evaluation of a third author, who acted as a judge. The same happened in the phase of inclusion/exclusion of articles after full reading.
The data from the selected articles were extracted for registration and compilation of information in a spreadsheet made for the present study in Microsoft Excel (Microsoft Corporation, Redmond, WA, USA).
All eligible articles evaluated UNHS programs and contained data such as the region of implementation of the program analyzed, components of the screening protocol followed by professionals, the economic analyzes performed and the cost-effectiveness perspective of the program. The primary outcome of interest was whether the implemented UNHS programs were evaluated as cost-effective, supported by economic cost-effectiveness relationships. Monetary value rates were reported when available. Items without dollar costs had the values converted to this currency using the Purchasing Power Parity (PPP) rate. ^14^ After the conversation to dollar, the implicit price deflator of gross domestic product ^15^ was used to adapt prices to 2012 levels, because the last economic analysis of the selected articles is from that year, minimizing distortions in the evaluation.
No quantitative analyzes and meta-analysis were performed in the present review due to the heterogeneity of the eligible studies and the scarce data available on cost-effectiveness. A qualitative analysis of the findings of the reviewed studies was performed individually, comparing the outcomes of interest to the present study.
The Consolidated Health Economic Evaluation Reporting Standards (CHEERS) ^16^ parameters were used to assess the methodological quality of the reviewed studies. Each cost-effectiveness analysis reported by each article included in the review was evaluated considering the CHEERS checklist, a tool developed by the International Society for Pharmacoeconomics and Outcomes Research, with 24 items subdivided into 6 main title and abstract; introduction; methods; results; discussion; and other. The items analyzed in each reviewed study were binarily scored. Each study received an aggregate score according to adherence to the criteria and any methodological concerns regarding each study were described.
A total of 1,291 records were found. Initially, the duplicate records were removed, leaving 773 articles for analysis of the title and abstract. After the initial reading, 721 articles were excluded for evading the proposed theme ( n = 518), being reviews ( n = 121), being in formats other than articles ( n = 69) or unavailable ( n = 13), resulting in 52 studies for the full text analysis. After their full-text reading, 29 articles were excluded for not meeting the eligibility criteria for cost-effectiveness analysis of UNHS programs. Therefore, 23 articles were included in the present review.
As one of the inclusion criteria adopted, all studies performed a cost-effectiveness or economic analysis of some UNHS program. The studies included in the present review were published between 1995 and 2017 and depicted data from several countries, especially United States of America (9 studies). But the included studies also covered countries from other continents, being characterized by heterogeneity. ( Table 1 ).
Variations and divergences were recorded between the protocols used to perform UNHS in the different sites and variations in their structures. Information on the team responsible for performing the UNHS was not present in four studies; in the others, this data was present, and the types of professionals varied, for example otolaryngologist, audiologist, and nurse.
In each study, the rates of coverage, referrals for retesting and diagnosis, identified hearing loss, and false-positive results were investigated. All reported program coverage (rates that were reported ranged from 25.68 to 99.5%); on the other hand, the least presented information was the false-positive rate.
The data on the UNHS flow varied, but most studies pointed out the execution of the first test within the recommended period, that is, before hospital discharge. Screening procedures also varied regarding the tests used, screening algorithms, audiometric frequencies tested, expected time between test and retest, hearing loss definition criteria, and frequencies indicated for screening. The Otoacoustic Emissions (OAE) was the most used screening test, followed by the Automatic Brainstem Auditory Evoked Potential (BAEP-A). All these data are presented in detail in Table 2 .
All the included studies mentioned a cost-effectiveness analysis, but none presented statistical elements for the analyzes described or to support the economic analyzes based on indicators better known in Health Economics, which are part of the Health Adjusted Life Years (HALY) group. The objective of this group of indicators is to present the well-being of the individual in the form of years, also considering the quality of life experienced. Included within HALY are Quality-Adjusted Life Years (QALY), an indicator that represents the well-being of the individual, and Disability-Adjusted Life Years (DALY), which calculates not the social utility of the individual, but rather their disability. Both are widely used as a cutoff point to identify acceptable strategies in the cost-effectiveness ratio. On the other hand, the ICER is an important incremental data, since it compares two different scenarios and guides resource allocation considering the best one. But this metric was not reported in the reviewed studies either.
The costs presented were reported in different local currencies, requiring first conversion and then deflation to allow some sort of comparison. The amounts were reported in international 2012 dollars.
The most present information in the reviewed studies were the expenses of the NHS program, expenses with the first hearing assessment of NHS, with retesting, diagnosis, intervention, and operational expenses – those related to the expected expenses to produce products and services. The sources of funding for identified UNHS programs, usually from national or local research or projects. Specific information on economic analysis identified in the implementation of NHS programs is in Table 3 .
There was no divergence between the reviewed studies regarding the positive cost-effectiveness ratio, considering the implementation of UNHS programs. There is an assertion of the use of these data to improve the programs and identify their weaknesses. The importance of reflecting on the cost-effectiveness of UNHS is perceptible, given the scarcity of epidemiological data on issues of this practice, such as the prevalence of hearing impairment in the population, and what is attributable to congenital hearing loss, often not detected and not treated early. And also due to the scarcity of audiological care data in children with Risk Factors for Hearing Loss (RFHL), among others.
Having a data system that associates the initial results of UNHS and the diagnostic assessments made in rehabilitation services is still a challenge for the best settings in the implementation of UNHS. Few articles mention information on data storage of UNHS programs and those that exposed this information were usually linked to databases of the very places of execution of this screening. In these sites, plausible factors for improvement were identified, such as control of neonate follow-up, logistical issues, and the false-positive rate. Table 4 identifies the conclusions of the economic analysis of each study reviewed, looking for positive and negative points and perspectives, as well as how UNHS data is stored.
According to the Joint Committee on Infant Hearing (JCIH), BAEP and OAE tests are recommended for screening children's hearing. The OAE is indicated for early identification of hearing impairment in newborns, and the BAEP-A should be done when, regardless of the results of the OAE test, the newborn has any RFHL, or for newborns who have failed the OAE. But there are several choices in the organizations of UNHS programs and in the design of protocols, as reported in the included studies. In a recent study ^40^ it was possible to identify that a two-stage OAE-BAEP-A protocol results in referrals for diagnostic tests four times more frequently when compared with the three-stage OAE-OAE-BAEP-A protocol. On the other hand, the application of several screening stages may increase the number of lost cases of hearing impairment in each assessment stage, either by non-attendance of the babies in the next assessment or by false-negative results.
This diversity may stem from the various existing programs, which may make it difficult for countries and institutions wishing to develop UNHS programs according to a single protocol to join, as they generally seek examples to serve as a guide for implementation. These differences in protocols should not confuse but rather guide the use and development of relevant protocols, ensuring that implemented UNHS programs reach benchmark standards. There are many reasons why countries choose one recommended protocol and not another. These relate to social context, available resources and constraints imposed in health settings. But the final choice should be limited not by resource constraints, but by the current evidence in the literature.
In addition to the type of program implemented, it is important to analyze its coverage. This is a qualitative measure, indicating that UNHS is available to all eligible infants born at the administering hospital or living in the administering region or country. Therefore, this reflects the quality of access and adherence to the UNHS program. There is inherent variability in the quality and validity of data in the selected studies, as program coverage encompassed different percentages. Of the 23 studies selected, 18 reported the percentage of coverage. Among these, three showed coverage ≥ 95% in their entirety and 1 only in the first phase. Five studies presented indexes ≥ 97%, and only 1 reported this index in the second stage. In the other articles, the index ranged between 25.68 and 93.60%.
It is extremely important that UNHS program managers control quality indicators through the percentage of all newborns who complete this assessment within the 1 ^st^ month of life, as well as all newborns who did not pass the initial hospital screening and required a new outpatient assessment, those who did not pass any subsequent new checks before audiological assessment, and the percentage of newborns who did not pass the initial screening and were then reassessed, ^5^ among others. Quality indicators for each phase are defined by scientific institutions and should be used as a parameter to help control the effectiveness of programs, determining whether interventions represent an improvement, and to ensure that improvements are sustained.
Methods for identifying areas for improvement, evaluating the selected protocol design, and ensuring the effective use of resources should be decisive components in all UNHS programs. The quality of the analysis ensures the support needed to achieve public policy efficiency and is directly related to the quality of the management system for the data explored in the research. A performance evaluation program that intends to present complete results, with a deep difference in costs and patient monitoring, requires an efficient organization of information. The absence of such a system may impair data accounting and compromise the evaluation of the flow, a factor that was evident in the heterogeneity and the omission of data found in some selected articles. It is not enough just to implement the program, but also to evaluate its performance in relation to costs, benefits to users and savings in the use of health resources, for example.
For countries implementing UNHS, the only method of ensuring effective use of resources is to collect data, monitor, and evaluate screening program performance, which can be done through digital platforms with broad access and communication. Thus, the quality of the program data management system is very important, and this aspect is still a cause for concern, as many structured programs do not monitor or evaluate their performance.
From the present review, it is clear that, in different countries, the professionals responsible for UNHS are also distinct, which includes trained volunteers or certified professional screeners. Detailed requirements for the training and skills development of these assessors are provided by some adopted protocols. The JCIH guidelines present comprehensive guidance on the role and competencies of different professionals as part of a multidisciplinary team approach, including audiologists, physicians, and nursing staff, to ensure the delivery of UNHS. This emphasizes the importance of the audiologist experienced in the assessment of newborns and young children, and the supervision of each component of the hearing screening program, especially in its implementation and, whenever possible, in each hospital.
In the present study, no statistical economic analysis was found that could prove the evidence of the cost-effectiveness of the programs. Only self-reported information on this cost-effectiveness relationship was found. Therefore, it is important to highlight that the principle of optimization and investment begins with economic analysis, and the impact on health can be considered when a budget is analyzed, or the minimization of costs considering an outcome. The present study found that the implementation of UNHS programs can have a positive impact on health services by identifying hearing loss early. Thus, they are cost-effective, especially for their results that have repercussions throughout the lives of the individuals. It is important to stress the importance of defining the context and its characterization, as well as the selection of the protocol to be used for structuring the UNHS program. On the other hand, as more countries and organizations develop their UNHS programs, there is a constant review of existing information and benefits of its applicability. ^41^
In analyzing the data, the expectation was to find the the expected mean QALYs and costs for the base-case cohort, the ICER to analyze screening strategies, and information on DALYs, with data on disability-adjusted years of life lost, to ensure that the intervention is cost-effective.
Hence, one can see the importance of including this unaddressed information, as these data show where governments need to act to strengthen their health systems and protect people from health care consequences and potential costs.
Even with the constant omissions of data, the results of the present review indicate that UNHS programs are generally cost-effective; that is, they state that the orientation to implement UNHS programs by analyzing resource allocation and comparing the efficiency of proposed interventions is positive.
Funding The authors received no financial support for the present research.