Italian validation of the Hearing Handicap Inventory for Elderly – Screening version (HHIE-S-It)| ACTA Otorhinolaryngologica Italica

Abstract

Objective. Validate the Italian version of the Hearing Handicap Inventory for Elderly - Screening version (HHIE-S-It).
Methods. After translation, psychometric properties and attributes were analysed by administering the HHIE-S-It to 167 elderly outpatients together with the Psychological General Well-Being Index (PGWBI).
Results. The Cronbach’s α coefficient was 0.908 for the total score, and 0.832 and 0.816 for its two subscales. Significant test-retest reliability was observed (p < 0.001). Moderate to high correlations were found between HHIE-S-It and pure tone average in the better ear (p < 0.001). The ANOVA test confirmed the significant difference in HHIE-S-It scores across groups according to the degree of hearing loss (p < 0.001). Only very low and low significant correlations were observed between HHIE-S-It and PGWBI. The criterion HHIE-S-It > 11 was observed as the best cut-off with highest sensitivity (86.4%), specificity (72.4%), positive predictive value (52.8%), negative predictive value (93.7%) and likelihood ratios (3.12 and 0.19).
Conclusions. Since the HHIE-S-It presented acceptable psychometric properties, its adoption is justified for both clinical and research purposes. Acceptable diagnostic attributes allow its use as a screening tool for age-related hearing loss.

Introduction

Age-related hearing loss (ARHL) or presbycusis refers to alterations of the central and peripheral auditory system that occur with ageing and that lead to hearing loss (HL) and difficulty in understanding spoken language. Its pathogenesis is complex and multifactorial. In fact, it originates from anatomical and functional changes, but other contributing factors include genetic alterations, hormones, exposure to ototoxic agents or loud noises, ear or brain infections and the presence of certain systemic diseases ¹. ARHL is the most prevalent condition affecting the elderly worldwide and the leading cause of disability for individuals older than 70 years ². In Europe, it is estimated that 20% to 30% of the elderly suffer from HL, defined as a hearing threshold above 30 dB, but it is impossible to establish the exact prevalence of presbycusis, due to the epidemiological consideration of other aetiologies of HL and the heterogeneity of data reported ³. However, two aspects need to be considered. Firstly, it is reasonable that audiology services are predicted to be increasingly overwhelmed by the increasing number of hearing-impaired elderly patients in the future, due to the progressive ageing of the general population. The second is that presbycusis is strongly associated with decline in cognitive functions, dementia, depression, and social isolation, as well as loss of communication ability and worsening social adjustment ⁴. Hence, the immediate and long-term impacts on health, society and the economy are already relevant ⁵. According to this scenario, ARHL must be addressed as a public health issue and an early identification of hearing disorders among the elderly is needed. Even if up to now no universal process for screening for presbycusis exists ⁶, the Hearing Handicap Inventory for Elderly - Screening version (HHIE-S) is one of the tools that can be used to determine whether a patient has a potential HL and needs further evaluation or audiological assessment ⁷. It was developed from the Hearing Handicap Inventory for Elderly (HHIE), an instrument designed to quantify the effects of hearing impairment on the emotional and social adjustment of elderly people ⁸. Although successful validation studies in the most widely spoken languages have been carried out, assessing psychometric proprieties and attributes of the HHIE-S, a validated Italian version is not yet available ^9-12.

Therefore, the first aim of the present study was to validate an Italian version of HHIE-S and to test its psychometric properties in a sample of outpatients referred to audiological services. As a second goal, we investigated whether the Italian version of the HHIE-S is an appropriate screening tool, focusing on its attributes and the threshold that can adequately discriminate normal from hearing-impaired elderly.

Materials and methods

Participants

The study design was cross-sectional and monocentric. A cohort of 167 consecutive outpatients attending the audiological centre at the Otorhinolaryngology Unit, Azienda Ospedaliero-Universitaria of Modena (Italy), were recruited from May 2020 to February 2021. They all shared the following minimal criteria: age ≥ 65 years, speaking Italian as their first language, willing to participate in the study, and not institutionalised or hospitalised in any health structures. No segregation of cases was made according to type, unilateral or bilateral HL. The following exclusion criteria were considered: cultural barriers or presence of major psychiatric or neurological disorders, subjects with fluctuating HL and balance disorders or wearing hearing aids/cochlear implants. Thus, the recruitment process was aimed at selecting a sample that could be representative of the community-dwelling Italian geriatric population. The minimal sample size was established according to a predetermined respondent-to-item ratio of 10:1 ¹³.

Procedures

After physical examination and immittance test, tonal audiometry was performed for each patient. The audiological evaluation was conducted using a two-channel diagnostic audiometer (INVENTIS Piano Plus, INVENTIS SRL). The patient was seated within a double-wall, soundproof booth that complies with ISO 8253 and in the window view of the examiner. The audiometric control board was visible only by the latter, who performed the evaluation following the Hughson-Westlake method to obtain air conduction thresholds for each ear. The pure tone average threshold (PTA) was computed for both ears considering the four frequencies (500 Hz, 1000 Hz, 2000 Hz, 4000 Hz). For statistical analysis, PTA of the better ear was used. According to the degree of hearing loss classification of the American Speech-Language-Hearing Association (ASHA) ¹⁴, a PTA ≤ 15 dB HL means normal hearing, 16 dB HL ≤ PTA ≤ 25 dB HL slight hearing loss, 26 dB HL ≤ PTA ≤ 40 dB HL mild hearing loss, 41 dB HL ≤ PTA ≤ 55 dB HL moderate hearing loss, 56 dB HL ≤ PTA ≤ 70 dB HL moderately severe hearing loss and a 71 dB HL ≤ PTA < 90 dB HL severe hearing loss, whereas a PTA over 90 dB HL represents profound hearing loss. The World Health Organization (WHO) definition of disabling hearing loss was considered to analyse the correlations between the questionnaire and hearing function. In particular, it refers to a PTA greater than 35 dB in the better hearing ear to separate patients with hearing disability from those without ¹⁵.

Questionnaires

The HHIE-S is a 10-item questionnaire that is self-administered by paper-and-pencil. It is composed of two 5-item subscales that assess the self-perceived impact of hearing loss in the emotional (E) and socio-situational (S) domains respectively. A “yes” response to an item is awarded four points, a “sometimes” two points and a “no” zero points. The score is a sum of all responses and range from 0 to 40 points indicating an increasing level of perceived handicap. The original version of HHIE-S was translated through the iterative translation process that complies with the standards established by the International Quality of Life Assessment project (IQOLA) ¹⁶, for academic use and under licence by the original Authors. Two bilingual audiologists translated the HHIE-S items into Italian. Conceptual equivalence, wording, comprehensiveness and feasibility of data collecting were assessed administering the provisional version to a sample of 10 healthy people. At the same time, the backward translation into English by two independent bilingual audiologists was performed. After comparison of different versions obtained with the original HHIE-S, a revision of the items was carried out. Therefore, a preliminarily administration was conducted in a pilot sample of 25 hearing-impaired patients with varying educational and social status, in a face-to-face interview format. Since few doubts regarding comprehensiveness were noted, only a minimal revision was performed and the definitive Italian version of the Hearing Handicap Inventory for Elderly – Screening version (HHIE-S-It) was obtained (Appendix A). In order to provide evidence of criterion-related validity of the HHIE-S-It, all patients were also asked to complete the Italian version of the Psychological General Well-Being Index (PGWBI) ¹⁷. It is an already validated instrument that measures the subjective perception of well-being across six domains: anxiety, depression, positive well-being, self-control, general health and vitality. It is composed of 22 items differently presented, such as statements or questions referring to the last four weeks of the subject’s lifetime, with responses reported on a six points Likert scale, with a score ranging from “0” to “5”. Each item assesses self-representations of intrapersonal affective or emotional states reflecting a sense of subjective well-being or distress. Response options and the scoring direction differ in order to inversely count answers of items regarding the distress. The items belonging to the same domain are balanced with respect to the direction of the responses. Since each domain is defined by a minimum of three to a maximum of five items, scores for each domain range from 0 to 15, from 0 to 20 or from 0 to 25. The total score, obtained from the sum of all responses, can reach a maximum of 110, representing the best achievable level of well-being.

Statistical analysis

All data were gathered anonymously in a Microsoft Excel^® database. Statistical analysis and graphical representation were performed with the Statistical Package for Social Sciences (IBM SPSS^® version 25.0 for Microsoft Windows^®). For epidemiological features, quantitative and qualitative variables were expressed as means with standard deviations (SD) and rates, respectively. Independent t-test was used to compare the continuous variables, whereas qualitative variables were compared using the Chi-square test. The score of the HHIE-S-It, its items, and subscales were also reported with descriptive statistics.

In order to fulfil the HHIE-S-It validation process, it was crucial that its psychometric proprieties were analysed ¹⁸. Thus, internal consistency, reliability, construct and criterion-related validity were evaluated. The internal consistency refers to the homogeneity of the questions comprising the questionnaire. It was assessed with Cronbach’s α coefficient where higher values of α mean a greater correlation of item to each other (inter-item correlations). While the use of cut-off is debated, in accordance with the international interpretation, the minimum acceptable of α is 0.7, an α from 0.8 to 0.9 is considered good and an α since 0.9 excellent ¹⁸. In addition, to verify if the removal of one or more items from the scale improves overall α coefficient, the statistical procedure “Cronbach’s α if item is deleted” was adopted. The reliability is essential to measure the stability of the scores over time. It indicates the confidence in achieving the same or similar responses if the questionnaire is distributed again to the same sample. In this study, the test-retest method by comparing scores across a repeated administration to 80 selected subjects after a six-week interval was adopted. Responders had no change in their hearing ability and the HHIE-S-It was administered again in a phone-interview format. The Pearson’s correlation coefficient and the intraclass correlation coefficient (ICC) were evaluated. The predetermined minimally acceptable value was set at 0.7 for both of them. In addition, the ceiling and floor effect were calculated by percentage frequency of highest or lowest possible score achieved by respondents respectively. If present, a percentage more of 15%, the reliability is reduced due to the difficulty in distinguishing patients with the lowest or highest possible score from each other.

The construct validity refers to the ability of an instrument to measure a theoretically derived hypothesis. Weinstein et al. assumed that higher scores of HHIE-S should correlate directly to the degree of hearing loss ⁷. In this study, Pearson’s correlation coefficient was used to disclose the presumed correlation between PTA and the total score of HHIE-S-It and its subscales. The criterion-related validity refers to the extent to which scores relate to a gold standard instrument in order to assess if the total score and its subscales were correlated with six domains of the aforementioned PGWBI (anxiety, depression, positive well-being, self-control, general health and vitality). Pearson’s correlation coefficient was used. Since HHIE-S-It measures the increasing level of perceived hearing handicap and PGWBI measures the level of well-being, negative correlations between the two instruments were expected. The level of statistical significance was considered reached if the p-value was < 0.05 in all procedures.

Finally, since the author proposed the HHIE-S as a screening tool to detect the degree of hearing complaints ⁷, in the present study the predictive validity was also assessed. It refers to the extent that it is valid to use the test score to predict the presence of a target condition. Over the years, several studies have evaluated the diagnostic attributes of HHIE-S in terms of sensitivity, specificity, predictive values and likelihood ratio ^{9,10,12,19-21}. However, most used different age populations and different classifications of hearing loss. Thus, to analyse the diagnostic ability of the HHIE-S-It and its discrimination threshold, a Receiver Operating Characteristics curve (ROC) was generated. Youden’s index J was applied to calculate the best cut-off for the ROC curve considering the WHO definition of disabling hearing loss ²². Therefore, the diagnostic properties were evaluated considering cut-offs used in previous studies and in present one ²³. Since the target condition in our case was the presence of hearing loss, the sensitivity and specificity are the ability to correctly identify patients with and without hearing loss. While the former is the ratio of patients with hearing loss and positive on the screening test to the number of all those with hearing loss, the latter is the ratio of patients without hearing loss and negative on the screening test to the number of all those without hearing loss. Sensitivity and specificity are intrinsic features of the test that are related to its accuracy and remain constant as long as the target condition and the test criterion remain constant. By themselves, however, they do not provide the basis for informed decisions following screening test results. In fact, positive test results could contain many false positive outcomes that are ignored in determining sensitivity. Similarly, negative test results could contain many false negative outcomes that are ignored in determining sensitivity. Hence, according to this limit, predictive values were also assessed, which are functions of the target condition’s prevalence in addition to sensitivity and specificity. Predictive values indicate the likelihood that a test can successfully identify whether people do or do not have hearing loss, respectively, based on their test results. In detail, the positive predictive value (PPV) is the ratio of test-positive subjects truly affected by hearing loss to the total number of patients who were positive on the screening test. Similarly, the negative predictive value (NPV) is the ratio of test-negative subjects truly not affected by hearing loss to the total number of patients who were negative on the screening test. In addition, likelihood ratios were assessed. In particular, the likelihood ratio for a positive test result is the ratio of the probability of being a positive screening result for patients with hearing loss to the probability for those without. Similarly, the likelihood ratio for a negative screening test result is the ratio of the probability of being a negative screening result for patients with hearing loss to the probability for those without. Thus, larger values for the likelihood ratio for a positive screening result are more desirable, as they indicate a higher accuracy of the screening test in obtaining positive results in those with hearing loss than in those without. Conversely, smaller values for the likelihood ratio for a negative result are more desirable, as they indicate a higher accuracy of the screening test in obtaining negative screening results in patients without hearing loss than in those with hearing loss.

Results

Of the 167 outpatients recruited for the study, 80 (47.9%) were males and 87 (52.1%) females. The mean age was 73.9 years (SD: 5.9; range: 65-90), whereas the mean value of PTA was 28.4 dB nHL (SD: 15.1; range: 5-120). Epidemiological features of the sample are represented in Table I. Considering PTA in the better ear and the ASHA classification of degree of hearing loss, 30 patients (17.9%) had normal hearing, 67 (40.1%) had a slight hearing loss, 37 (22.2%) had a mild hearing loss, 26 (15.6%) had a moderate hearing loss, 7 (4.2%) had a moderately severe hearing loss. No participant had a severe or profound hearing loss (Fig. 1). Males were not significantly older than females (t = 1.883, df = 164.6, p = 0.061), but the degree of hearing loss in the better ear was significantly worse in males (t = 2.504, df = 134.1, p = 0.013). The score of the HHIE-S-It and its subscales are reported by descriptive statistics in Table II. No significant difference was found between genders regarding the HHIE-S-It total score (t = 1.486, df = 152.1, p = 0.139), the S score (t = 1.830, df = 160.2, p = 0.069) and E score (t = 1.719, df = 156.4, p = 0.088).

Internal consistency and reliability

Cronbach’s α coefficient was 0.908 for the total score, and 0.832 and 0.816 for the E and S subscales respectively. In the procedure “Cronbach’s α if item is deleted”, removal of item 6 improved overall α coefficient of the total score to 0.916 and of the S scale to 0.854. After its removal, no other item increased α again. The Pearson’s correlation coefficient ρ between the total scores of the HHIE-S-It and its subscales was 0.965 for the E subscale (2-tail significance < 0.001) and 0.967 for the S subscale (2-tail significance < 0.001) respectively. A significant correlation between the two subscales was identified also: the ρ was 0.867 (2-tails significance < 0.001). The test-retest reliability of HHIE-S-It and its subscales, evaluated by the ICC and Pearson’s correlation coefficient, was found to be correlated at a significant level as shown in Table III. The floor effect was 21.6%, 33.5%, 23.9% for total score of the HHIE-S-It, E and S subscales respectively, while the ceiling effect was 2.9%, 2.4%, 2.9% for the three different scores.

Construct validity

Pearson’s correlation coefficient between PTA in the better ear and HHIE-S-It total score was 0.628 (2-tail significance < 0.001). Highly significant correlations were also identified between PTA and the E and S subscales of the HHIE-S-It: ρ was 0.544 (2-tail significance < 0.001) and 0.643 (2-tail significance < 0.001) respectively. As confirmed by the one-way ANOVA test, the HHIE-S-It total score (F = 26.3; p < 0.001) and scores of its two subscales (F = 19.1; p < 0.001 for E; F = 29.5; p < 0.001 for S) were significantly higher according to degree of hearing loss (Fig. 2).

Criterion-related validity

Negative correlations were found between the HHIE-S-It and its subscales (S, E) scores and all domain scores of the PGWBI (Tab. IV). In particular, significant correlations were identified between the HHIE-S-It total score, S subscale and the Vitality of the PGWBI (p < 0.005), between the E score and Vitality domain (p < 0.05), and between the three scores of the HHIE-S-It and Self-Control domain (p < 0.05). Moreover, significant correlations were seen between total score and S subscale of the HHIE-S-It and the total score of the PGWBI (p < 0.05).

Predictive validity

In the sample of the present study, the prevalence of disabling hearing loss was 26.3%. The ROC curve showed an AUC of 0.860, which is significantly higher than (p < 0.001) chance, i.e. AUC = 0.5 (Fig. 3). The Youden index J was 0.587 and the associated criterion was HHIE-S-It total score > 11, corresponding to a sensitivity of 86.4% and a specificity of 72.4%. PPV and NPV were 52.8% and 93.7%, respectively. LR+ was 3.12 whereas LR- was 0.19. According to previous studies, a criterion of HHIE-S-It total score > 8 was considered ^{7,9,10,12,19,21}. In this case, a sensitivity of 86.4% and a specificity of 67.7% were observed. PPV and NPV were 49.4% and 93.3%, respectively. LR+ was 2.68 and LR- was 0.20.

Discussion

In this study, 167 elderly outpatients were recruited in accordance with the predetermined respondent-to-item ratio of 10:1 ¹³. Only subjects aged 65 years or older without hearing aids or cochlear implants were considered. While males were not significantly older than females (p = 0.061), the degree of HL in the better ear was significantly worse in males (p = 0.013). Since the prevalence of the HL is nearly twice in males, this result was not surprising ²¹. However, no significant differences were found between sex regarding the score of HHIE-S-It and S and E subscales (p = 0.139; p = 0.069; p = 0.088). Hence, the statistical analysis was conducted considering the entire sample regardless of gender. According to the ASHA’s HL degree classification ¹⁴, 137 (82%) subjects presented HL but most, 130 (77.8%), had slight to moderate HL. None had severe or profound HL. Differently, considering the actual definition of the WHO ¹⁵, the prevalence of disabling HL was 26.3% (44 subjects). This discrepancy gives an idea of how the use of different types of definitions and classifications influences the epidemiological data of the ARHL.

Other validation studies considered psychometric properties ¹¹, attributes as screening tool ⁹ or both ^10,12. The internal consistency was good to excellent since Cronbach’s α coefficient was 0.908 for the total score, and 0.832 and 0.816 for the E and S subscales, respectively. Table V compares the internal consistency in present study with that observed in other validation studies ^10-12. Only the removal of item 6 slightly improved the α coefficient of the total score and S subscale (0.916 and 0.854). This item “Does a hearing problem cause you to attend religious services less often than you would like?” investigates aspects most complained about by elderly hearing-impaired patients, such as speech perception in competing background noise, deteriorated signal and challenging social environments. Moreover, as observed during preliminary administration, patients were inclined to misunderstand the question, answering based on their usual attendance at religious services. Nevertheless, according to Ventry and Weinstein, it is “necessary and desirable” to identify all specific situations that could be detrimental to social adaptation, although not so frequently encountered by older people in their everyday life ⁸. Hence, considering these aspects and the slight improvement of Cronbach’s α coefficient, item 6 was not removed from the HHIE-S-It. A floor effect was observed for HHIE-S-It and its subscales whereas no ceiling effect was reported. This could lead to difficulty in distinguishing patients with the lowest possible scores and may be considered a slight limit. The test-retest reliability was used to measure the reproducibility of the instrument. It was adequate since a significant ICC and Pearson’s correlation coefficient were found between the HHIE-S-It and its subscales (Tab. III).

Ventry and Weinstein developed the HHIE-S to measure hearing disability. They assumed that higher scores should correlate directly with the degree of HL ⁷. In the present study, very high correlations between the total score and S and E and between the two subscales were observed (p < 0.001). In addition, moderate to high correlations were found between PTA in the better ear and HHIE-S-It (p < 0.001) and between PTA and the two subscales (p < 0.001). The ANOVA test confirmed the significant difference in HHIE-S-It scoring across groups according to ASHA’s HL degree classification (p < 0.001). Figure 2 shows this difference graphically. These observations confirmed the construct validity of the instrument. The criterion-related validity was also assessed administering in addition the PGWBI, an already validated instrument. Since the HHIE-S-It measures the increasing level of perceived hearing handicap and PGWBI measures the level of well-being, negative very low and low significant correlations were observed between HHIE-S-It and Self Control (p < 0.05), HHIE-S-It, S and Total score of PGWBI (p < 0.05), HHIE-S-It, S and Vitality (p < 0.005), E and Vitality (p < 0.05). This shows that if quality of life needs to be considered, a specific questionnaire should be given in addition to the HHIE-S-It.

Since the HHIE-S is intended as a screening tool ⁶, several studies have evaluated the diagnostic attributes ^{9,10,12,19-21}. Most used different age populations and different classifications of HL ²¹. Lichtenstein et al. were among the first to evaluate the diagnostic performance of the HHIE-S by considering different definitions of HL ¹⁹. They codified the presence of hearing disability according to the instrument scoring. A score from 0 to 8 was considered indicative of no hearing handicap, from 10 to 24 of mild to moderate handicap, and a value from 26 to 40 indicative of severe handicap. Subsequently, scores above 8 became indicative of the presence of hearing impairment, and the diagnostic properties of the instrument, even in other validation studies ^9,10,12, were analysed by considering a cut-off > 8. However, the WHO recently changed the definition of disabling HL to the presence of a PTA greater than 35 dB in the better hearing ear ¹⁵. Thus, in the present study the actual WHO definition was considered. The ROC curve and Youden’s index J were applied to calculate the best cut-off for the HHIE-S-It. The criterion total score > 11 was observed as the best cut-off with highest sensitivity (86.4%), specificity (72.4%), PPV (52.8%), NPV (93.7%), LR+ (3.12) and LR- (0.19). In addition, the diagnostic properties of the criterion total score > 8 were analysed and were found to be smaller.

While many years have passed and different classifications of the degree of HL have emerged, Figure 4 shows that the HHIE-S-It with criterion total score > 11 has diagnostic properties comparable to those of the original and most widely spoken language versions. In fact, the HHIE-S-It presented Likelihood ratios (LRs) comparable to those observed in other studies. Tomiaka et al. ¹⁰ reported LR+ and LR- equal to 3.60 and 0.20 considering hearing loss as a PTA greater than 40 dB in the better ear and LR+ and LR- equal to 2.90 and 0.70 considering hearing loss as a PTA greater than 25 dB in the better ear. Servidoni and Conterno ²⁰ reported a LR+ and LR- of 3.56 and 0.15 considering hearing loss as a PTA greater than 40 dB in the better ear.

Furthermore, some aspects need to be taken into consideration. The AUC of 0.860 indicated a very good diagnostic accuracy. The observed LRs indicated that HHIE-S-It could be considered a valid screening tool ²⁴. Indeed, screening tests typically should present some advantages compared to diagnostic tests, such as involving less load on the health care system, being more affordable and available, as well as less invasive, less dangerous, less time-consuming, and less uncomfortable for patients ²³. Screening tests are also well-known for being imperfect and sometimes ambiguous. Hence, sensitivity and specificity should be considered appropriated as comparable to HHIE-S and its most widely spoken language versions. Predictive values should be considered appropriated in addition. In particular, the moderate PPV (52.8%) might be acceptable since that diagnostic test is easily and quickly performed, and not stressful, expensive, or unavailable ²³. Furthermore, a high NPV might be acceptable because ARHL is a serious, often asymptomatic or underestimated issue that can be treated effectively saving large amounts of public financial resources ²⁵.

Finally, two limitations should be taken into account. The floor effect observed could lead to a difficulty in distinguishing patients with the lowest possible scores. The HHIE-S-It cannot provide any information about the outcome of hearing aids since no patient was wearing them in our sample.

Conclusions

In the current study, we translated and validated the Italian version of the HHIE-S (HHIE-S-It). Psychometric proprieties and diagnostic attributes were assessed properly in a sample of 167 elderly outpatients considering the current hearing loss degree classification. Cronbach’s α coefficient was excellent for the total score and good for its two subscales. While test-retest reliability and construct validity were acceptable, the criterion-related validity showed only very weak to weak correlations with some domains of an instrument that measures the level of well-being. The AUC, sensitivity, specificity, predictive values and likelihood ratios demonstrated that for the HHIE-S-It the criterion total score > 11 is the best cut-off to separate patients with hearing disability from those without. Therefore, the HHIE-S-It was found to be a valid instrument with acceptable psychometric properties that allow its use for both clinical and research purposes for Italian-speaking subjects. In the same way, the diagnostic attributes were found to be acceptable enough to allow its use as a screening tool according to the uses recommended by current guidelines ⁶. In order to properly interpret the results, clinicians need to be aware that the criterion is total score > 11 and that the instrument cannot provide any information about the outcome of hearing aids.

Acknowledgements

The authors thank Franca Laura Artioli, Mariagrazia Totaro, Patrizia Cardini, Paola Allegra and Francesco Ostello, audiometric technicians of the Operative Unit of Otolaryngology, Azienda Ospedaliero-Universitaria di Modena, Modena, Italy, for their contribution in performing audiometric evaluations.

Conflict of interest statement

The authors declare no conflict of interest.

Funding

Internal departmental funding was used to support this study without commercial sponsorship. This research received no other specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Author contributions

EA: investigation, data curation, formal analysis, visualization, writing – original draft; LS: supervision, writing – review and editing; SP: methodology, conceptualization, supervision, writing – review and editing; CC: investigation, data curation, writing – original draft; CG: investigation, data curation; EG: methodology, resources; DM: methodology, project administration, supervision, writing – review and editing; RN: validation, visualization, writing – review and editing.

Ethical consideration

The study was approved by the local Ethics Committee (Protocol AOU 0012599/20 of 6^th May 2020). All procedures performed were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments.

Written informed consent was obtained from each participant/patient for study participation and data publication.

Figures and tables

Figure 1.Distribution of hearing loss in the sample based on PTA according to the ASHA’s degree of hearing loss. Normal hearing: from -15 to 15 dB HL (30 subjects, 18.0%); Slight hearing loss: from 16 to 25 dB HL (67 subjects, 40.1%); Mild hearing loss: from 26 to 40 dB HL (37 subjects, 22.2%); Moderate hearing loss: from 41 to 55 dB HL (26 subjects, 15.6%); Moderately severe hearing loss: from 56 to 70 dB HL (7 subjects, 4.2%); Severe hearing loss: from 71 to 90 dB HL (0 subjects, 0.0%); Profound hearing loss: from 91 dB HL (0 subjects, 0.0%).

Figure 2.Distribution of HHIE-S-It total score according to the degree of hearing loss based on the ASHA classification. Normal hearing: from -15 to 15 dB HL (n = 30); Slight hearing loss: from 16 to 25 dB HL (n = 67); Mild hearing loss: from 26 to 40 dB HL (n = 37); Moderate hearing loss: from 41 to 55 dB HL (n = 26); Moderately severe hearing loss: from 56 to 70 dB HL (n = 7); Severe hearing loss: from 71 to 90 dB HL (n = 0); Profound hearing loss: from 91 dB HL (n = 0). The HHIE-S-It total score was significant higher according to degree of hearing loss. The one-way ANOVA test confirmed the difference between the HHIE-S-It total score, the E score and the S score within the five groups (p < 0.001).

Figure 3.The ROC curve of the HHIE-S-It. The Receiver Operating Characteristics curve shows the diagnostic ability of the HHIE-S-It total score in detecting the hearing loss, in term of true-positive rate (sensitivity) and false-positive rate (1 - specificity). The Youden index J is the point where the ROC curve is at maximum distance from the no discrimination line (chance line). It represents the HHIE-S-It value with the highest test accuracy. The World Health Organization (WHO) definition of disabling hearing loss was considered to elaborate the ROC curve.

Figure 4.Diagnostic proprieties of different versions of the Hearing Handicap Inventory - Screening Version (HHIE-S). For the Italian version (HHIE-S-It) the criterion was total score > 11 whereas for the other version the criterion was total score > 8. Different definitions of hearing loss were used among the studies. °Pure Tone Average > 35 dB in the better ear; °°Hearing threshold at 1000 and 2000 Hz > 40 dB in the better ear or both ears; ^▪Hearing threshold at 500, 1000 and 2000 Hz > 25 dB in the better ear; ^••Pure Tone Average > 40 dB in the better ear; ^•Pure Tone Average > 25 dB in the better ear; ※Pure Tone Average > 20 dB in one or both ears; PPV, positive predictive value; NPV, negative predictive value.

References

1. Wang J, Puel JL. Presbycusis: an update on cochlear mechanisms and therapies. J Clin Med. 2020; 9:218. DOI
2. The GBD 2019 Hearing Loss Collaborators. Hearing loss prevalence and years lived with disability, 1990-2019: findings from the Global Burden of Disease Study 2019. Lancet. 2021; 397:996-1009. DOI
3. Roth TN, Hanebuth D, Probst R. Prevalence of age-related hearing loss in Europe: a review. Eur Arch Otorhinolaryngol. 2011; 268:1101-1107. DOI
4. Sharma RK, Chern A, Golub JS. Age-related hearing loss and the development of cognitive impairment and late-life depression: a scoping overview. Semin Hear. 2021; 42:10-25. DOI
5. The World Health Organization. The World Health Organization: Geneve; 2017.
6. Gurgel RK, Briggs SE, Dhepyasuwan N. Quality improvement in otolaryngology-head and neck surgery: age-related hearing loss measures. Otolaryngol Head Neck Surg. 2021; 165:765-774. DOI
7. Ventry IM, Weinstein BE. Identification of elderly people with hearing problems. ASHA. 1983; 25:37-42.
8. Ventry IM, Weinstein BE. The hearing handicap inventory for the elderly: a new tool. Ear Hear. 1982; 3:128-134. DOI
9. Salonen J, Johansson R, Karjalainen S. Relationship between self-reported hearing and measured hearing impairment in an elderly population in Finland. Int J Audiol. 2011; 50:297-302. DOI
10. Tomioka K, Ikeda H, Hanaie K. The hearing handicap inventory for elderly-screening (HHIE-S) versus a single question: reliability, validity, and relations with quality of life measures in the elderly community, Japan. Qual Life Res. 2013; 22:1151-1159. DOI
11. Öberg M. Validation of the swedish hearing handicap inventory for the elderly (screening version) and evaluation of its effect in hearing aid rehabilitation. Trends Hear. 2016; 20:2331216516639234. DOI
12. Duchêne J, Billiet L, Franco V. Validation of the french version of HHIE-S (hearing handicap inventory for the elderly - screening) questionnaire in french over-60 year-olds. Eur Ann Otorhinolaryngol Head Neck Dis. 2022; 139:198-201. DOI
13. Nunnally JC. McGraw-Hill: New York; 1978.
14. Clark JG. Uses and abuses of hearing loss classification. ASHA. 1981; 23:493-500.
15. The World Health Organization. Deafness and Hearing Loss.Publisher Full Text
16. Ware JE, Gandek BL, Keller SD, IQOLA Project Group. Quality of life and pharmacoeconomics in clinical trials. Lippincott - Raven Publishers: Philadelphia, New York; 1996.
17. Grossi E, Mosconi P, Groth N. Istituto di Ricerche Farmacologiche “Mario Negri”: Milan; 2002.
18. Terwee CB, Bot SD, de Boer MR. Quality criteria were proposed for measurement properties of health status questionnaires. J Clin Epidemiol. 2007; 60:34-42. DOI
19. Lichtenstein MJ, Bess FH, Logan SA. Diagnostic performance of the hearing handicap inventory for the elderly (screening version) against differing definitions of hearing loss. Ear Hear. 1988; 9:208-211. DOI
20. Servidoni AB, Conterno LO. Hearing loss in the elderly: is the hearing handicap inventory for the elderly - screening version effective in diagnosis when compared to the audiometric test?. Int Arch Otorhinolaryngol. 2018; 22:1-8. DOI
21. US Preventive Services Task Force, Krist AH, Davidson KW. Screening for hearing loss in older adults: us preventive services task force recommendation statement. JAMA. 2021; 325:1196-1201. DOI
22. Fluss R, Faraggi D, Reiser B. Estimation of the Youden Index and its associated cutoff point. Biom J. 2005; 47:458-472. DOI
23. Trevethan R. Sensitivity, specificity, and predictive values: foundations, pliabilities, and pitfalls in research and practice. Front Public Health. 2017; 5:307. DOI
24. Šimundić AM. Measures of diagnostic accuracy: basic definitions. EJIFCC. 2009; 19:203-211.
25. Tordrup D, Smith R, Kamenov K. Global return on investment and cost-effectiveness of WHO’s HEAR interventions for hearing loss: a modelling study. Lancet Glob Health. 2022; 10:E52-E62. DOI