|Year : 2022 | Volume
| Issue : 1 | Page : 45-51
Speech recognition, quality of hearing, and data logging statistics over time in adult cochlear implant users
Giulia Elvira Malzanni, Caterina Lerda, Rosa Alessia Battista, Chiara Canova, Omar Gatti, Mario Bussi, Lucia Oriella Piccioni
Department of Otolaryngology, Head and Neck Surgery, San Raffaele Scientific Institute, Vita-Salute San Raffaele University, Milan, Italy
|Date of Submission||17-Mar-2022|
|Date of Acceptance||25-Mar-2022|
|Date of Web Publication||25-Apr-2022|
Dr. Giulia Elvira Malzanni
Department of Otolaryngology, Head and Neck Surgery IRCCS San Raffaele Scientific Institute, Via Olgettina 60, Milan 20132
Source of Support: None, Conflict of Interest: None
Objective: The objective of this study was to examine improvement in speech recognition and “the Speech, Spatial and Qualities of Hearing Scale” (SSQ) scores in monolateral cochlear implant (CI) users and its correlations to data logging statistics. Materials and Methods: A prospective observational, nonrandomized, study was conducted analyzing speech tracking with shielded mouth in noise (SMn), speech audiometry, and SSQ questionnaire (Italian version) at 1, 3, and 6 months postactivation of CI of 31 patients. The overall data of average daily use and acoustic scene analyses were extracted from data logging system. Data of 6-month cumulative speech in quiet-, speech-in-noise-, and noise-listening time of 19 patients were correlated to speech perception scores and SSQ scores. Results: An improvement was observed in all outcomes (speech tracking with SMn, speech audiometry, and SSQ scores). Listeners used their device on for a mean time of 13.45 h per day. Quiet scene was the most represented listening environment. A significant positive correlation was found between noise exposition (in hours) and hearing quality score of SSQ and between exposition to 60–69 dB noise and scores resulted from each part of SSQ. Discussion: The analysis failed to demonstrate a correlation between auditory performances and both exposition to speech scene and total exposition. However, time spent in noise using the device correlated with SSQ scores. Conclusion: Cochlear implantation ensures good results in speech recognition and quality of life, with progressive scores' improvement after activation. A predominant use in quiet was observed together with a correlation between hearing performance and exposition to noise.
Keywords: Cochlear implant, data logging, listening environment, speech audiometry, speech recognition, speech tracking, SSQ questionnaire
|How to cite this article:|
Malzanni GE, Lerda C, Battista RA, Canova C, Gatti O, Bussi M, Piccioni LO. Speech recognition, quality of hearing, and data logging statistics over time in adult cochlear implant users. Indian J Otol 2022;28:45-51
|How to cite this URL:|
Malzanni GE, Lerda C, Battista RA, Canova C, Gatti O, Bussi M, Piccioni LO. Speech recognition, quality of hearing, and data logging statistics over time in adult cochlear implant users. Indian J Otol [serial online] 2022 [cited 2022 May 27];28:45-51. Available from: https://www.indianjotol.org/text.asp?2022/28/1/45/343765
| Introduction|| |
Cochlear implant (CI) is the first example of a neural prosthesis that can substitute a sensory organ: bypass the malfunctioning auditory periphery of profoundly deaf people to electrically stimulate their auditory nerve.
Different factors can affect the overall hearing performance capabilities in CI recipients such as age, duration of deafness, preverbal or postverbal onset of deafness, etiology, cognitive abilities, and related disabilities. On this topic, a retrospective multicenter study from 2012 concluded that durations of severe-to-profound hearing loss of more than 40 years negatively influenced performance, together with older age at implantation and some unfavorable etiologies of HL (meningitis, temporal bone fractures, acoustic neuroma, and auditory neuropathy spectrum disorders). However, nowadays it has been demonstrated that even elderly patients, subject who has long-standing deafness, can reach good hearing performances. Conversely, etiology seems to have smaller effects, though bacterial labyrinthitis resulted in poorer outcomes in a previous study.
Furthermore, the exposition to a rich auditory environment, plays a role in obtaining better performances. CI users need to be massively exposed to complex listening situations in order to improve listening and speech performance. For this reason, clinicians recommend full-time use of CI to patients, defined as continuative use ranged from 8 to 10 h per day.
Nowadays, thanks to data logging, it is possible to objectify CI use recording through the CI processor the exposure time to a particular listening condition during patient's daily life. Both Busch et al. and Cristofari et al. in 2017 used data logging to demonstrate that older CI users (>19 years old) spend more time in “quiet” and “noise” while preschoolers and school-aged children (age 0–18 years) usually spend more time in “speech in noise.” Busch et al. found that children (<6–17 years old) spend more time in “music” and “noise” while adults (age 18–64 years) spend more time in “quiet,” hypothesizing that the difference in exposure to complex listening situations may be related to different auditory outcomes.
In adult CI users, the relationship between postoperative speech recognition, linguistic abilities, and device use has not much been investigated. Only recently, analyzing data logging statistics in a cohort of CI recipients 1 year after implantation, Schvartz-Leyzac et al. found a weak but positive correlation between speech in quiet and postoperative sentence recognition. Moreover, a stronger correlation was observed between number of hours using the device per day and postoperative speech recognition performance.
The aims of the present study are to quantify substantial performance improvements in terms of speech, language, and hearing threshold and to relate data logging statistics with the speech therapy and audiometric assessments. Moreover, the authors want to demonstrate the effective improvement of the patient's quality of life, thanks to a repeated administration of the Speech, Spatial and Qualities of Hearing Scale (SSQ) questionnaire.
| Materials and Methods|| |
A prospective observational, nonrandomized, study was conducted. All patients signed an informed consent form. The study adhered to the tenets of the Declaration of Helsinki. The data were anonymized or maintained with confidentiality. Data logging statistics and extraction techniques include the environmental scene classifier within the Cochlear CP900, CP1000 and Kanso processors (Cochlear LTD., Sydney, Australia) that identify six different sound environments: speech in quiet, speech in noise, noise, music, quiet, and wind.
Forty subjects were enrolled in the study, but 31 patients had complete data for analysis.
The inclusion criteria were as follows: (1) 18 years of age or older at the time of activation, (2) first CI ear, (3) at least 6 months of CI use, (4) postverbal hearing loss with adequate language development, and (5) Italian as native language. Exclusion criteria included: (1) a documented device failure that occurred at any time during the data collection time period; (2) use of combined electro-acoustic stimulation strategy; (3) cognitive impairment, mental delay, or severe comorbidities; (4) use of sign language as mean of communication; and (5) not respect the exact timing of the controls at 1, 3, and 6 months.
The etiology of deafness was unknown in 13 cases (42%), 6 patients (19.3%) reported a sudden hearing loss, 4 patients (12.9%) were affected by otosclerosis, 4 (12.9%) had previous surgery for chronic otitis media, while 4 other subjects (12.9%) reported other causes of hearing loss (Cogan syndrome, petrous bone fracture, measles, and genetic mutation). The study sample included 14 males (45.2%) and 17 females (54.8%) between the ages of 25 and 77 years with a mean age of 58 (±15) years at the time of data retrieval. The mean age at surgery was 57 (±15) and the mean duration of deafness (age at implant – age at onset of deafness) was 17.4 years (±13.4). Eighteen patients (58%) had a bimodal stimulation modality (hearing aids [HAs] + CI) while 13 (42%) had a unilateral CI. All patients underwent monolateral cochlear implantation by a single surgeon (senior author). One patient received Nucleus® CP900 sound processor (3.2%), 27 received Nucleus® CP1000 sound processor (87.1%), and 3 received Nucleus® Kanso™ sound processor (9.7%) (Cochlear LTD, Sydney, Australia). The characteristics of the sample are summarized in [Table 1].
Speech therapy evaluation
The speech therapy evaluation was conducted at the preoperative stage and at 1, 3, and 6 months after activation by administering the speech tracking test, as it is described in the Test Abilità Uditive Varese document. This test is divided into four parts: lip reading in quiet, lip reading in noise, shielded mouth in quiet, and shielded mouth in noise (SMn). The competition noise is placed at a distance of 1 m behind the patient in a central way, at an intensity of +60 dB. The therapist instead places himself at a distance of 1 m in front of the patient. The score is calculated based on the number of correct words repeated in 1 min; if the patient does not recognize a word, the therapist continues with the next ones, without stopping. Subsequently, the score is referred to the normative data present in the literature, thus placing the patient in a sufficient or insufficient performance to guarantee a comfortable speech intelligibility. In the present study, the condition that was taken into consideration was the SMn that is the most difficult condition during speech tracking test. A sufficient score in adult patients, relatively to this part, is 60 (correct words repeated in a minute).
During each evaluation, the SSQ self-assessment questionnaire was submitted to the patient. This questionnaire appeared to be the best tool to assess everyday hearing functioning. It investigates the subjective experience of the patient and interrogates about many difficult listening situations, paying special attention to spatial hearing (SH) ability, listening to music, and listening effort.
SSQ contains 49 questions, divided into 3 parts (“speech” with 14 questions, “spatial” with 17, and “hearing qualities” with 18). The questionnaire investigates a range of hearing disabilities across several domains. The score is calculated based on every single question, to which the patient can answer expressing an opinion based on a Likert scale from 1 to 10. Higher scores indicate better skills. The overall SSQ score is calculated by dividing the total points by the number of questions. An average, comparable to subsequent checks, will then be made for each individual part of the questionnaire: speech perception (SP), SH, and hearing qualities (HQs).
The audiometric evaluation was conducted at the preoperative stage and at 1, 3, and 6 months following the implant activation. The preoperative assessment includes a comprehensive audiometric evaluation with air and bone conduction thresholds between 125 and 8000 Hz (tonal audiometry), together with speech recognition scores (speech audiometry). At each follow-up, the audiometric tests refer only to the use of the single CI in open field in a soundproof booth, excluding the contralateral ear removing HA and masking with competitive noise. In order to perform an accurate investigation among the patients, the tonal audiometry results were recorded as pure-tone average, calculated considering thresholds at 500, 1000, 2000, and 4000 Hz. Speech audiometry was conducted using simple two-syllable Italian words, and the results were recorded as percentage of word recognized at different intensities. The most significant was the percentage of words recognized at 60 dB SPL (the closest to everyday speech).
The data logging system is a part of the Cochlear CP900, CP1000 and Kanso sound processors that scan the acoustic environment by analyzing the frequency spectrum of microphone input signals. The patient data were obtained from Custom Sound software (Cochlear, Sydney, Australia). It classifies the sound environments into one of six scenes (speech in quiet, speech in noise, music, quiet, noise, and wind). In addition, the data logger registers the time intervals over which the signal level in the speech processor lies in one of the following six level bands: 40 dB (1), 40–49 dB SPL (2), 50–59 dB SPL (3), 60–69 dB SPL (4), 70–79 dB SPL (5), and 80 dB SPL (6). The daily exposure time to different acoustic scenes was extracted at every mapping session; 6 months' use data were then cumulated. Nineteen (61.3%) of the 31 patients had complete data for data logging analysis.
The Kolmogorov–Smirnov normality test was performed in the sample included in study regarding the variable examined. It gave a positive outcome, suggesting that our variables follow a normal distribution in our population. Thus, a repeated-measures analysis of variance was performed to compare SMn (words repeated per minute), speech audiometry (% of words repeated at 60 dB SPL), and SSQ scores (divided into SP, SH, and HQ parts) in the different period of follow-up (1, 3, and 6 months). The correlation between data logging scores (loudness subcategories of speech in noise, speech in quiet, and noise), speech tracking scores in particular SMn, and SSQ scores was tested by using the Pearson's correlation test. Significance level was set at P = 0.05. All analyses were performed using the MedCalc Statistical Software version 16.8 (MedCalc Software, Ostend, Belgium; https://www.medcalc.org; 2016).
| Results|| |
Changes in SMn score of speech tracking test, speech audiometry scores (% of word recognition at 60 dB), and SSQ scores (divided into SP, SH, and HQ parts) over time (at 1, 3, and 6 months) are shown in [Table 2], with average score, standard error (SE), and 95% confidence interval (95% CI).
|Table 2: Changes in shielded mouth in noise (words repeated/min), speech audiometry, and speech, and spatial quality scores over time|
Click here to view
Score in SMn test of speech tracking improved from a mean of 19.94 words repeated per minute (SE: 2.343) at 1 month postimplantation to 28.77 (SE: 1.9362) at 3 months and continued to improve to 37.5 (SE: 1.4825) at 6 months. The pairwise comparisons between scores at 1, 3, and 6 months postimplantation showed a statistically significant improvement (P < 0.0001 and P = 0.0003) in all comparisons [Figure 1].
|Figure 1: Shielded mouth in noise results with pairwise comparisons over time. WpM: number of words per minute. *for P < 0.05|
Click here to view
Furthermore, 60-dB speech audiometry's results improved from a mean of 0.60 (SE: 0.0629) at 1 month, 0.78 (SE: 0.0653) at 3 months, to 0.89 (SE: 0.0474) at 6 months postimplantation, showing a global statistically significant improvement (P = 0.0006 and P = 0.0013), except the one between 3 and 6 months postimplantation (P = 0.1379), as shown in [Figure 2].
|Figure 2: Speech audiometry with pairwise comparisons over time. *For P < 0.05|
Click here to view
The mean score of the SSQ's “speech perception” part showed a progressive improvement, resulting 3.3394 (SE: 0.4261) at 1 month, 4.9044 (SE: 0.4187) at 3 months, and 6.5328 (SE: 0.3119) at 6 months. A similar score's increase was observed in the “spatial hearing” part: from a mean score of 2.7433 (SE: 0.4346) at 1 month to 3.9889 (SE: 0.4592) at 3 months and 5.2383 (SE: 0.4059) at 6 months. Accordingly, in the “hearing quality” part, the mean score was 3.7772 (SE: 0.4602) at 1 month, 4.5800 (SE: 0.4059) at 3 months, and 6.1650 (SE: 0.3144) at 6 months. The pairwise comparison between scores of SP, ST, and HQ parts at 1, 3, and 6 months found a statistically significant improvement (P < 0.0001, P = 0.0004, and P = 0.0002), except between the mean score of HQs at 1 and the one at 3 months (P = 0.2377). Results with pairwise comparisons are summarized in [Figure 3].
|Figure 3: SSQ (speech perception, spatial hearing, and hearing qualities) with pairwise comparisons over time. *For P < 0.05|
Click here to view
The data logging analysis allowed to calculate the device's daily use and to classify the sound environments into one of six scenes. Listeners used their device for 12.03 h a day on average during the 1st month postactivation. The mean use improved at 3 months postactivation (13.20 h a day) and continued to improve at 6 months (13.45 h a day). [Table 3] shows data regarding the average number of hours for each type of sound environment as well as the corresponding percentage of time spent in each environment, based on the average number of hours the device was used per day.
Quiet scene was the most represented listening environment (48% of daily use at 6 months), followed by speech (22.9% in noise and 11.1% in quiet), noise (13.7%), and music (4.3%). Wind scene had a very low prevalence (near to 0%). Time spent in each environment did not change much over time.
The correlation between data logging scores (subcategories of speech in noise, speech in quiet, and noise at different loudness levels), speech tracking scores (SMn), speech audiometry, and SSQ scores was tested with Pearson's correlation test, in order to explore potential relationships between postoperative performance and data logging information. We hypothesized that subjects who listened more (speech in noise, speech in quiet, and noise) would have better speech recognition scores. According to our assumption, results showed a significant positive correlation between exposition to noise (in hours) and HQ score [correlation coefficient: 0.491, P = 0.033, [Figure 4]]. Considering subcategories of loudness, a significant positive correlation was found between exposition to noise between 60 and 69 dB and scores of each part of SSQ [Figure 5]. The correlation coefficient between noise 60–69 dB and SP score was 0.59 (P = 0.0079), the one between noise 60–69 dB and SH score was 0.492 (P = 0.03239), and the one between noise 60–69 dB and HQ score was 0.559 (P = 0.0129). In addition, subjects with a higher speech exposition tended to have higher SMn scores, without reaching statistical significance [Figure 6]. On the contrary, speech audiometry results were not related to data logging information (speech or noise exposition).
|Figure 4: Scatterplot showing correlation between exposition to noise and SSQ hearing quality score (correlation coefficient = 0.491, P = 0.033*)|
Click here to view
|Figure 5: Scatterplots showing correlation between exposition to noise at 60 to 69 dB and SSQ scores: (a) Speech perception (correlation coefficient = 0.59, P = 0.0079*), (b) spatial hearing (correlation coefficient = 0.492, P = 0.03239*), and (c) hearing qualities (correlation coefficient = 0.559, P = 0.0129*)|
Click here to view
|Figure 6: Scatterplot showing correlation between exposition to speech and shielded mouth in noise scores (correlation coefficient = 0.189, P = 0.4381)|
Click here to view
| Discussion|| |
The purpose of this study was to analyze performance and quality of life improvement after 1, 3, and 6 months of cochlear implantation and potential effects of the listening environment on speech and hearing performances. Specifically, it was compared in terms of how the patient's performance changed during these sessions.
Performance improvement includes speech tracking scores (SMn), speech audiometry, and SSQ scores. In our opinion, speech tracking technique (ST) is quite dated, since it was originally developed by De Filippo and Scott in 1978 for the measurement of HA performance and has then been introduced in the evaluation of CI recipients, before and after implantation. For this reason, in our study, we took in consideration the listening situation of SMn, which is the most difficult hearing condition. SMn scores (words repeated per minute) showed a significant improvement from 1 to 3 and to 6 months. These results are consistent with previous studies that analyzed speech tracking performance improvement over time in a sample of CI recipients., However, in the present work, a more complex listening condition was used (listening with the mouth shielded in the noise) than those used in older works, which considered only lip reading alone and lip reading with the aid of the CI. As a consequence, our analysis is more significant as it demonstrates an improvement in a more difficult listening condition.
Audiometric evaluation showed an improvement in speech audiometry scores, which was significant comparing the results at 1 month with those obtained at 3 months and at 6 months. These findings are consistent with other studies' results.,
No significant differences were obtained only by comparing the results of speech audiometry at 3 months with those obtained at 6 months. This can be explained by the plateau effect obtained in speech recognition after about 3 months, leading to a harder identification of any improvement. In addition, speech audiometry is a too simplistic evaluation and does not allow to evaluate the different listening conditions of everyday life. On the contrary, speech tracking is more sensitive in recognizing improvements in auditory and linguistic performance as the listening conditions used are more varied and complex. They include hearing with or without lip reading, hearing in quiet or in noise, and hearing with different reading speed, according to the patients' needs. We believe that our data make it possible to justify improvement, both at specific listening level (simple words) and at a complex level (text or conversation with several interlocutors).
Thanks to the repetitive administration of SSQ questionnaire, it has been possible to investigate improvements in hearing abilities and quality of hearing. A statistically significant improvement was observed in the score, relative to the part of SP, obtained at 1, 3, and 6 months. This suggests gain in the ability to hear speech in various real-life situations: hearing in quiet, hearing in noise, hearing in multiple speech stream processing, hearing at the phone. A comparable improvement was obtained in the comparison between the results at 1, 3, and 6 months of the part relating to SH, suggesting improvement in localizing sounds' source and ability to discriminate distance and movement. In the HQ part, a significant improvement was observed between the scores obtained at 3 months and those at 6 months, as well as between 1 month and 6 months' scores. On the other hand, there was no significant improvement between the score obtained at 1 month and at 3 months. As displayed in [Table 2] and [Figure 3], the greatest improvements in the SSQ scores were obtained between 3 and 6 months (especially in the HQ part). This result can be justified by a greater psychophysical device's tolerance established over time, together with a greater awareness of the results. We have to consider that this part of the questionnaire (HQs) investigates more difficult skills, such as abilities to perceive sound quality and naturalness, identification of sound, and segregation of sounds which are obtained after longer times of device use. It is also important to remember that several factors, which were not examined in the present study, can contribute significantly to postoperative speech recognition scores, including duration of deafness, etiology of deafness, and scalar location., These variables could explain the differences in postoperative performance noted in the patients.
Data logging information of the study sample shows a full-time use of the speech processor (>13 h/day); the mean use appears higher than the one reported by other studies conducted in adult population, and tends to increase over time. The increase in usage time the mean age of the sample (58 years old), is consistent with other studies' results, which show higher expositions to quiet scene in adult population, but is also in line with the particular historical moment in which the study was conducted, the COVID-19 pandemic, that forced many people to limit their social life and to work from home. On the other hand, this result demonstrates how difficult it is for adults to be exposed to complex and stimulating auditory contexts.
The data analysis failed to demonstrate a correlation between auditory performances and exposition to speech scene or total exposition, unlike what has been observed in other studies., Only a tendency to get higher scores on SMn in subjects with higher temporal exposure to speech was detected. However, this result has to be confirmed by further studies with larger samples.
On the contrary, a significant positive correlation was found between exposure time to noise scene at a loudness level between 60 and 69 dB and SSQ scores. In fact, the exposure to noise, in particular at a modest intensity, constitutes a form of stimulation of the auditory path that could contribute to a sound enrichment and therefore to an improvement in SP, spatial perception, and HQ. However, further research is needed to better understand the nature of this relationship, and whether it is simply causal or correlational in nature.
| Conclusion|| |
Cochlear implantation ensures good results regarding speech recognition and quality of life, with scores progressively improving after activation. Speech tracking scores appear to be the most sensitive in recording patients' speech improvement, while SSQ questionnaire appears adequate in investigating quality of hearing. About data logging, a predominant use in quiet was observed together with a correlation between hearing performance and exposition to noise. These results have crucial implications for the counseling and rehabilitation of adult CI recipients.
Not required. The current study is a purely observational study (with no element of intervention).
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Blamey P, Artieres F, Başkent D, Bergeron F, Beynon A, Burke E, et al.
Factors affecting auditory performance of postlinguistically deaf adults using cochlear implants: An update with 2251 patients. Audiol Neurootol 2013;18:36-47.
Buchman CA, Gifford RH, Haynes DS, Lenarz T, O'Donoghue G, Adunka O, et al.
Unilateral cochlear implants for severe, profound, or moderate sloping to profound bilateral sensorineural hearing loss: A systematic review and consensus statements. JAMA Otolaryngol Head Neck Surg 2020;146:942-53.
Blamey P, Arndt P, Bergeron F, Bredberg G, Brimacombe J, Facer G, et al.
Factors affecting auditory performance of postlinguistically deaf adults using cochlear implants. Audiol Neurootol 1996;1:293-306.
Busch T, Vanpoucke F, van Wieringen A. Auditory environment across the life span of cochlear implant users: Insights from data logging. J Speech Lang Hear Res 2017;60:1362-77.
Cristofari E, Cuda D, Martini A, Forli F, Zanetti D, Di Lisi D, et al.
A multicenter clinical evaluation of data logging in cochlear implant recipients using automated scene classification technologies. Audiol Neurootol 2017;22:226-35.
Schvartz-Leyzac KC, Conrad CA, Zwolan TA. Datalogging statistics and speech recognition during the first year of use in adult cochlear implant recipients. Otol Neurotol 2019;40:e686-93.
Gatehouse S, Noble W. The Speech, Spatial and Qualities of Hearing Scale (SSQ). Int J Audiol 2004;43:85-99.
Burdo S, Poggia C, Giuria P. T.A.U.V. (Test Abilità Uditive Varese). I Care 1995:2;3-36.
De Filippo CL, Scott BL. A method for training and evaluating the reception of ongoing speech. J Acoust Soc Am 1978;63:1186-92.
Jenkins H, Chmiel R, Jerger J. Speech tracking in the evaluation of a multichannel cochlear prosthesis. Laryngoscope 1989;99:245-51.
Robbins AM, Osberger MJ, Miyamoto RT, Kienle ML, Myres WA. Speech-tracking performance in single-channel cochlear implant subjects. J Speech Hear Res 1985;28:565-78.
Favaretto N, Marioni G, Brotto D, Sorrentino F, Gheller F, Castiglione A, et al.
Cochlear implant outcomes in the elderly: A uni- and multivariate analyses of prognostic factors. Eur Arch Otorhinolaryngol 2019;276:3089-94.
Sladen DP, Carlson ML, Dowling BP, Olund AP, DeJong MD, Breneman A, et al.
Cochlear implantation in adults with asymmetric hearing loss: Speech recognition in quiet and in noise, and health related quality of life. Otol Neurotol 2018;39:576-81.
Holden LK, Finley CC, Firszt JB, Holden TA, Brenner C, Potts LG, et al.
Factors affecting open-set word recognition in adults with cochlear implants. Ear Hear 2013;34:342-60.
Finley CC, Holden TA, Holden LK, Whiting BR, Chole RA, Neely GJ, et al.
Role of electrode placement as a contributor to variability in cochlear implant outcomes. Otol Neurotol 2008;29:920-8.
Battista RA, Ferraro M, Piccioni LO, Malzanni GE, Bussi M. Personal protective equipment (PPE) in COVID 19 pandemic: Related symptoms and adverse reactions in healthcare workers and general population. J Occup Environ Med 2021;63:e80-5.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6]
[Table 1], [Table 2], [Table 3]