Public online information about tinnitus: A cross-sectional study of YouTube videos.

PURPOSE: To examine the information about tinnitus contained in different video sources on YouTube.
MATERIALS AND METHODS: The 100 most widely viewed tinnitus videos were manually coded. Firstly, we identified the sources of upload: consumer, professional, television-based clip, and internet-based clip. Secondly, the videos were analyzed to ascertain what pertinent information they contained from a current National Institute on Deafness and Other Communication Disorders fact sheet.
RESULTS: Of the videos, 42 were consumer-generated, 33 from media, and 25 from professionals. Collectively, the 100 videos were viewed almost 9 million times. The odds of mentioning "objective tinnitus" in professional videos were 9.58 times those from media sources [odds ratio (OR) = 9.58; 95% confidence interval (CI): 1.94, 47.42; P = 0.01], whereas these odds in consumer videos were 51% of media-generated videos (OR = 0.51; 95% CI: 0.20, 1.29; P = 0.16). The odds that the purpose of a video was to sell a product or service were nearly the same for both consumer and professional videos. Consumer videos were found to be 4.33 times as likely to carry a theme about an individual's own experience with tinnitus (OR = 4.33; 95% CI: 1.62, 11.63; P = 0.004) as media videos.
CONCLUSIONS: Of the top 100 viewed videos on tinnitus, most were uploaded by consumers, sharing individuals' experiences. Actions are needed to make scientific medical information more prominently available and accessible on YouTube and other social media.

Introduction

Tinnitus is the auditory perception of a ringing sound without an actual acoustic stimulus. Some people describe the sound they hear as a ringing, others as buzzing, roaring, hissing, or clicking. The sound can vary in frequency from high to low, and can be present not only continuously or intermittently, but also unilaterally or bilaterally. About one in five people in USA suffer from tinnitus and they report a deterioration of quality of life, psychological problems, as well as physical ones, ranging from pain to insomnia.[1] In a cross-sectional study on the prevalence of tinnitus in the USA, Bhatt et al.[2] estimated it to be present in about 9.6% of adults in the past year, with 7% of the sufferers considering it to be a big problem, 42% a small problem, 27% having had tinnitus for over 15 years, and 36% having symptoms constantly. Just fewer than 50% had reported it to their physician, and the treatment most proposed (45%) had been medication, while hearing aids were a distant second (9%), and cognitive behavioral therapy was rarely mentioned (0.2%).[1]

Tinnitus is quite prevalent among older adults (ages 60 and over), especially those who have experienced ear or hearing problems.[3] For example, a recent retrospective file review of 2758 patients at a clinic during a contiguous 12-month period noted a 16.8% prevalence in the patients, seen in 39% of the female patients, and with 97.5% of the patients also presenting with hearing loss.[4] It is possible that tinnitus reflects an inner-ear microcirculatory dysfunction in the elderly, especially in those with impaired control of blood pressure and left ventricular dysfunction. As such, the presence of tinnitus may be clinically important as an early diagnostic indicator for patients in need of intensive management of mild-to-moderate chronic heart failure.[5] Clearly, tinnitus is associated with some of the health problems experienced by older adults such as hypertension and vestibular disorders,[3] and contributes to a diminution of quality of life.

Tinnitus may be one of the symptoms of a different condition, such as presbyacousis (age-related hearing loss), exposure to loud noise (long or short term), accumulation of earwax, otosclerosis (hardening of the middle ear ossicles), circulatory system disorders (from hypertension to arteriovenous malformation), may be caused by injury to the ear, other chronic health problems, or medications.[1]

A number of treatments have been proposed for tinnitus, with little if any, success, and fall in three general categories: medications, noise manipulation, and alternative treatments. Noise manipulation involves not only masking devices that suppress the ringing sound, but also hearing aids that correct for accompanying hearing loss, yet an individualized tinnitus retraining with counseling has also shown some acceptable results for certain patients. Antidepressant and antianxiety medications have shown some promise with severe cases.[3]

During the period 2011–2015, 34 patents were studied in a review of recent therapies for inner ear diseases, including tinnitus.[6] The conclusion was that better insight into the physiology at the molecular level, and considering the challenges involved, is directing efforts toward new directions. For example, the direct stimulation of the cochleovestibular nerve resulted in successful outcome in 60% of the patients, but unfortunately also in hearing damage in 86% of them, making it an untenable treatment.[7] Currently, no treatment exists for tinnitus that is received positively by all patients. Certainly, most patients adopt one or several changes in their lifestyle: (a) reductions in their use of alcohol, caffeine, and nicotine, (b) refraining from loud noises, (c) masking the noise with music or other ambient sounds, and (d) managing stress, often through counseling, education, or support groups.

Prevention may be a crucial component in this health matter, because the underlying disorder that results in tinnitus often may be preventable. Bhatt et al. found that prolonged noise exposure correlated with occurrence of tinnitus and therefore could be targeted for prevention.[2] Hearing protection from loud noises and loud music, as well as maintenance of cardiovascular health may help prevent the appearance of tinnitus.[3]

The internet has been the main resource for immediate access to health-related data.[8] In particular, social media have been used by researchers and practitioners in healthcare and public health settings to facilitate both disease management among patients suffering from chronic illnesses and health promotion among individuals at risk of acquiring chronic illnesses.[910] YouTube is a popular social media platform and is ranked the second most popular website globally.[11] A systematic review published in 2015 found that YouTube has been increasingly used as a platform for disseminating health information.[12] However, health-related misinformation is also prevalent on YouTube, while, in general, videos from governmental and professional organizations can provide high-quality, trustworthy information to the public.[12] Nevertheless, our recent studies found that among the most popular 100 YouTube videos pertinent to a variety of health topics, those uploaded by healthcare and public health professionals are not as popular as videos uploaded by consumers or media organizations.[13141516]

Given that the above general facts about tinnitus are accessible to health professionals and the public via other sources, the objective of this study was to examine the information about tinnitus contained in the 100 most widely-viewed English language videos on YouTube. Based on prior research,[13141516] we hypothesized that there would be a paucity of content uploaded by professional sources, and that videos would contain mostly anecdotal accounts uploaded by the general public (consumers).

Materials and Methods

The term “tinnitus” was used to search for the 100 most widely viewed videos on YouTube. This was determined by sorting for the number of views. All videos included in this sample were in English. A total of 23 non-English videos were excluded in the process of coming up with the 100 for our sample. In addition, 60 videos were excluded for being solely a sound that could mask tinnitus with no other content or information. Once the sample was comprised, basic descriptive data were recorded and included the title, uniform resource locator (URL), date of upload, length of the video, number of views of the video, as well as the number of thumbs up and thumbs down. The next step in the process was to code the content of each of the videos. Categories of information in this category were largely influenced by a fact sheet published by The National Institute on Deafness and Other Communication Disorders.[3] This entailed first identifying the sources of upload, which were grouped into the following categories: consumer (member of the lay public); professional (a credentialed person, qualified to discuss the topic); television-based clip (any clip that originated from television); and internet-based clip (any clip that originated from an internet channel or website).

Professional analyses of the symptoms of tinnitus and patients’ concerns regarding accurate diagnosis and potential treatments for tinnitus all would include information regarding the parameters outlined below.[3] Therefore, the second step involved the analysis of the videos to ascertain whether they contained the pertinent information on tinnitus, as listed below. People who search for information on tinnitus do so to learn about its nature and symptoms, its causes, its treatment, possible ways to prevent it, problems that may co-exist with it, and possibly how it has affected other fellow citizenry. If videos contained the below information, they would be considered valuable, informative, and useful to the public seeking information regarding this matter.

Mentions of the following content/information regarding tinnitus were noted:

Regarding the nature and symptoms of tinnitus: subjective tinnitus; objective tinnitus; pulsatile tinnitus; pain.

Regarding the causes of tinnitus: loud noise; acoustic neuroma (auditory nerve/vestibular schwannoma/brain tumor); hearing loss; Ménière’s Disease; diseases of heart/blood vessels; diseases of thyroid vessels; drugs/medications; temporomandibular joint disorders.

Regarding the treatment of tinnitus, mentions were noted of using: a hearing aid; counseling; sound generators; a cochlear implant; antianxiety/antidepressant medication; acoustic neural stimulation; nutritional supplements.

Mentions of prevention.

Regarding associated/co-morbid symptoms: insomnia; difficulty coping.

Regarding the purpose of the video: if this video is about a personal experience; whether or not the purpose of the video is to sell a product or service and if so, whether the product or service assures the viewer that it cures tinnitus.

Lastly, we noted whether the video was featuring a celebrity (as this often raises public awareness of a topic).

A random sample of 10 videos were selected and recoded by a second researcher to determine the inter-rater reliability of the manual coding. The level of agreement was established using the percentage of agreement and Cohen’s unweighted Kappa. Among the all 27 manually coded variables noted above, only the source variable showed 80% agreement with Kappa value equaling 0.64, while the rest of variables achieved 100% agreement.

Statistical analysis was conducted using the software SAS 9.4 (SAS Institute, Cary, North Carolina). The length of videos, number of views, number of views per day, thumbs up, and thumbs down. These distributions were skewed; hence, non-parametric methods were used for their analyses. The Kruskal–Wallis H test, a non-parametric alternative of analysis of variance (ANOVA) test, based on ranks was performed across the five source categories. Only one (length of video) test suggested that there is a significant difference; hence, a pairwise analysis was performed using Bonferroni Post Hoc test. The correlations among the length of videos, number of views, number of views of per day, thumbs up and thumbs down analysis was performed via Spearman’s rho (a non-parametric alternative of Pearson correlation based on ranks). All the manually coded themes of the videos were converted into multiple binary variables and each binary variable representing one theme (“yes” or “no”). Univariate logistic regression analysis was performed; the source of the video was the predictor variable, while the manually coded content variables are the binary outcome variables. Some source categories have very limited counts, and this will affect the odds ratio (OR) calculation. In the light of this, two source categories were merged. The two categories that were merged were television (n = 6) and internet (n = 27) into one category named “media” (n = 33). The OR of a specific category of video source (professional and consumer) showing a specific type of content was calculated using “media” as reference.

Ethics approval

The Institutional Review Board of William Paterson University does not review the studies, in which human subjects are not involved.

Results

Of the 100 manually coded videos, 42 were consumer-created, 33 were media videos, and 25 were professional videos. The collective number of views of the videos was almost 9 million. Table 1 presents the descriptive statistics of the continuous variables including video lengths (duration), number of views, number of views of per day, thumbs-up, and thumbs-down of these 100 videos. Consumer videos attracted 3.17 million views (35.4% of the total views), followed by professional videos with 3.15 million views (35.18%), while media had 2.63 million views (29.42%). Video lengths ranged from 0.28 to 29.63 min.

Table 1

Descriptive statistics of 100 most viewed tinnitus-related YouTube videos in English by their sources

Source of upload	n	Video length (min)					Number of views
		Mean [SE]	Median	Range	95% CI	Total (%)	Mean [SE]	Median	Range	95% CI	Total (%)
Consumer	42	5.35 [0.76]	4.08	0.28–20.85	3.82–6.87	224.49 (48.39)	75,557.33 [10,000.26]	46,147.50	14,196.00–240,608.00	55,361.39–95,753.27	3,173,408.00 (35.40)
Media	33	3.20 [0.58]	2.08	0.52–17.10	2.03–4.37	105.47 (22.73)	79,911.61 [17,020.57]	54,624.00	24,613.00–468,277.00	45,241.84–114,581.37	2,637,083.00 (29.42)
Professional	25	5.36 [1.18]	3.92	0.44–29.63	2.92–7.80	133.98 (28.88)	126,138.88 [29,989.01]	67,843.00	18,253.00–550,154.00	64,244.61–188,033.15	3,153,472.00 (35.18)
Total	100					463.94					8,963,963.00
Source of upload	n	Thumbs-up					Thumbs-down
		Mean [SE]	Median	Range	95% CI	Total (%)	Mean [SE]	Median	Range	95% CI	Total (%)
Consumer	42	374.86 [105.74]	178.00	0–4256.00	161.31–588.41	15,744.00 (38.93)	31.33 [5.95]	15.50	0–155.00	19.31–43.35	1316.00 (43.75)
Media	33	476.52 [345.21]	83.00	1.00–11,488.00	-226.65–1179.68	15,725.00 (38.88)	26.40 [7.01]	6.00	0–168.00	12.11–40.68	871.00 (28.96)
Professional	25	359.12 [99.45]	198.00	0–2247.00	153.86–564.38	8978.00 (22.19)	32.84 [7.36]	24.00	0–144.00	17.65–48.03	821.00 (27.29)
Total	100					40,447					3008
Source of upload		n	Views per day
			Mean [SE]		Median		Range		95% CI		Total (%)
Consumer		42	60.02 [10.09]		39.00		10.00–377.00		39.65–80.40		2521.00 (24.75)
Media		33	67.03 [18.77]		34.00		10.00–591.00		28.81–105.26		2212.00 (21.72)
Professional		25	218.08 [124.19]		40.00		10.00–3145.00		-38.24–474.40		5452.00 (53.53)
Total		100									10,185

CI = confidence interval, SE = standard error.

Table 2 shows the results of comparison of these variables between the three source groups. The distributions of length of videos (min) were the only one that was statistically significant (Kruskal–Wallis χ 2 = 6.67, P = 0.04). Pairwise comparisons using Bonferroni Post Hoc test found that there was no statistically significant difference in the video length between the sources. The inconsistent results between the overall test and the pairwise comparisons are due to two reasons: (1) Bonferroni test is believed to be conservative thus less powerful; and (2) the overall signal is not strong enough to claim a large difference with a P-value of 0.04.

Table 2

Results from Kruskal–Wallis H test of comparing views per day video length (min), number of views, thumbs-up and thumbs-down of top 100 viewed tinnitus-related YouTube videos in English across three different source categories

Meta-data	χ 2 value	P-value
Views per day	0.71	0.70
Video length (min)	6.67	0.04*
Number of views	1.76	0.41
Thumbs-up	6.17	0.05
Thumbs-down	3.36	0.19

* <0.05.

The total number of thumbs-ups garnered by all the videos regardless of source is 40,447 [Table 1]. Consumer videos accounted for 15,744 (38.93%) thumbs-up and 1316 (43.75%) thumbs-down. There was no difference between the categories for both thumbs-up (Kruskal–Wallis χ 2 = 6.17, P = 0.05) and thumbs-down (Kruskal–Wallis χ 2 = 3.36, P = 0.19) as shown on Table 2. Interestingly, the videos from media sources have nine fewer thumbs up compared to videos uploaded by consumers, despite the fact that videos uploaded by consumers having close to 1 million more views than videos from media sources. Table 3 gives the results of Spearman correlation. Evidently, the more times a video was viewed, the more thumbs-up the video attracted (Spearman’s rho = 0.38, P < 0.0001) and also more thumbs-down (Spearman’s rho = 0.32, P = 0.001). As a result, thumbs-up and thumbs-down are positively correlated (Spearman’s rho = 0.42, P < 0.0001). Interestingly, a longer duration of a video is slightly correlated with more thumbs-up ratings (Spearman’s rho = 0.18, P = 0.08) but not with more thumbs-down ratings (Spearman’s rho = 0.10, P = 0.35).

Table 3

Correlation matrix of all five continuous variables with Spearman’s rho and P-value

	Thumbs-up	Thumbs-down	Number of views	Views per day	Video length (min)
Thumbs-up	1.00	0.43	0.38	0.45	0.18
		<0.0001*	<0.0001*	<0.0001*	0.08
Thumbs-down	0.43	1.00	0.32	0.37	0.10
	<0.0001*		0.0013*	0.0002*	0.35
Number of views	0.38	0.32	1.00	0.76	0.07
	<0.0001*	0.0013*		<0.0001*	0.50
Views per day	0.45	0.37	0.76	1.00	−0.01
	<0.0001*	0.0002*	<0.0001*		0.90
Video length (min)	0.18	0.10	0.07	−0.01	1.00
	0.08	0.35	0.50	0.90

* <0.05.

The frequency of mention of pertinent information regarding tinnitus was obtained for each of the three sources of video uploads, and is shown on Table 4. Regarding the nature of tinnitus and associated parameters, almost all the videos across sources mentioned subjective tinnitus (93/100), whereas only 5 out of 100 videos featured a celebrity. A third of the videos mentioned fear of being diagnosed with tinnitus (34/100). Regarding symptoms of tinnitus, half of the videos mentioned pain as a symptom of tinnitus. Regarding causes of tinnitus, exactly half of the videos also mentioned loud noise as a cause of tinnitus. Only four videos (4%) cited acoustic neuroma (brain tumor) as cause of tinnitus.

Table 4

Frequency count of 100 tinnitus-related videos in English by their sources and contents

Content	Source category of video
	Consumer	Media	Professional	Total (%)
(1) Symptoms
Mentions subjective tinnitus*
Yes	40 (95.24)	30 (90.91)	23 (92)	93 (93)
No	2 (4.76)	3 (9.09)	2 (8)	7 (7)
Mentions objective tinnitus
Yes	16 (38.10)	18 (54.55)	23 (92)	57 (57)
No	26 (61.90)	15 (45.45)	2 (8)	43 (43)
Mentions pulsatile tinnitus*
Yes	0 (0)	1 (3.03)	2 (8)	3 (3)
No	42 (100)	32 (96.97)	23 (92)	97 (97)
Pain mentioned (if any)
Yes	24 (57.14)	13 (39.39)	13 (52)	50 (50)
No	18 (42.86)	20 (60.61)	12 (48)	50 (50)
Mentions anxiety/fear of diagnosis/screening
Yes	13 (30.95)	11 (33.33)	10 (40)	34 (34)
No	29 (69.05)	22 (66.67)	15 (60)	66 (66)
(2) Causes
Mentions loud noises as a cause of tinnitus
Yes	21 (50)	13 (39.39)	16 (64)	50 (50)
No	21 (50)	20 (60.61)	9 (36)	50 (50)
Mentions acoustic neuroma (brain tumor) as cause*
Yes	2 (4.76)	2 (6.06)	0 (0)	4 (4)
No	40 (95.24)	31 (93.94)	25 (100)	96 (96)
Mentions hearing loss as a cause of tinnitus
Yes	8 (19.05)	5 (15.15)	8 (32)	21 (21)
No	34 (80.95)	28 (84.85)	17 (68)	79 (79)
Mentions Ménière’s disease as a cause of tinnitus*
Yes	1 (2.38)	1 (3.03)	1 (4)	3 (3)
No	41 (97.62)	32 (96.97)	21 (96)	97 (97)
Mentions diseases of heart/blood vessels as a cause*
Yes	3 (7.14)	4 (12.12)	1 (4)	8 (8)
No	39 (92.86)	29 (87.88)	24 (96)	92 (92)
Mentions diseases of thyroid vessels as a cause*
Yes	0 (0)	1 (3.03)	0 (0)	1 (1)
No	42 (100)	32 (96.97)	25 (100)	99 (99)
Mentions diseases of hormonal change as a cause*
Yes	1 (2.38)	2 (6.06)	2 (8)	5 (5)
No	41 (97.62)	31 (93.94)	23 (92)	95 (95)
Mentions drugs as a cause of tinnitus*
Yes	4 (9.52)	4 (12.12)	4 (16)	12 (12)
No	38 (90.48)	29 (87.88)	21 (84)	88 (88)
(3) Treatment
Treatment mentioned: hearing aid*
Yes	2 (4.76)	5 (15.15)	3 (12)	10 (10)
No	40 (95.24)	28 (84.85)	22 (88)	90 (90)
Treatment mentioned: counseling
Yes	13 (30.95)	11 (33.33)	10 (40)	34 (34)
No	29 (69.05)	22 (66.67)	15 (60)	66 (66)
Treatment mentioned: sound generators white nose
Yes	6 (14.29)	7 (21.21)	7 (28)	20 (20)
No	36 (85.71)	26 (78.79)	18 (72)	80 (80)
Treatment mentioned: cochlear implants*
Yes	1 (2.38)	0 (0)	1 (4)	2 (2)
No	41 (97.62)	33 (100)	24 (96)	98 (98)
Treatment mentioned: antianxiety/depressants*
Yes	3 (7.14)	2 (6.06)	0 (0)	5 (5)
No	39 (92.86)	31 (93.94)	25 (100)	95 (95)
Treatment mentioned: acoustic neural stimulation*
Yes	4 (9.52)	2 (6.06)	2 (8)	8 (8)
No	38 (90.48)	31 (93.94)	23 (92)	92 (92)
(4) Comorbid symptoms
Mentions that tinnitus causes insomnia
Yes	14 (33.33)	9 (27.27)	9 (36)	32 (32)
No	28 (66.67)	24 (72.73)	16 (64)	68 (68)
Mentions that tinnitus sufferers have difficulty coping
Yes	33 (78.57)	18 (54.55)	20 (80)	71 (71)
No	9 (21.43)	15 (45.45)	5 (20)	29 (29)
(5) Mentions prevention of tinnitus
Yes	17 (40.48)	11 (33.33)	14 (56)	42 (42)
No	25 (59.52)	22 (66.67)	11 (44)	58 (58)
(6) Nature of video
Purpose of the video is to sell a product or service
Yes	10 (23.81)	13 (39.39)	6 (24)	29 (29)
No	32 (76.19)	20 (60.61)	19 (76)	71 (71)
Video is about an individual experience
Yes	26 (61.90)	9 (27.27)	6 (24)	41 (41)
No	16 (38.10)	24 (72.73)	19 (76)	59 (59)
(7) Features a celebrity*
Yes	2 (4.76)	2 (6.06)	1 (4)	5 (5)
No	40 (95.24)	31 (93.04)	24 (96)	95 (95)

*These content categories had at least one cell <5. We did not calculate odds ratios in Table 5 for these categories due to very small samples.

Table 5

Odds ratios of certain content and theme between sources (media as the reference group)

Content category (Ref.: media)	Odds ratio (95% CI)	P-value
(1) Symptoms
Mentions objective tinnitus
Consumer	0.51 (0.20, 1.29)	0.16
Professional	9.58 (1.94, 47.42)	0.01*
Mentions pain (if any)
Consumer	2.05 (0.81, 5.19)	0.13
Professional	1.67 (0.58, 4.76)	0.34
Mentions anxiety/fear of diagnosis/screening
Consumer	0.90 (0.34, 2.38)	0.83
Professional	1.33 (0.45, 3.92)	0.60
(2) Causes
Mentions loud noises as a cause
Consumer	1.54 (0.61, 3.88)	0.36
Professional	2.74 (0.93, 8.01)	0.07
Mentions of hearing loss as a cause
Consumer	1.32 (0.39, 4.48)	0.66
Professional	2.64 (0.74, 9.38)	0.14
(3) Treatment
Treatment mentioned − counseling
Consumer	0.90 (0.34, 2.38)	0.83
Professional	1.33 (0.45, 3.92)	0.60
Treatment mentioned − sound generators white nose
Consumer	0.62 (0.19, 2.06)	0.43
Professional	1.44 (0.43, 4.83)	0.55
(4) Comorbid symptoms
Mentions causes insomnia
Consumer	1.33 (0.49, 3.62)	0.57
Professional	1.50 (0.49, 4.60)	0.48
Mentions difficulty coping
Consumer	1.12 (0.51, 4.73)	0.03*
Professional	1.20 (0.61, 3.89)	0.05
(5) Nature of video
Purpose of the video is to sell a product or service
Consumer	0.48 (0.18, 1.30)	0.15
Professional	0.49 (0.15, 1.54)	0.22
Video is about an individual experience
Consumer	4.33 (1.62, 11.63)	0.004*
Professional	0.84 (0.26, 2.78)	0.79
(6) Prevention mentioned
Consumer	1.36 (0.53, 3.52)	0.53
Professional	2.55 (0.87, 7.43)	0.09

CI = confidence interval. * <0.05.

Regarding the treatments for tinnitus, not many videos mention about different treatment, except counseling, for which, there were 34 (34%) videos mentioned about counseling as a treatment for tinnitus. Regarding the prevention of tinnitus, 42 (42%) videos mentioned prevention and among which, 14 videos are from professional-generated videos. The percentage of professional videos discussing about the prevention of tinnitus is the highest among three source groups (56%).

The likelihood that pertinent information about tinnitus would be mentioned in a video depending on its source of upload was calculated via ORs for the three sources. The pertinent information is listed in Tables 4 and 5. Videos from media sources were used as the reference category in calculating the ORs of consumer and professional videos (shown in Table 5). Note that some content categories have at least one cell size <5, so we did not calculate ORs in 5 for these categories due to very small samples.

The odds of mentioning objective tinnitus in professional videos were 9.58 times that of media videos [OR = 9.58; 95% confidence interval (CI): 1.94, 47.42; P = 0.01], whereas, the odds of mentioning objective tinnitus in consumer videos were 51% of media videos, but this was not statistically significant (OR = 0.51; 95% CI: 0.20, 1.29; P = 0.16). In addition, consumer videos were found to be 4.33 times as likely to carry a theme about an individual experience of tinnitus as media videos (OR = 4.33; 95% CI: 1.62, 11.63; P = 0.004), whereas there was no difference between media videos and professional videos in this regard.

Discussion

Most viewed videos by source

The videos created by commercial interests were the ones viewed the most, with almost 1 million views more than the second placed videos (from media sources). However, the difference between the total numbers of thumbs up for consumer and media videos was only 19. While most videos from consumers were based on personal experiences with tinnitus with few graphics with less appeal, videos from media sources were visually more pleasing and interesting. Further, males were more likely to feature in tinnitus videos than females (39% of the people who featured in the videos were males as compared to the 22%).

Video information regarding the nature of tinnitus

The videos overwhelmingly (93 out of 100) mentioned tinnitus as a subjective condition. This gives credence to the widespread belief that tinnitus is a subjective condition based on the perception of the affected individual.[17] Subjective tinnitus is a type of tinnitus where only the individual with the symptom perceives the auditory stimulus. This is the most prevalent kind of tinnitus, and can be due to issues of the outer, middle, or inner ear, as well as by pathologies of the auditory nerves or pathways (the brain structures that interpret nerve signals as sound). It is not surprising that 23 out of 25 videos from professional sources describe objective tinnitus compared to other sources of videos. The plausible explanation for this is that professionals, particularly physicians, diagnose most conditions objectively.[8] In contrast, consumer videos are more likely to mention subjective individual experience and difficulty in coping with tinnitus.

Video information regarding the causes and prevention

Video information regarding the causes and prevention did not differ significantly among video sources given the current data set of top 100 most viewed videos. However, professional-generated videos had slightly greater percentages of presenting causes and preventions.

Video information regarding the treatment options for tinnitus

Studies have shown that there is no consistent alleviation or total elimination of the sensation of tinnitus[18] and the aim of treatment is usually the scaling down of reactions to tinnitus. Although various medications are used for tinnitus, none has been approved exclusively for its treatment, and may better serve the treatment of comorbid conditions such as anxiety and depression that are often associated with tinnitus.[18] It is also important to note that the result of the statistical analysis of the YouTube videos on tinnitus showed that it was overwhelmingly stated across all sources (mostly by more than 90%) that none of the treatments mentioned was a cure to tinnitus, underscoring the point that tinnitus is a difficult condition to treat.

Limitations

The limitations of our current study include using an arbitrary cut-off point of the 100 most popular videos and using only those that were recorded in the English language. As this is a cross-sectional study, longitudinal changes in view count, and other meta-data of the videos are not captured. The authors did not have access to data regarding the demographics of the viewers and the time and location at which they viewed the videos. The authors did not code video content for misinformation.

Conclusion

The challenges faced by clinical and public health professionals in engaging the public via the social media include the prevalence of poor quality information therein and the potential risks to patients of being misled by poor quality information. From the results of our research, we found only one fourth mostly viewed English-language tinnitus-related videos are uploaded from professionals, while the rest are mostly individual personal experience shared by consumers. Unlike evidence-based medicine, social media users often emphasize anecdotal reports, and depend greatly on individual patient reports for dissemination of medical information.[19] Therefore, consumers need to be cautious of health information gleaned from YouTube as it may have limited value. Meanwhile, to make professional videos more popular among all tinnitus-related videos, healthcare and public health professionals should consider engaging key opinion leaders on social media (such as celebrities) in their health campaigns and have them posted these videos on their social media channels. To have the support of key opinion leaders to post and re-post a campaign’s video on their social media channels at the beginning of a campaign (“seeding strategy” in the parlance of marketing) was found to be vital in making a video viral,[20] and should be considered an integral part of any health campaign strategy using social media as a means of health communication.

To conclude, professional videos about tinnitus contained more pertinent information compared to consumer videos or media videos. Actions are needed to make scientific medical information more prominently available and accessible on YouTube and other social media.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.