The Development and Validation of a Questionnaire Measuring Digital Eye Strain and Risk Level (DESRIL-27)

Background: Daily exposure and overuse of digital devices are associated with several health-related issues to human eye, muscles and bones that are collectively termed as Digital Eye Strain and related Musculoskeletal Disorders. These problems have become worse during covid-19 pandemic due to global shift towards working online. This paper introduces a validated research instrument called DESRIL-27 to be used in assessing digital eye strain and related musculoskeletal disorders, with their associated risk factors in the working population. Methods: DESRIL-27 was developed from thorough review of salient literatures and in consultation with a team of experts to ensure the content validity. This 27-item tool consists of two scales, namely symptom severity and risk level scales. Reliability and validity analyses were conducted to assess the psychometric properties of DESRIL-27. Results: Cronbach’s alpha was 0.91 for Symptom Severity and 0.88 for Risk Level from the scales in the questionnaire. Principal Component Analysis results ranged from acceptable to very good. Conclusions: The procedures adopted for the validation and the results of the analysis reveal that DESRIL-27 is a reliable and valid research instrument. It can be safely used in future research studies to assess the visual and musculoskeletal problems and their risk factors in digital screen users. The resultant knowledge can help individuals, health practitioners, and related policy makers to prevent and control these problems at workplace.


Background
The increasing use of digital devices for work and social purposes in response to technological advancement, often for prolonged hours each day, is now normal among individuals of all ages.
Research shows that using digital devices (desktops, laptops, tablets, e-readers and cell phones) for more than 4 hours daily, increases the risk of developing a clinical syndrome called Computer Vision Syndrome or CVS [1,2]. It has been de ned as "a group of eye-and vision-related problems that result from prolonged computer, tablet, e-reader and cell phone use" [3]. CVS is synonymous with the terms Visual Fatigue (VF) and Digital Eye Strain (DES), re ecting the variety of digital devices to be the potential source of these problems [4]. DES encompasses a range of ocular and visual symptoms, which include eyestrain, tired eyes, sore eyes, dry eyes, watery eyes, irritated eyes, blurred vision, and double vision [2,5]. Computer use has also been reported to be associated with a range of musculoskeletal disorders (MSD) especially pain, stiffness and numbness in neck, shoulders and back [6]. A systematic review revealed that digital devices, speci cally the use of smartphones, may contribute to the occurrence of musculoskeletal changes in the head-neck, shoulder-arm, and hand-thumb areas [7]. The substantial growth in digital device usage is especially evident during the COVID-19 pandemic, with home-isolation and social distancing leading to a necessitated use of digital devices to work or study online, posing a greater risk of developing digital eye strain-related symptoms as well as numerous social, psychological and physical issues. While symptoms are usually transient, the condition can cause signi cant and frequent discomfort for sufferers, leading to increased disability, reduced occupational productivity and absenteeism from the workplace [8,9].
Digital eye strain has been regarded as one of the biggest occupational hazards of the century [10].
Globally, around 60 million people suffer from DES and an estimated 1 million new cases occur each year [11]. A systematic review of 104 studies on DES between 1980 -2014 (around the period where computers have become a necessity to daily life) reported the prevalence of DES to range between 64-90% of the population and has been accounted as the most reported health problem in 70% of computer users [2]. According to The Occupational Safety and Health Administration of the US Government (OSHA), 90% of 70 million US workers who use computers for more than three hours per day, experience some form of CVS [12,13]. Meanwhile, a study reported that 22% of computer workers complained of various musculoskeletal disorders such as neck, back and shoulder pains, and carpal tunnel syndrome, attributed mainly due to poor workstation ergonomics and long sitting periods [14].
Many of the current validated questionnaires on DES assess vision related symptoms only and do not include items for musculoskeletal symptoms or risk factors for DES [15,16]. Hence, researchers had to use multiple questionnaires in their studies to assess DES and associated risk factors [17,18]. In this study, we report the development and validation of a questionnaire called the Digital Eye Strain and Risk Level Questionnaire (DESRIL-27) that incorporates visual and musculoskeletal symptoms of DES and the assessment of risk factors.

Study Design
A questionnaire validation study was conducted in three stages: designing the questionnaire by experts, pretesting questionnaire to assess comprehensibility and nally, psychometric assessment. The study protocol was conducted in accordance to the Declaration of Helsinki and ethical approval was obtained from the PAPRSB Institute of Health Sciences Research Ethics Committee (IHSREC) and University Research Ethics Committee (UREC), Universiti Brunei Darussalam (UBD/PAPRSBIHSREC/2021/11).
Designing the questionnaire by expert committee The questionnaire was designed using the "Table of Speci cation" approach [19][20][21] to ensure content validity [22,23]. The approach started with setting the scope of measurement scale. The rst scale (Symptom Severity Scale) measured symptoms related to eye, vision and musculoskeletal system. The second scale (Risk Level Scale) measured risk level regarding work environment (workspace, device or equipment, work hours/breaks/exercises) and ergonomic factors, speci cally workstation set-up in digital screen users ( Table 1). The areas in each scale were discussed and agreed upon, followed by points in each area. Finally, corresponding items were identi ed by experts for each point. In addition, we assessed the available items from existing validated questionnaires and adopted relevant items, for example, Computer Vision Syndrome Questionnaire (CVS-Q) and Computer Vision Symptom Scale (CVSS17) [15,16] and recommendations from American Optometric Association and Centers for Disease Control and Prevention [3,24]. A total of 36 items were included in the rst draft of the questionnaire and the number reduced to 27 items in the nal questionnaire. Pretesting the questionnaire A comprehensibility assessment (pretesting) of the drafted questionnaire was conducted to check the ease of understanding and acceptability for all components of the questionnaire including introduction, instructions, items and responses [23]. We conducted two sessions of pretests including a total of 14 participants. The participants for both sessions were recruited by purposive sampling and were comprised of administrative staff of Universiti Brunei Darussalam (UBD) from a low educational background, using digital screens regularly and were cooperative to give feedback to the questionnaire. After the rst session, the questionnaire was revised according to the feedback given by the participants followed by a second session, which was conducted with the revised version. The questionnaire was revised again according to the feedback from the second session. We removed four items which were identi ed to have poor comprehensibility with participants and better similar items were present in the questionnaire. The revised questionnaire had 32 items.

Psychometric assessment
Following pre-testing, we conducted a psychometric assessment to determine whether the items should be revised, omitted or kept unchanged [23]. The revised questionnaire with 32 items was tested among 42 members of administrative staff of UBD from various educational backgrounds, who had been working with digital screens for at least one year and were recruited by purposive sampling. The purpose of sampling here, was to recruit a range of participants from those who had minimal use of digital screen to those who had maximal use, almost all the time. The members who participated in pretests were not included in this study. A researcher was present to address any queries and clari cations by the participants who answered the paper-based questionnaire. With the results of psychometric analysis, the questionnaire was revised, and the nal revision was obtained which had 27 items; 16 items in Symptom Severity Scale and 11 items in Risk Level Scale ( Table 1).
Scoring of the questionnaire Symptom Severity Scale has two components, frequency and intensity. There are four responses for frequency of symptom: never, sometimes, often and almost always, and are given scores 0, 1, 2 and 3 respectively. Intensity scale has four responses: mild, moderate, intense and very intense, and are given scores 1, 2, 3 and 4 respectively. Each item score was calculated by multiplying frequency and intensity scores, having values from 0 to 12. Therefore, the Symptom Severity Scale ranges from 0 to 192 for the 16 item-scale. Higher score re ects higher symptom severity.
For Risk Level Scale, there are 11 items with 4 responses. Three items have the following responses: very comfortable, comfortable, slightly uncomfortable, and very uncomfortable, and are given a score of 0 to 3 respectively. Remaining 8 items have the following responses: more than adequate, adequate, slightly inadequate, and very inadequate and are given a score of 0 to 3 respectively. Therefore, the Risk Level Scale ranges from 0 to 33 for the 11 item-scale. Higher score re ects higher risk level.

Statistical Analysis
All analysis was conducted using R (ver. 4.1.1) and RStudio for Mac (ver. 4.0.2). The package "psych" was used to assess psychometric properties of the questionnaire [25]. Quantitative variables were presented with means and standard deviations or median and IQR (Interquartile range) as appropriate.
Qualitative variables were presented with frequency distribution and percentages. Analyses on psychometric properties included internal consistency reliability analysis for reliability, principal component analysis for construct validity and item and scale level descriptive statistics to assess the sample distribution adequacy. Each scale was described with mean and standard deviation, minimum and maximum score, and oor and ceiling percentages. Floor percentage was calculated as the percentage of respondents at the lowest scale point (score zero for Symptom Severity Scale and zero for Risk Level Scale). Ceiling percentage was calculated as the percentage of respondents at the highest scale point (score 192 for Symptom Severity Scale and 33 for Risk Level Scale). These statistics assess the adequacy of purposive sampling that aims to have a good range of sample in the validation study.

Results
A total of 42 respondents participated in the psychometric assessment. The median age of the participants was 35.5 years, and all were full-time employees working in the administration o ces of UBD. Majority of the respondents was females (64.3%), education below bachelor degree (88.1%), income less than Brunei Dollars (BND) 2000 (88.1%), using spectacles (62.0%), had no regular eye checkup (81.0%) and used digital screens for more than 5 hours per day (66.7%). Detailed description of the respondents is presented in Table 2. Psychometric properties of the questionnaire Initially, 32 items were analyzed for psychometric properties. There were three poorly performed items, with corrected item-total correlation and poor factor loadings (<0.40 in both) on the respective component in Principal Component Analysis. Items' review showed that there were similar items available. Therefore, these three items were dropped from the questionnaire, as it did not compromise the content validity.
As the Cronbach's alpha was still high, 0.91 for Symptom Severity Scale and 0.88 for Risk Level Scale, we identi ed similar items in the questionnaire and dropped two items without compromising the content validity. Finally, with the 27-item questionnaire, it satis ed a good reliability and construct validity with some minor or acceptable weakness. There were four items S2, S5, S8 and RL1, which we considered to have marginally acceptable corrected item-total correlation whereas their factor loading was good in respective component. The eigen value extracted by Component 1 was 7.50 units and by Component 2 was 5.94 units. Variance extracted by Component 1 was 28% and by Component 2 was 22%, giving a total of 50% of variance extracted by the two components. We present details of psychometric statistics of the DESRIL-27 in Table 3.  Table 4 shows descriptive statistics of Symptom Severity and Risk Level Scales. It revealed that for the Symptom Severity Scale, the lowest score was 1 and the highest score was 97, while the scale ranges from 0 to 192. The sample taken occupied well in the lower half of the scale but not adequately occupied in the upper half of the scale. With regard to Risk Level Scale, the lowest score was 1 and the highest score was 28, while the scale ranges from 0 to 33. Though it did not reach the lowest and highest of the scale, the sample taken spread or occupied well in the scale range. However, the psychometric analysis gave a good outcome in this study. While reviewing the literature on effects of digital screen use on human eye and musculoskeletal system, it was found that there is a gap in the knowledge to understand the symptoms and risk factors due to lack of validated questionnaires. Several studies used instruments which did not follow the standard patterns of questionnaire validation and due to this reason, there is discrepancy among the results, with prevalence of DES to be 19.6% in one study, and more than 80% in another study [29,30].
When we compare our instrument to other validated questionnaires, it is found that the reliability of DESRIL-27 is superior to CVS-Q and is equivalent to CVSS17. For our instrument, the reliability coe cient (Cronbach's alpha) for symptom severity of DES is 0.91 whereas for CVS-Q and CVSS17, it is 0.78 and 0.92 respectively. Regarding ICC (Intraclass correlation) which is done for test-retest reliability [31], the CVS-Q has a value of 0.80 and CVSS17 has a value of 0.85. It shows that the ICC values of both these instruments are also comparable to our reliability coe cient [15,16].
This newly developed and validated questionnaire DESRIL-27 is unique in the sense that, in addition to eye problems, it includes items to assess the effects of musculoskeletal symptoms associated with the use of digital screens, which other questionnaires such as CVS-Q and CVSS-17 lack in [15,16]. Hence, this questionnaire can be used as a guide for employers to create awareness on the importance of ful lling the ergonomic requirements while at work with digital screens.
The methodology used for the questionnaire validation is considered to be the main strength of the study.
All steps followed in this study were according to the standard practice of validation of health questionnaires [27]. Both pretests and the psychometric assessment were conducted in the presence of the researchers. The questionnaire was paper based, easy to understand, with a 12-minute completion time on average.
A limitation of the study is that the diagnosis of DES is through self-reporting where reporting bias could arise. There could also be some variation in different culture and nature of workplace which could introduce measurement errors. Therefore, this questionnaire should be further evaluated in other populations and also in different work settings in future studies.
With COVID-19 pandemic continuing its presence, working or schooling from home and communication using digital technologies will inevitably be part of the new norm, where many individuals may suffer from the physical effects of overuse of digital devices. By gaining information on the symptom severity and risk levels, individuals may get away from prolonged use of digital devices. This may guide them to be more empowered in management of their health by ensuring visual and musculoskeletal health awareness. In addition, employers can explore strategies to manage DES and MSD effectively in the workplace such as adopting proper ergonomic measures, incorporating regular screen breaks and practicing the 20-20-20 rule in the workplace [32]. Good health of workers is not only bene cial for themselves but also critical for the growth of the organizations as well as the society as a whole.
Future studies on impact of digital devices and health can include assessment of ergonomic and environmental factors. Comparative studies can also be done among different occupations in the same country and among same occupations in different countries. This will not only help determine the magnitude of DES and MSD but also the associated risk factors in other parts of the world.
The researchers have prepared a manual and a coding guideline for the scoring items to assist the researchers in conducting their future studies using DESRIL-27. All this material along with the questionnaire are available on our website (https://sites.google.com/view/thedesril-27project/).

Conclusion
This study has successfully shown that DESRIL-27 is a reliable and valid instrument for assessment of DES and MSD symptom prevalence and risk factors. DESRIL-27 provides opportunities for other researchers to conduct prevalence studies on DES and MSD among individuals from various professions working with digital screens. This in turn may lead to further research opportunities to determine the causative factors for DES and MSD due to digital screen use. By doing so, this will enable stakeholders to manage the visual and musculoskeletal health of employees by implementing necessary modi cations in the work environment.