A Usability Evaluation Instrument for Pain Management Mobile Applications: An Elderly’s Perspective

Abstract Technology offers new opportunities in healthcare to manage chronic conditions cost-effectively. In the adoption of mHealth services, one of the main barriers is usability, especially for users having limitations or special needs, such as the elderly. Unlike mHealth apps specifically designed for the elderly to monitor and manage chronic conditions such as cardio-vascular health, diabetes, and hypertension, there is a lack of research in pain management, including usability evaluation. Moreover, there is a lack of comprehensive usability evaluation tool. This research focuses on the formulation and assessment of the usability evaluation instrument for the mHealth apps, targeting elderly users. An instrument comprising 8 usability constructs and 58 distinct items was formulated. The content of the instrument was verified by five experts. Face validity was performed with 16 elderly pain sufferers on pre-selected pain apps, it also aids in validating the reliability of the instrument. The research also discusses the mapping of the instrument’s items with the usability measures and elderly barriers influencing technology, with graphical visualization and scores of various artefacts. The graphical details would be helpful in identifying the areas for usability improvement from the elderly’s perspective and the scores would be helpful in comparing usability of different applications.


Introduction
According to WHO (World Health Organization), elderly, i.e., 65 years old and above people are rapidly becoming a majority of the worldwide populace (Chuna & Patterson, 2012).There are several requirements to sustain the elderly's living, which include health monitoring, information seeking, social interaction, relaxation activities, security, and protection (Hassan & Md Nasir, 2008;Plaza et al., 2011).In this context, ICT (Information and Communication Technology) plays a significant role.A new domain named Gerontechnology has been evolving which refers to the technology that facilitates the aging society as per their needs.It means incorporating research, development, and design in the engineering disciplines keeping in view the peculiarity of the aging phenomenon.Among the areas of research covered by Gerontechnology, it specifically includes areas of independent living, caregiver support, and healthcare (Plaza et al., 2011).
However, the elderly being resistant to the adoption of new technologies are categorized as late majority or laggards according to Rogers Innovation Diffusion theory (Wong, 2011).According to Rogers (1995), adoption of any innovation is influenced by five factors: relative advantage, compatibility, trialability, observability, and complexity.The first four factors have a positive correlation with the rate of adoption of any innovation, whereas the adoption rate declines with increased complexity.
In recent years, studies have shown the elderly's growing interest in mobile health technology, proving the myth of the elderly's less diversion towards the technology to be less legitimate (Huguet et al., 2015).Besides, mobility and independent living mobile technology could be a promising tool in the self-management of chronic diseases, thus enhancing the quality of living of the elderly (Holzinger et al., 2007;Lindeman, 2015;Plaza et al., 2011).Nevertheless, certain barriers described in the next section put hurdles in the acceptance of mobile technology by the elderly.It is necessary to understand the complexities, uncertainties, and difficulties encountered by the elderly to give them user-friendly solutions and increase motivation and acceptance in adopting the solution.
This research is confined to provide a usability assessment tool that covers major aspects of mobile app's usability evaluation for the elderly.It will be significant for the professionals to involve users in a comprehensive usability evaluation of healthcare mobile applications concerning the elderly.The instrument also contributes to the quantification of usability measures as well as the elderly barriers of the mHealth (mobile health) application.The visualization of assessment results will help to identify the areas of improvement.Researchers and designers could use this instrument to modify or develop the app in accordance with the tool's highlighted areas of improvement.The solution could also be helpful for the clinicians to choose among the existing pain management apps and recommend the most suitable app for elderly patients as a self-management regime.

Background
This section reviews literature related to the elderly's barriers to technology use, usability of mHealth technology from the elderly's perspective, and pain management mobile applications and the elderly.

Elderly's barriers to technology use
There are several specific technological issues for the elderly, which include, cognition, motivation, physical and perception issues (Czaja et al., 2013;Holzinger et al., 2007;W. A. Rogers & Fisk, 2010).
Cognitive issues are related to a reduced capacity of working, prospective, semantic, and procedural memory, attention, processing of time-restricted discrete information, ability to recall, increased time to learn new skills, and decline in numeracy and representational fluency (Wildenbos et al., 2018).As cognitive performance slows down with age so for the complex applications, it is recommended to simulate user interaction and alter the design of the applications upon feedback to produce a simplified design.A simple application can help to match and enhance the performance of elders, as implicit learning ability does not reduce by aging (Holzinger et al., 2007).
The motivational issues are comprised of poor understanding of benefits, problems with beliefs, attitudes, anxiety, fear, computer literacy, and acceptability (Holzinger et al., 2007;Wildenbos et al., 2015).Most elders have poor or less knowledge of the benefits of using a mobile application.They also feel uncomfortable learning new technology and lose concentration easily and become bored while learning.Sometimes they don't have trust in their ability to learn or capability to operate certain devices (Holzinger et al., 2007).
The physical issues are problems with the physical body parts of elders that results in reduced flexibility and motor skills (Holzinger et al., 2007).Physical barriers also affect slower movement and reflexes, stiffer muscles and joints, tremors (in hands), and a diminished balance or difficulty in holding a device (Wildenbos et al., 2018).Also, basic ICT usage might result in a challenge for most elders, since they might have a difficult time locating items on the screen, and will require substantial effort to have the hand-eye coordination required to use mobile devices (Mendez & Mendoza, 2013).Such issues lessen the confidence of elders to learn new things and absorb knowledge (Holzinger et al., 2007).
The perception issues include barriers to vision and audition (Holzinger et al., 2007).The abilities that decrease with age include visual acuity, visual accommodation, color vision, contrast detection, dark adaptation, glare, motion perception, and peripheral vision (Holzinger et al., 2007).The factors where the necessity increases with age involved demand for illumination increases with age, and a significant reduction of auditory facilities with age (Holzinger et al., 2007).
The aforementioned barriers offer numerous rational reasons for why mHealth applications are not adopted by the elderly.Longer learning times and low visual acuity were the two most often mentioned causes among those found.It is possible to correlate various conditions that an older adult might experience and understand which aspects of usability might impede by this condition (Paiva et al., 2020).

Usability of mobile health technology from elderly perspective
Some of the foundational studies (Darroch et al., 2005;Goodman et al., 2005;Siek et al., 2005;Sterns, 2005) addressed the usability of handheld technology by the elderly and showed that portable technology can be helpful in providing mobility and independence for the elderly.However, if the objectives and needs are not fulfilled and the solution is not user-friendly and does not provide ease to use, it will hinder the elderly's acceptance of technology (Nimmanterdwong et al., 2022).Thus, usability is essential to technology adoption and is a significant field of research (Caboral-Stevens et al., 2015).
A few previous studies investigated the usability evaluation of mobile technology with respect to the elderly.Becker and Webbe (2006), proposed a framework based on Schneiderman's model, comprising four main components: user profile, environmental factors, usability quality factors, and technology objective; to uncover the potential barriers associated with the interface design.The framework could be used as a code of conduct to execute and report a usability study involving PPC (Pocket PC) and elderly caregivers; and define measurable outcomes for a usability study involving the elderly.Mendez and Mendoza (2013) proposed a conceptual framework based on the principles of ISO/IEC 2000 (International Organization for Standardization/International Electrotechnical Commission), abstraction hierarchy commonly found in the cognitive systems (Rasmussen, 1986) and Ham's hierarchical usability model (Heo et al., 2009).This framework identifies a relationship of the elderly barriers with the usability criteria and usability properties.As this model presents a complex multi relation scheme, it would be difficult to implement this model and a comprehensive checklist is required for an elaborative evaluation scheme.
Another study (Wildenbos et al., 2015) suggested a framework for the evaluation of mHealth tools for elderly patients.It is a theoretical framework that explains the agerelated barriers impacting mHealth's ease of use.Based on the literature, the researchers identified the age-related barriers and developed an app for the elderly.The usability issues were analyzed via a cognitive walkthrough.The issues were then categorized based on the mHIMSS (mobile Healthcare Information and Management Systems Society) themes and severity ranking was performed as per Neilson's classification of severity.The usability issues were further mapped with the potential underlying feature of the elderly that could be the reason behind that usability issue.This study concluded that mHIMMS design guidelines are not sufficient for developing highly usable apps and this framework could give insight that can help in the (re)design of the app.
Wildenbos et al.'s work was further extended by the researchers where a framework MOLD-US was presented to summarize the evidence of the effect of aging barriers on the use of mHealth (Wildenbos et al., 2018).A scoping review was carried out to formulise the physical and functional age-related barriers and their causes, due to some common medical conditions, in accordance with the usability issues of the elderly in using mHealth apps.The usability aspects in this framework are based upon Neilson's quality components: errors, efficiency, learnability, memorability, and satisfaction (Nielsen, 2012).The correlation between aging barriers and medical conditions is visualised in the framework.However, it is not necessary that a specific disease only leads to that particular barrier as specified in the framework.The elderly could have multiple morbidities and a mHealth intervention should address multiple aging barriers.Table 1 summarises the usability evaluation measures of the aforementioned studies.
It has been observed that all the aforementioned usability frameworks do not include the same set of attributes.Learnability, efficiency, and satisfaction are the common attributes among all the frameworks; whereas, retention overtime/memorability, error rate, effectiveness, functional support as per user needs, and control to operate are not taken into account in some of the frameworks.This highlights the need for a common and comprehensive usability evaluation framework.In the aforementioned studies, there is also a lack of involvement of elderly patients using healthcare mobile applications in the validation of the proposed frameworks.Studies on the realistic use by the elderly are required to understand how age obstacles affect elderly patients' use of mHealth applications.
The frameworks also provide theoretical knowledge of usability evaluation without a comprehensive assessment tool for mapping usability evaluation with the age-related barriers.Such assessment tool will be helpful in analyzing the areas of improvement and (re)designing the mHealth applications with respect to the elderly population.Despite few available guidelines (Aging & Medicine, 2001;Silva et al., 2015) for performing usability evaluation with respect to the elderly, little research has been conducted towards the application of the usability assessment tool with respect to the elderly barriers to technology use.There is a gap in this area of research.

Pain management mobile applications and elderly
Self-management of chronic diseases could be made effective by giving easy communication, accessible information, and simple data storage and retrieval techniques to the patients, caretakers, and clinical experts.Mobile health technology could prove beneficial in this context and capable of reforming patients' conduct towards their wellbeing.Similar to the youth population, elders may be benefited from mobile health technology (Bakshi et al., 2015).According to WHO, Learnability "Easiness of learning the system and retrieving information from the system" Learnability "How easy is it for users to accomplish basic tasks the first time they encounter the design?"

Speed of Performance
"The time it takes to complete the benchmarked tasks." Efficiency "Support of efficient interaction" Efficiency "Amount of effort required to use the system and usefulness in meeting the user's needs and goals." Efficiency "Once users have learned the design, how quickly can they perform tasks?"Error Rate "The number, type, and severity of errors made by a typical user when completing the benchmarked tasks" --Errors "How many errors do users make, how severe are these errors, and how easily can they recover from the errors?"Retention over time "The knowledge that is maintained after a period of non-use." --Memorability "When users return to the design after a period of not using it, how easily can they re-establish proficiency?"Subjective Satisfaction "The user's feedback on how well he or she liked various aspects of the user interface" Satisfaction "Ergonomic support" Perceived User Experience "Level of pleasantness in using the system and the quality of the system's design."the elderly, are rapidly becoming a majority of the worldwide populace as well as the population group suffering from the chronic diseases than young adults (Chuna & Patterson, 2012), having three diseases on average (Mira et al., 2014) where the pain is their common illness.
As identified from a literature review of usability issues in healthcare mobile applications for the elderly (Shah & Chiew, 2017), numerous elderly-specific healthcare mobile applications exist for certain diseases like cardiovascular health, diabetes, fall detection and prevention, and Alzheimer's disease.However, far too little attention has been paid to the area of mHealth and pain management in relevance to the elderly age group, although chronic pain ranks six among the major chronic conditions (Gereau et al., 2014).
An integrative review was conducted in a study (Bhattarai & Phillips, 2017) that evaluates the role of digital health technologies in the management of pain in the elderly.The review revealed that most studies involved computer-based interventions.Only one study (Parker et al., 2013), which was qualitative (focus group discussion), involved smartphone-based intervention in which the barriers and facilitators for the usage of mHealth for pain management by the elderly were identified.This review also affirms a research gap in the area of testing and evaluation of the exponentially growing mobile apps for pain management, involving the elderly population.
Another integrative review was conducted in a study (Dunham et al., 2021) that objectively analyzed the quantitative and qualitative research that is currently available regarding the advantages of smartphone applications for the treatment of chronic pain in the elderly.The review's outcomes highlight the general dearth of studies supporting older adults who have long-term diseases like chronic pain.
Large-scale research on the unique requirements of elderly adults with chronic pain is lacking.
Hence it is required to elicit the elder specific utility and usability requirements to develop an optimized solution for pain management using mobile technology and integrate the apps into the clinical practice (Bhattarai & Phillips, 2017).Inter-professional collaboration alongside real end users is important in this perspective as well (Nimmanterdwong et al., 2022).
Thus, in the purview of the aforementioned research gaps this research endeavors to: formulate and validate a comprehensive tool for the usability evaluation, in particular for the elderly; mapping of the usability constructs with the usability measures and elderly barriers influencing usability; followed by usability evaluation of pain management mobile applications involving the elderly population

Methodology
This study is exploratory and experimental in nature.Taking into consideration various parameters of this research, mixed method approach was adopted.Mixed method approach involves a blend of quantitative and qualitative techniques which integrate the facts based on both approaches (Collins et al., 2007;Creswell & Creswell, 2017).A systematic literature review-based research approach was adopted to formulate the usability evaluation instrument.The instrument was validated using qualitative and quantitative approaches, i.e., via a questionnaire survey and usability evaluation study (statistical method).The reliability of the instrument was then analysed quantitatively (questionnaire).
The approach followed by this study is shown in Figure 1.The research first explores the elderly's barriers to technology use, the proposed frameworks of usability evaluation, the interface design guidelines for the elderly, and the elderly's usability issues in healthcare mobile applications.
In the second phase, the method for formulating the usability evaluation instrument was determined.Questionnaire was chosen as a mean of usability evaluation instrument for the elderly.
The third phase involved the formulation of the usability evaluation questionnaire.In this context, the usability evaluation attributes, identified via systematic literature review of the pain management mobile applications (Shah & Chiew, 2019) and the systematic literature review of the elderly and healthcare mobile applications (Shah & Chiew, 2017), were grouped into key themes in accordance with their similarities and proximity, termed as usability constructs.The items (questionnaire statements) of the usability constructs were formulated then by taken into account the respective usability constructs, and the usability evaluation attributes and findings, or usability recommendations identified from the SLRs.Moreover, QUIS (Questionnaire for User Interface Satisfaction) (Shneiderman & Norman, 1992) and general guidelines for the interface design specific for the elderly (Aging & Medicine, 2001;Silva et al., 2015) were also considered.The measurement method was designed in the form of closed-end questions with a range of choices using either Likert or semantic differential scale, as it would be simpler for respondents to record their opinion/perception towards the usability of mHealth applications (Brinkman, 2009;de Vet et al., 2011).
In the fourth phase, the proposed usability evaluation questionnaire was validated using quantitative approaches, i.e., via a questionnaire survey (content validity with domain experts and face validity with non-domain experts and elderly participants), and involved elderly pain sufferers to evaluate the selected commercially available pain management mobile applications (statistical method).The cumulative responses of the domain experts, non-domain experts, and the elderly participants provided the assessment of the proposed usability evaluation instrument, reflected their agreement on the comprehensiveness, appropriateness, and relevancy of the items to the constructs.The suggestions of experts are incorporated in the finalized version of the usability evaluation instrument The fifth phase involved mapping of the usability constructs (based on the concept and definition) with the usability measures of ISO/IEC 25010 and the elderly barriers that influence usability hinder the elderly's technology use.Each individual item under all the usability constructs was treated separately in the mapping.
In the sixth phase, the proposed usability evaluation instrument was implemented based on the mapping criteria to calculate numeric values for the usability measures and the barriers (cognitive load, motivational load, perceptual load, and physical load) in technology use, for healthcare mobile applications with respect to the elderly population.
Lastly, in the seventh phase, usability evaluation was performed involving the elderly pain sufferers.The elderly participants evaluated pain management mobile applications via the usability evaluation instrument.Scores were assigned to each item of the questionnaire as per the participant's response.The mean values for the respective entities were then calculated and the evaluation results visualised.Post evaluation session face validity of the proposed usability evaluation instrument was performed with the participants and a quantitative evaluation of reliability was also performed using Cronbach's alpha coefficient, a.

Formatting the elderly's usability evaluation instrument
Evaluation in this research used a questionnaire, which is a common tool for evaluating a product by HCI (Human Computer Interaction) practitioners (Bark et al., 2006).

Conceptualization
In this phase, the constructs intended to be measured are identified and elaborated as items of the questionnaire (de Vet et al., 2011).The usability evaluation attributes used in the studies, identified via a systematic literature review of the pain management mobile applications and the systematic literature review of the elderly and healthcare mobile applications (Shah & Chiew, 2017, 2019) were grouped as per similarity and proximity in their meaning under main themes, termed as the usability constructs in this research.Eight usability constructs are proposed: readable design/typography, text writing style, layout, navigation, learn and use, memory load and mental effort, user guidance and error handling, and cues, instructions, and user control.The definitions of these constructs and grouping of the usability evaluation attributes identified from the literature are shown in Table 2.
For this research, closed-end questions were selected with a set of options as it will be easier for the elderly respondents to respond while assessing the usability of mHealth applications.The items of the usability evaluation instrument were developed by keeping in view the usability constructs and usability evaluation attributes shown in Table 1.Findings or usability recommendations from the SLRs (Systematic Review of Literature) (Shah & Chiew, 2017, 2019) were also taken into account to formulate the items of the usability evaluation instrument.QUIS (Questionnaire for User Interface Satisfaction) (Shneiderman & Norman, 1992) and general guidelines for the interface design specific for the elderly (National Institute on Aging & National Library of Medicine, 2001;Silva et al., 2015) were also considered.
A point to note is that no comprehensive guidelines exist in literature about learning, memory load, user guidance, error handling, cues, instructions, and user control as per the elderly population.The proposed comprehensive elderlyspecific usability evaluation instrument is either helpful for the mHealth app designers and developers, or for the usability experts to perform a heuristic evaluation.Moreover, it is simple enough to be assessed by the elderly in the usability studies.

Measurement method
In this research, the approach of using attitude scales (de Vet et al., 2011) has been adopted to evaluate the opinion/perception of the elderly participants towards the usability of the pain management mobile applications.Each item was provided with a set of fixed choice answers using either Likert or semantic differential scale.The complete questionnaire (after validation by the experts) is shown in Appendix A.

Validation of the usability evaluation instrument
Five domain experts and three non-domain experts were invited by email to take part in the expert review.Three domain experts (usability) and two non-domain experts (linguistics and medical) accepted the invitation.Generally, responses from two or three experts are acceptable for validation purpose (Lam et al., 2018).The domain experts evaluated the comprehensiveness of the identified constructs and items, the relevancy of the items to the constructs, and the appropriateness of the usability evaluation instrument with respect to elderly people.
A 5-point Likert scale was used to obtain the experts' opinion, i.e., strongly agree, agree, neither agree nor disagree, disagree, and strongly disagree.The experts were also required to identify areas of weakness by giving recommendations on ways of improving the instrument to ensure transparency and comprehensibility (Lynn, 1986).The Navigation It is the way to proceed within large or complex information space, i.e., to move within the elements of a screen, forward and back, and from one screen to another (Sikorski, 2002).
Navigation, user interaction, controllable, easy to use/user-friendliness/ease of data input/ cumbersome to use, functionalities (submit, clear, cursor movement), issues with the interface, self-explanatory menu structures, menu labelling, recognition of touch sensitive areas, large size of operating elements, display size 5.
Learn and use It is a measure of how quickly and effectively an application can be understood to use by the user without prior knowledge or at the first time of use (Foraker Labs, 2012).
Learnability, knowledge, ease of learning, quick learn to use, need to learn, ability to use the application without prior knowledge, fast achievement of a first feeling of success, requirement of technical support for system usage, intervention required 6.
Memory load and mental effort The amount of effort involved with thinking and reasoning, that might overlap with other thought processes, i.e., how much a user has to keep in mind while performing a task (Foraker Labs, 2012).
Memory retrieval (ability to accurately recall the answer), complexity, workload demand, minimalist display, simple comprehensibility and interpretability of displayed images and depictions, recognisability 7.
User guidance and error handling It is the way to decrease the possibility of occurrence of mistakes by assisting users in providing them ease to complete the task; or by guiding the users for the management of error situation by presenting some output in response to users' input (Sikorski, 2002).
Error/error prevention/system protection against making errors, responsiveness of the screens to touch, helpfulness or usefulness of information, help, help documentation and support, communication and support, feedback 8.

Cues, instructions and user control
It is the degree to which an application enables the users to acquire additional guidance via information available within the application regarding various features and their step-by-step usage so that the disabled users or the users with the widest possible range of skills and attributes could make better control and use of the application (Association, 2005).
Alarm and notifications setting, controllable, additional notes, information to support device familiarity (e.g., dragging finger instead of tapping and holding device surface) questionnaire for the assessment of the proposed usability evaluation instrument for elderly people is given in Appendix B. The relevancy of the items to the constructs was validated on a 4-point ordinal scale, i.e., not relevant, somewhat relevant, quite relevant, and highly relevant (Chiwaridzo et al., 2017;Davis, 1992;Lam et al., 2018;Vrbnjak et al., 2017).Moreover, the items were also validated for their clear meaning and understandability.The experts were also provided with the option to share their suggestions for improvement, addition, or deletion of any items.Furthermore, the subjective assessment was performed as per Oluwatayo (2012) and Wynd and Schaefer (2002) to evaluate the overall comprehensiveness, appropriateness, and understandability of the proposed usability evaluation instrument with the non-domain experts (medical doctor and linguistic) and the elderly people (participated in the usability evaluation).They were also asked to make suggestions for improving the instrument.
The validation process is depicted in Figure 2. The questionnaire is attached in Appendix C. For the inter-rater agreement, the simplest degree of measurement, i.e., percentage as specified in Bowling ( 2014) has been used.
The demographic details of the experts (Appendix D), socio-demographic details of the elderly participants (Appendix E), validation result of the comprehensiveness and appropriateness of the usability constructs and items of the usability evaluation instrument, and relevancy of the items to the constructs are attached in Appendices F to G respectively.Result of subjective assessment of the proposed usability evaluation instrument by non-domain experts and elderly participants is shown in Appendix H.
The accumulative responses of the domain experts, nondomain experts, and the elderly participants provided the assessment of the proposed usability evaluation instrument, reflected their agreement on the comprehensiveness, appropriateness, and relevancy of the items to the constructs.The suggestions of experts are incorporated in the finalized version of the usability evaluation instrument.The accessibility aspects (as highlighted by one of the experts) were also considered while the mapping of the usability constructs with the usability measures and elderly barriers influencing usability.

Reliability of the usability evaluation instrument
Usability evaluation of existing pain management mobile applications was also performed by involving real-time elderly pain app users to evaluate the reliability of the proposed usability evaluation instrument.The quantitative evaluation of reliability was performed using a famous reliability statistic Cronbach's alpha coefficient, a (Cronbach, 1951).For this purpose, sixteen elderly participants 65 years old or older, diagnosed with chronic pain, understand English and have literacy in smartphone usage were involved to perform the usability evaluation of existing pain management mobile applications.Participants were excluded if they have active mental health problems, e.g., depression or anxiety or have a cognitive and visual impairment.Patients were also provided with the participant information sheet and the consent form to proceed with the usability evaluation study.
On the other hand, commercially available pain management mobile applications were screened and selected from Google Play Store and Apple's App Store based on a specified time frame, keywords, and inclusion/exclusion criteria.The search was conducted in May 2017 that was revised and updated in January 2019.The keywords involved in the electronic search were: pain, pain manage, pain management, pain track, pain log, pain diary, pain record, pain scale, pain severity, and pain journal.From the search results, an app was excluded if the app was: (a) paid; (b) outdated i.e., not updated for 2 years or more; (c) for general health record; (d) for any specific pain type (such as headache, migraine, back pain, fibromyalgia); (e) specifically, for kids; (f) designed for the health care professional only; (g) for pain management program only, i.e., describing any techniques for pain management, e.g., guided image therapy, pain-relieving breath, hypnotherapy, exercise package; (h) only based on pain information or based on pain management reference resource only; (i) for the introduction of any clinic or institute or academy; (j)for a private study only.
Twenty-one apps were identified from Google Play Store; the apps were sorted in descending order of the number of downloads.After removing the outdated, paid and duplicate versions of the apps, 11 unique apps were identified.34 apps were identified from iTunes Store; the apps were sorted in descending order of customer rating.After removing the outdated, paid and duplicate versions of the apps, 16 unique apps were identified for further analysis.
As our study was intended to perform usability evaluation via the proposed instrument and also a comparative study among the four selected commercially available pain management mobile applications (i.e., two apps from iTunes Store and two apps from Google Play Store), the criteria to recruit 16 ± 4 participants were adopted to serve the purpose, as per literature (Alroobaea & Mayhew, 2014;Six & Macefield, 2016).There were 12 participants for the Android apps and 4 participants for the iPhone apps.
To narrow down the scope, the top five apps were first selected from each platform's unique app list, based on the number of downloads and customer rating/stars.The apps were further screened with respect to their functionality as per the conceptual utility model proposed by Shah & Chiew (2019b) and pain data input methods discussed in Shah & Chiew (2019a) to select two apps from each platform.In this context, Manage My Pain Lite and Pain Companion were identified from Google Play Store; and Pain Diary and Community CatchMyPain and Chronic Pain Tracker Lite were identified from App Store, for further assessment with the proposed usability evaluation instrument.
The usability evaluation study was preceded by a demo session where participants were presented with the selected pain apps and provided with hands-on training on the use of these apps.The mobile applications were preinstalled on an Android or iPhone devices of the researcher and each participant evaluated both apps depending on the platform (operating system) of the participant's personal smartphone.
As the quantitative evaluation of reliability was performed using Cronbach's alpha coefficient, a few items with zero variance were removed from the scale in the reliability tests.The test result shows that the coefficient values were greater than 0.7 (0.948, 0.898, 0.875, 0.906) for each pain management mobile application respectively.It indicates a high level of internal consistency and ensures that the evaluation results are reliable with an acceptable value of Cronbach's alpha.
The internal consistency of the usability constructs was also determined via the accumulative evaluation results of the four pain management mobile applications used in the usability evaluation.Results indicated an acceptable value of Cronbach's alpha coefficient for each usability construct, i.e., readable design (0.945), test writing style (0.908), layout (0.941), navigation (0.658), learn and use (0.847), memory load or mental effort (0.844), user guidance and error handling (0.610) and cues, instructions and user control (0.550).The a values of five usability constructs are above 0.7, while three usability constructs are below 0.7, nevertheless, the a value of 0.5 and above is sufficiently good to measure attitude.Thus, the a values of all usability constructs are acceptable (Tuckman & Harper, 2012).All the items also seemed retainable, as deleting any result substantially decreases the a (indicating the importance of those items), or a slight increase in a or has no effect on the value of a.

Mapping of the usability constructs with the usability measures and elderly barriers influencing usability
Usability constructs of the proposed usability evaluation instrument are mapped with the six ISO/IEC 25010 usability measures (appropriateness recognisability, user interface aesthetics, operability, learnability, user error protection, accessibility) and the elderly barriers (cognition, motivation, perception, and physical) that impact usability and hinder the use of technology by the elderly.The mapping was performed based on the definition and concept of the entities.Since WCAG (Web Content Accessibility Guidelines) is known as a mobile accessibility assessment standard, its guidelines were also used to map the items with the usability measure "accessibility."Each individual item of the proposed tool is independently handled in the mapping.
Barriers are the restricted capabilities or proficiency that could affect usability measures with respect to the constructs.This mapping would help to observe the usability measures impacted by the elderly barriers.In relation to text-writing (usability construct), for instance, motivation (barrier) can impact learnability (usability measure).Quantifying the items of each construct will help to calculate the numerical values of cognitive load, motivational load, perceptual load, and physical load corresponding to the use of an application.
Mapping of the usability constructs and items with the ISO/IEC 25010 usability measures and the elderly barriers influencing usability is shown in Table 3.

Implementation of the elderly's usability evaluation instrument
The elderly's usability evaluation instrument was implemented based on the scenario as shown in Figure 3.
The usability evaluation instrument is comprised of eight usability constructs.There are multiple items "i" in each usability construct "UC." Table 4 shows the usability constructs and the related items "q."Each item has a fixed set of options and the options were allocated with a weightage for computation.Table 5 displays the various response options and their numerical values.
A score was given to each usability construct based on the mean weighing value of all the related items.Thus, the score of a usability construct "UC x " is defined as: Each item was mapped with one or multiple usability measures and the elderly barriers to calculate numeric values of the six usability measures and the four elderly barriers in technology.
The proposed usability assessment instrument comprises a total of 58 items, which can be represented as a set The usability measures could likewise be described as set UM ¼ UM 1 , UM 2 , UM 3 , UM 4 , UM 5 , UM 6 , where UM 1 ¼ Appropriateness recognisability, UM 2 ¼ User interface aesthetics, UM 3 ¼ Operability, UM 4 ¼ Learnability, UM 5 ¼ User Error Protection and UM 6 ¼ Accessibility.
Thus, UM x will be defined as: total number of mapped in ½ For instance, appropriateness recognisability UM 1 will be calculated as follows: Similarly, the elderly barriers in technology use could also be represented as set EB ¼ EB 1 , EB 2 , EB 3 , EB 4 , where EB 1 ¼ Cognition, EB 2 ¼ Motivation, EB 3 Physical, EB 4 ¼ Perception.
Therefore, EB x will be defined as: total number of mapped in ½ For example, cognition EB 1 will be calculated as follows: The mean score s calculated for each UC x , UM x and EB x were interpreted as very poor (1.0 < ¼ s < ¼ 1.8), poor (1.8 < s < ¼ 2.6), average (2.6 < s < ¼ 3.4), good (3.4 < s < ¼ 4.2) and excellent (4.2 < s < ¼ 5.0).The overall usability score U of the mobile application is calculated via the mean UM x scores of the six usability measures and the overall usability level will be assigned as per the defined range of scores.
The assessment results of the proposed usability evaluation tool are visualized as graphs and tables.A box plot graph first presents an overview, followed by a tabular view that shows the overall mean scores and corresponding level of each entity.A sample of graphical and tabular representation is shown in Figures 4 and 5, respectively.
A detailed report concerning a specific entity from the tabular view could be observed by clicking on the link given in the last column.A graphical summary is represented in the form of a bar graph.
Figure 6 shows a sample bar chart of the entity "cognition."Additional information on the scores and level of the corresponding usability constructs and items could also be seen in tabular form.Figure 7 shows the table representing one of the barriers "cognitive load."

Usability evaluation results of the pain management mobile applications
Usability evaluation was conducted to assess the selected pain management mobile applications involving elderly patients.The quantified scores of the pain management mobile applications, evaluated using the usability evaluation instrument, are summarized in Tables 6-8 in relevance to the elderly barriers influencing usability, usability measures, and the usability constructs respectively.
The mean scores in Table 6 represent less cognitive, motivational, physical, and perceptual load for the elderly as almost all scores were interpreted as "Good."The scores indicate the elderly's barrier in using the apps.The higher score represents a reduced respective load.Thus, among the Android apps, Manage My Pain Lite shows slightly less cognitive and motivational load compared to Pain Companion.In the meantime, the physical load appears to be the same in both apps and the perceptual load appears to be less in Pain Companion than in the other app.On the other hand, among the iPhone apps, Chronic Pain Tracker Lite has a relatively reduced cognitive, motivational, and perceptual load than Pain Diary and Community CatchMyPain.On the contrary, Pain Diary and Community CatchMyPain are   slightly better in terms of the physical load than Chronic Pain Tracker Lite.Similarly, comparison in terms of usability measures and usability constructs between different apps can be done based on the respective scores.Further breakdown of the evaluation results will help to identify the areas of improvement in relevance to the elderly population.The detailed results as shown in Figure 7 help to identify specific items that are well-accepted by the users, or need further improvement despite the overall usability barrier, measure, or construct scores being excellent or good.In this way, the implementation of the usability evaluation instrument enables viewing the detailed report for each entity under the elderly barriers, usability measures, and usability constructs, with respect to a particular app and identifying the areas for improvement.

Discussion
The proposed instrument is comprised of usability constructs and items from the elderly's perspective for the usability evaluation of healthcare mobile applications, in particular pain management mobile applications.In the context of medical information and in relevance to elderly users, text display or presentation (i.e., the typography and layout of text), simplicity and intuitiveness of the text to be grasped easily by the individual, and the text's effectiveness are three important areas of consideration (Hartley, 1999).To overcome the issues that arise with age-related transitions, there are effective means of presenting the text (Czaja & Sharit, 1998).Thus, healthcare mobile applications should be more accessible by adopting the  instructions and recommendations specific to the elderly (Hartley, 1999).
Similarly, simple and understandable design and layout and increasing the ease of navigation are also crucial to provide a usable application to the elderly (Arnhold et al., 2014).Accessibility, which evaluates the application with respect to the individuals with reduced skills (Billi et al., 2010), is generally a less examined usability characteristic (Petrie & Kheir, 2007).Nonetheless, certain accessibility concerns may impact non-disabled users as well (Petrie & Kheir, 2007).Although evaluation of usability does not imply that everyone's issues are addressed (Billi et al., 2010), yet making technological solutions for healthcare needs accessible to the elderly is still crucial in an aging society where the elderly are the majority of pain sufferers (Holzinger et al., 2011).
Furthermore, learning innovative approaches and strategies is another challenge owing to cognitive limitations in the elderly.Providing easy-to-learn pain self-management mobile applications reduces cognitive stress and increases technology acceptance among elderly consumers.User guidance and error prevention is also essential aspect of usability concerning elderly users.For instance, notifying consumers before they make any deletion, allows them to ensure their action and contributes to a low error rate and increased usability (Norman, 2013).
The proposed instrument was also validated by the domain experts from the perspective of the elderly.The experts agreed on the comprehensiveness of the identified constructs and items, the relevancy of the items to the constructs, and the appropriateness of the usability evaluation instrument.The effectiveness of the instrument's text was also evaluated by the medical doctor, language expert, and elderly participants.The validation indicates the applicability of the proposed instrument for the effective assessment of the usability of healthcare mobile applications from the elderly's perspective.
This usability evaluation instrument can be used by the elderly working with the usability experts or mobile application developers to produce or enhance the mHealth apps in accordance with their needs.Thus, this instrument is a remedy for the lack of the elderly population's involvement in the design and development of pain management mobile applications although pain occurs mostly in the elderly.The participation of the elderly population in the usability evaluation studies can help to improve the elderly's motivation and acceptance of pain management mobile applications.The reliability test also indicates that the proposed instrument is reliable for usability evaluation in the elderly's context.
From the evaluation results with the proposed instrument and real users, it can be concluded that the common usability issues faced by the elderly users with pain management mobile apps are the lack of guidance, instructions, and cues to guide the users in using the apps.This in turn has resulted in issues in learning and operating the apps.Researchers and app designers can leverage on the proposed instrument to eliminate these potential usability issues from the app.Clinical practitioners can rely on the instrument to filter apps fulfilling desirable usability measures before recommending the apps to their patients.Moreover, measurement of the items, via usability evaluation instrument, also contributes to the quantification of usability measures as well as the elderly barriers of the mHealth application.The visualization of assessment results helps to identify the areas of improvement.
Though efforts have been made to maintain the rigor of the study, yet there were certain shortcomings to acknowledge.One of the possible limitations is that the formulation of instrument was guided by the SLRs with scope restricted to five databases (although relevant literature was identified from snowball sampling) and only patient-oriented pain management mobile applications were addressed.Moreover, the researchers faced difficulty in getting cooperation from the elderly to participate in the study.It affected the sample size.Thus, the sample size for the usability evaluation study was limited to only sixteen old age participants.Nevertheless, valuable data was derived from the usability evaluation study.
The percentage of less literate participants in the usability evaluation study was low.Thus, their opinions in using pain management apps might be under represented.Also, participants evaluated the pain apps relevant to the platform as per their owned smartphones.The four selected apps were platform-specific and not cross-platform.Hence, the users' perspective may be influenced by their smartphone platforms and may introduce bias.
There are some potential future research goals for overcoming aforementioned limitations.Applications that support healthcare providers in their clinical decisions could also be considered for review in future to obtain more insights about pain management, as pain management is a two-way process between patients and healthcare providers.Future research may also replicate the study protocol by including more old age respondents specifically with lower educational level or illiterate elderly and report the results of the usefulness and practicality of the proposed usability evaluation instrument.Also, usability evaluation study may generate more insightful results by involving user evaluating cross-platform apps.
Although the formulation of instrument was guided by the SLRs that were focused on pain management mobile applications, the proposed instrument could be helpful in the usability evaluation of other mHealth apps from the elderly's perspective.Future research is required to evaluate various mHealth apps other than pain management from the elderly's perspective via the proposed usability evaluation instrument.

Conclusion
In this research, a usability evaluation instrument that maps the usability measures with the elderly's barriers is proposed.The content of the instrument was verified by five experts (usability, linguistics, and medical).A usability evaluation session, followed by face validity, was performed with sixteen elderly participants on the pre-selected pain management mobile applications.In this research, the reliability of the proposed usability evaluation instrument was tested and the identification of usability issues of the pre-selected applications was demonstrated.The instrument also provides graphical visualization of the mapping of usability measures with the elderly's barriers.The graphical details help in identifying the areas of usability improvement and the scores helps in comparing usability of different applications.
The usability assessment instrument serves as a comprehensive basis for evaluating healthcare mobile applications with respect to the elderly population.It will be significant for the professionals to involve real users in the usability evaluation of pain management mobile applications and modify or develop the app in accordance with the instrument's highlighted areas of improvement, or to assure usability of the app.The instrument also helps clinicians to choose among the existing pain management apps and recommend the most usable app to elderly patients as a selfmanagement regime.

Satisfaction"
How pleasant is it to use the design?"-Effectiveness "Visual support of task goals" of control users have to choose and decide how to proceed."

Figure 1 .
Figure 1.Process flow of the formulation of usability evaluation instrument for the elderly.
For elderly participants' recruitment, ethical approval was granted by the Medical Research Ethics Committee of UMMC (University of Malaya Medical Centre) (MREC ID NO. 201881-6552).The medical doctors of the Department of Primary Care Medicine, rehabilitation physician of the Advanced Pain Care and Rehabilitation center, Department of Rehabilitation Medicine, and consultant neurosurgeon, Pediatric Neurosurgery and Interventional Pain Management, Department of Surgery were approached to seek their permission and assistance to recruit patients.

Figure 3 .
Figure 3. Implementation model of the usability evaluation instrument for the elderly.

Figure 6 .
Figure 6.Detailed report of cognitive load for Manage My Pain Liteoverview in form of bar chart.

Figure 7 .
Figure 7. Detailed report of cognitive load for Manage My Pain Litetabular view.

Table 1 .
Summary of the usability evaluation measures in the usability frameworks.

Table 2 .
Usability constructs and grouping of the usability evaluation attributes identified from literature.

Table 3 .
Mapping of the usability constructs with the usability measures and the elderly barriers in the use of technology.

Table 3 .
Continued.Multiple ways to access the information (e.g., table of contents, list of links to all the related subpages, search function etc.)

Table 6 .
Mean scores and corresponding levels of the elderly barriers influencing usability.

Table 7 .
Mean scores and corresponding levels of the usability measures.

Table 8 .
Mean scores and corresponding levels of the usability constructs.