Home telemonitoring for heart failure and COPD patients: Evaluation of the usability and acceptance of the cVitals application

(1)

Home telemonitoring for heart

failure and COPD patients

Evaluation of the usability and acceptance of

the cVitals application

(2)

(3)

2

Home telemonitoring for

heart failure and COPD patients

Evaluation of the usability and acceptance of the cVitals application

Master Dissertation Medical Informatics

Student: Hester Albers Studentnumber: 10980717 Email: h.albers@amc.uva.nl Mentor: Gaby Wildenbos

Academisch Medisch Centrum (AMC) Amsterdam Department of Medical Informatics

Email: g.a.wildenbos@amc.uva.nl

Tutor:

Linda Dusseljee-Peute

Email: l.w.peute@amc.uva.nl Independent expert: Martine Breteler Focuscura Email: m.breteler@focuscura.nl Independent expert: Monique Jaspers

Email: m.w.jaspers@amc.uva.nl

SRP address:

Focuscura Odijkerweg 1

3972 NE Driebergen Rijsenberg

Duration: May 2016 - January 2017 Date of submission: 16 January 2017

(4)

3

Acknowledgements

I would like to express my gratitude to all who helped me finish this thesis. Without the guidance of all my supervisors, help of all FocusCura employees and the support of my family and friends, I would never be able to complete this thesis.

In particular I would like to thank my supervisors at the AMC. Linda Dusseljee-Peute for her guidance, advices and comfort in times when I most needed it. You helped me to structure my thoughts and motivated me to go on. Gaby Wildenbos for her support and quick answering during this study. Monique Jaspers for her ideas and support at the beginning of this study. Furthermore, I would like to thank my supervisor from FocusCura Martine Breteler, whose explanations, revisions and comments really helped this study along.

Moreover, I would like to thank all the involved FocusCura employees in helping me making this study possible. Without their help I would never been able to incorporate all these patients in the evaluation.

Evaluation of cVitals depended largely on the HF and COPD patients of the cooperating care organisation. A big thanks to all who took the time to participate in the usability or acceptance evaluation.

(7)

6

Summary

Background: Home telemonitoring (HTM) for the chronically ill is increasingly used as a

solution to the growing problem of an aging population and its upcoming increase in chronic diseases and rising healthcare expenses. Various forms of HTM applications are developed to promote self-management for people with chronic diseases. However, these systems are not yet widely used by the chronically ill. Research has shown that there are several barriers that affect the acceptance (intention to use) by this mainly elderly audience, namely: computer anxiety, low expectations of usability and perceived safety. Additionally, older users often experience difficulties using monitoring systems. Therefore, it is important that the acceptance and usability of existing telemonitoring systems are evaluated to enhance and promote the use of these systems. This study focuses on the evaluation of cVitals, a HTM application aimed to support Chronic Obstructive Pulmonary Disease (COPD) and Heartfailure (HF) patients. The first aim of this study is to evaluate which usability issues COPD and HF patients encounter using cVitals. The second objective is to evaluate which factors affect the acceptance of cVitals by COPD and HF patients. Furthermore, associations between the usability and acceptance of cVitals are described.

Methods A first usability analysis of cVitals was carried out by interviewing the designer of

cVitals by use of the Healthcare Information Management and Systems Society (HIMSS) expert analysis. Next a think-aloud study with 10 COPD and HF patients was performed. The usability issues and suggestions for improvements by users were clustered among the HIMSS usability guidelines and were classified with a severity rating. When an issue or suggestion could not be clustered amongst the HIMSS guidelines, a new cluster was described. To evaluate the factors that influenced the acceptance of cVitals, a questionnaire was made based on the extended Unified Theory of Acceptance and Use of Technology (UTAUT) model. Literature was used to construct the extended version of the UTAUT, which is especially aimed at acceptance of health information systems by older adults. The extended UTAUT questionnaire was send out twice to HF and COPD patients: when they just started using cVitals and at 3 months follow-up. Analysis was carried out using a principal component analysis.

Results The usability evaluation identified 51 usability issues and 14 suggestions for

improvement. Primarily problems were clustered onto ‘no knowledge of functionalities’ and

‘effective information presentation’. These issues were mostly related to the cognitive and

physical barriers of older users. Patients reported high satisfaction with the system. The acceptance evaluation showed that intention to use cVitals by new users is influenced by the following factors: hedonic motivation (0,46), performance expectancy (0,37), effort expectancy (0,26), and facilitating conditions (0,23). Among these factors, hedonic motivation exerted the strongest effect. A significant difference between the first and follow-up measurement was found in the factor habit.

Conclusion The think-aloud method was shown to be an effective usability method for

identifying usability issues of varying severity. Additionally, this study confirms the value of the extended UTAUT model in identifying factors influencing acceptance in the context of HTM applications. The study has implications for the development of cVitals and other HTM applications. Knowledge gained from this study is both beneficial to designers of HTM applications and researchers evaluating these systems.

(8)

7

Samenvatting

Achtergrond Telemonitoring vanuit huis voor chronisch zieken wordt steeds vaker ingezet als

oplossing voor het groeiende probleem van de vergrijzing en de daarbij komende toename van chronisch zieken en de oplopende zorgkosten. Verschillende vormen van telemonitoring worden ontwikkeld om zelfmanagement van chronisch zieken te bevorderen. Echter worden deze systemen nog niet veel gebruikt door chronisch zieken. Uit onderzoek is gebleken dat er meerdere barrières zijn die invloed hebben op de acceptatie bij deze voornamelijk oudere doelgroep, namelijk: moeilijkheden in gebruik, computer angst, lage verwachting van bruikbaarheid en waargenomen veiligheid. Hierom is het van belang dat huidige telemonitoring systemen worden geëvalueerd op acceptatie en bruikbaarheid zodat deze doorontwikkeld kunnen worden om het gebruik van systemen te stimuleren. Dit onderzoek richt zich op de evaluatie van cVitals, een telemonitoring applicatie die vanuit huis gebruikt wordt. Het eerste doel van dit onderzoek is om te evalueren welke problemen COPD en HF patienten ervaren bij het gebruik van cVitals. Het tweede doel is om te evalueren welke factoren van invloed zijn op de acceptatie van cVitals. Tevens word de relatie tussen de gebruiksvriendelijkheid en acceptatie van cVitals beschreven.

Methoden Evaluatie van de problemen bij gebruik van cVitals werd in eerste instantie

uitgevoerd middels een interview met de designer van cVitals. Hierbij werd geëvalueerd of de designer de Healthcare Information Management and Systems Society (HIMSS) richtlijnen voor gebruiksvriendelijkheid gebruikte. Hierna is er ‘think-aloud’ studie met 10 COPD en Hartfalen (HF) patiënten uitgevoerd. De problemen bij gebruik en suggesties voor verbeteringen die door gebruikers werden geclassificeerd middels de HIMSS richtlijnen en een ‘severity rating’. Als er een probleem of een suggestie niet kon worden geclassificeerd onder de HIMSS richtlijnen, werd er een nieuw cluster beschreven. Om de factoren die van invloed zijn op de acceptatie van cVitals te identificeren werd een vragenlijst gemaakt op basis van het verlengde Unified Theory of Acceptance and Use of Technology (UTAUT) model. Literatuur werd gebruikt om de verlengde versie van de UTAUT te maken, welke special gericht is op de acceptatie van zorgsystemen die door oudere volwassenen worden gebruikt. Deze vragenlijst is twee keer afgenomen: wanneer HF en COPD patienten net begonnen waren met het gebruik van cVitals (T0) en na 3 maanden opnieuw (T1). Een principale-componenten analyse is uitgevoerd om de data te analyseren.

Resultaten De evaluatie van de gebruiksvriendelijkheid liet 51 gebruikersproblemen en 14

suggesties voor verbetering zien. De classificaties met de meeste gebruikersproblemen waren 'geen kennis van functionaliteiten' en 'effectieve informatie-presentatie'. Deze problemen waren meestal gerelateerd aan de cognitieve en fysieke barrières van oudere gebruikers. Een hoge tevredenheid met het systeem werd gemeten. Uit de acceptatie evaluatie bleek dat de intentie om cVitals te gebruiken door nieuwe gebruikers wordt beïnvloed door verschillende factoren, namelijk: drijfveren (0,46), prestatie verwachting (0,37), inspanningsverwachting (0,26) en beschikbare faciliteiten (0,23). Drijfveren blijkt het sterkste effect op acceptatie te hebben. Een significant verschil tussen de scores van T0 en T1 is gevonden op de factor gewoonte.

Conclusie Deze studie toont aan dat de ‘think aloud’ methode een effectieve methode is voor

het identificeren van gebruikersproblemen. Tevens bevestigt deze studie de waarde van het verlengde UTAUT model om de factoren die van invloed zijn op acceptatie op het gebied van telemonitoring systemen te identificeren. Deze studie heeft gevolgen voor de ontwikkeling van

(9)

8 cVitals en andere telemonitoring systemen. Opgedane kennis uit dit onderzoek is zowel gunstig voor ontwerpers van HTM applicaties en onderzoekers die deze systemen evalueren.

(10)

9

1. Introduction

In the Netherlands, half of the population nowadays lives with a chronic disease (1). Prevalence of chronic diseases increases with age. Chronic Obstructive Pulmonary Disease (COPD) and Heart Failure (HF) are two chronic diseases which are highly prevalent in the Netherlands. It is estimated that around 142.000 people live with heart failure and 596.000 with COPD (2, 3). COPD and HF occurs mainly in older adults (aged above 50). Chronic diseases affect the functional health status and quality of life of a person dramatically. In addition, chronic diseases affect not only the individual, the impact on society is huge as well, especially in healthcare costs. Healthcare costs include indirect and direct costs. Indirect costs include for example loss in labor due to absence and disability. Direct costs include the expenditure on healthcare, in 2006 spending on common chronic diseases was 17% of all healthcare costs. An essential step of a structured care approach for coping with chronic diseases is to support self-management (4). Therefore, the Dutch government described 3 objectives. One of these objectives includes:

“75% of the chronically ill (diabetes, COPD) and frail elderly, who want and are able to do, carry out independent measurements within 5 years, often in combination with data remote monitoring by the caregiver (5)”.

Currently almost half of the population of older adults (50+) and chronically ill are independently measuring certain health values. However, only 1-5% of this population submits these measurements electronically (website or mobile app) to their caregiver (6).

New technologies can play a vital role to empower self-management for chronic patients at home, and telehealth technologies have been growing dramatically. Home telemonitoring (HTM) is a way of addressing the new needs of healthcare in an ageing population. It is defined as “the use of telecommunication technologies to transmit data on patients health status (e.g. questionnaires and vital signs measurements) from home to a healthcare center (7)”. These measurements can be performed automatically and constant or manually by patients.

At the same time mobile technologies are evolving rapidly. Today, our mobile technologies have become almost indispensable. Through our mobile phone and tablets, we make use of several apps that make our lives healthier and easier. We are increasingly using health apps and self-measuring equipment. This mobile revolution is offering an opportunity to provide medical support through monitoring to the chronically ill. In some instances, mobile apps may allow HTM to replace time consuming office visits (8).

Despite the rapid technological evolvement, HTM adoption rates remain low due to the traditionally conservative older users. HTM will only be a successful intervention if it is adapted to the needs and preferences of the end users. Usability and acceptance among users of these technologies are therefore essential. When considering older users acceptance of mHealth, user engagement is especially challenging, since older users are traditionally reluctant to accept innovative solutions as mHealth (9). Barriers of older adults to accept HTM are computer

(11)

10 anxiety, low expectations of the usefulness, perceived security, and difficulties using these systems (10). Therefore, previously developed theories about technology acceptance should be used to study those factors. The Unified Theory of Acceptance and Use of Technology (UTAUT) is considered to be the leading theory in analysis of acceptance. This model is able to analyse the different factors which influence the intention to use and use technologies. Insight in the factors influencing acceptance may help in adapting the system to increase successful implementation of the application.

Ease of use is an essential factor in acceptance of HTM systems (11, 12). Nevertheless, users of HTM systems mostly report on difficulties in use with these systems (7, 13, 14). HTM devices are of limited value when it does not adequately address the usability needs of the target audiences. “App designers and content developers tend to focus on the “more obvious” types of usability and accessibility (i.e., visual, auditory and motor), and often overlook or give little attention to the cognitive accessibility aspects of their content and user interfaces (8)”. Several guidelines exist to support developers in design their apps. The Healthcare Information and Management Systems Society (HIMSS) provides a usability guideline based on 9 principles of usability (15). However, this guideline does not include all aspects mentioned by older adults regarding usability. Therefore, a framework on mHealth for older adults has been developed. It can be used to support usability evaluators and mHealth app designers in understanding the intrinsic aging barriers that are inherent to usability issues encountered by older users (16).

These theories are not yet incorporated in much evaluation studies of HTM. Whereas traditional health interventions are based upon theory and research, HTM services lack research of the effects and acceptance among users (17, 18). Hence this thesis is of importance, since it describes an evaluation of a HTM application case study, supported by a theoretical framework of technology acceptance and usability. The case study that is analysed is “cVitals”, which aims to help health professionals to follow and guide patients with chronic diseases who live on their own in management of their disease. Furthermore, it assists the chronically ill with managing their disease by offering them: a service to send and measure health values, an overview of their health values and information about their disease and lifestyle. The two largest user groups are HF and COPD patients. Both diseases are chronic and often occur in the older population.

1.1 Research objectives

The first aim of this study is to evaluate the usability of the cVitals application by identifying usability problems that HF and COPD patients can encounter, and translate them to specific design recommendations. The second aim of this study is to gain insight into the influencing factors regarding acceptance of the cVitals. Since ease of use is essential in the acceptance of HTM applications, associations between the acceptance and usability results will be described as well.

This study is beneficial for the development, implementation and evaluation of the cVitals and other HTM applications. Stakeholders include the development team, product manager, the end-users and developers and researchers of other HTM applications. The identification of

(12)

11 usability problems can help the development team by adapting cVitals in such a way that it aligns with the user needs. In the end this is beneficial to the end-users because the app is adapted to their needs. These insights are also valuable to developers of other HTM applications since it identifies the needs and insights of older adults as users. Insights into the influencing factors regarding acceptance cVitals can help development teams and researchers in better understanding why and how people use HTM applications and possibly adapt the way HTM applications are implemented and used. The research questions addressed in this study are:

1. What are the usability problems HF and COPD patients may experience using cVitals at home?

a) How does the designer of cVitals (implicitly/explicitly) incorporate the extended HIMSS guidelines in redesigning cVitals?

b) What kind of usability problems, according to the extended HIMSS guidelines and mhealth framework, do HF patients encounter and how severe are these problems regarding the use of cVitals?

c) What kind of usability problems, according to the extended HIMSS guidelines and mhealth framework, do COPD patients encounter and how severe are these problems regarding the use of cVitals?

2. What are the influencing factors regarding acceptance of HF and COPD patients when using CVitals at home?

a) What are influencing factors of the extended UTAUT model on intention to use of cVitals by HF patients?

b) What are influencing factors of the extended UTAUT model on intention to use of cVitals by COPD patients?

1.2 Chapter organisation

Chapter 2 starts with an introduction into the HTM field. This chapter also focuses on cVitals and its main user groups. Chapter 3 describes theories about technology acceptance and usability. Chapter 4 contains the methods of the evaluation of cVitals; chapter 5 contains the results. Chapter 6 presents the discussion of this study including associations between the usability and acceptance of cVitals, strengths and weaknesses, and suggestions for further research. Chapter 7 provides the conclusions.

(13)

12

2. Home telemonitoring for COPD and HF patients

HTM is a new approach which empowers patients to manage their disease (7). HTM includes the use of monitoring devices, wearable sensors and intelligent sensors for the transmission of health data from the patient’s home to a health professional in a care center. Several forms to monitor patients exist: video-consultation, with or without transmission of vital signs, mobile telemonitoring, automated device-based tele monitoring, interactive voice response, and Web-based tele monitoring (19). HTM is a mHealth solution. mHealth is a term which can be interpreted in different ways. This thesis defines mHealth as: “medical and public health practice supported by mobile devices, such as mobile phones, patient monitoring devices, personal digital assistants (PDAs), and other wireless devices (20)”. Mobile HTM is promised to be a solution for delivering cost effective quality care. Benefits of monitoring devices include reducing risk of hospitalisations, healthcare costs, emergency visits, and improving the patients’ health and quality of life for HF and COPD patients (7, 13, 19). HTM for chronic conditions has attracted quite an amount of research over the years both in the form of primary studies and systematic reviews. On the other hand, this is also associated with an increase in contradictory findings in terms of HTM effects (19). Additionally, the benefits from wider diffusion and use of HTM have not been fully achieved yet. HTM is expected to be an integral part of healthcare within the upcoming years. However, the funding of healthcare is complex and fragmented. This makes it more difficult to implement and finance new technologies on a structural basis. New technologies are often funded for a shorter period as pilots, by the healthcare organisation itself or by private funds (21). Though, advances are made and HTM technologies are getting more structurally financed by healthcare insurance companies. An example of this is cVitals, a HTM application of FocusCura, which is one of the first services which will be structurally financed by a health insurance.

OVERVIEW

This chapter contains relevant background information on home telemonitoring and the evaluated application cVitals. The emphasis lies on applications for older COPD and HF patients. It also gives an

(14)

13

2.1 cVitals

2.1.1 Background

cVitals is a HTM application which aims to empower patients by letting them participate in their own health process. It helps patients to manage their disease and to communicate with their health professional. In turn, it helps health professionals to follow and guide patients with chronic diseases who live individually in management of their disease. Patients can perform measurements at home, for example blood pressure, and send it to their health professional. Agreement on fixed measurement time points is reached before a patient starts with the cVitals. Health professionals can follow the patient’s results and trends of the measurements. The health professional receives an (red) alarm when a measurement is above or below the threshold. An (orange) alarm is generated when a patient did not sent their measurements on time. Subsequently, the health professional can decide to contact the patient. Patients receive an email when they failed to send their measurements on time. See figure 1 for an overview of the inclusion and use of the cVitals. Due to the fact that measurements are performed regularly, earlier detection of severe symptoms by the health professional should be possible. When patients have problems using the app, they can call the service desk of FocusCura.

The interface of cVitals is different for the health professional than for patients. This study will only focus on the interface for the user. The application consist of four main screens: measurements to perform, graphical view on performed measurements, information about a Figure 1 cVitals flow diagram

1. Physician prescribes cVitals and allows for measurements and threshold values 2. Patient uses cVitals

3. For measurement values below or above the threshold, a message is send to the care professional and allows the care professional to have contact with the patient

(15)

14 patient’s disease and a possibility to fill in absence and personal data, see figure 2. A patient completes a questionnaire or specific measurement and send it to his/her health professional on the main screen. Sending measurements is the basic functionality of cVitals, while the functionalities on the other screens can be seen as extra functionalities. The graphical view on performed measurements is a screen which provides a graphical overview of all measurements over time. The screen with information about the disease guides patients to a website with information about the characteristics of the disease. Unnecessary alarms can be generated when patients do not sent their measurements on time. The absence screen can be used to avoid this. When an user is absent for a long time, he/she can fill in the date of absence. This will be automatically communicated to the professional.

Figure 2 Screenshots cVitals native app

A: Main screen with measurements to perform

B: Screen which leads to information about a patient’s disease C: Screen with graphical view on performed measurements D: Screen with personal data and possibility to fill in absence

cVitals is installed by a technician of FocusCura at the home of an user. The technician explains how the application can be used. cVitals is installed on an iPad (native app) or used in a web browser (webapp). This study is aimed at evaluating both versions, since both are used by the end users. No major differences exist between the web- and native app, the biggest difference is that the four main buttons are on the top of the screen instead of the bottom at

(16)

15 the webapp. The native app is only available for IOS. Users receive the iPad from their care organisation, which borrows it from FocusCura. The measuring equipment is not attached, users can buy it themselves, it is given by the care organisation (which borrows it from FocusCura) or they can fill out their measured values manually. Most of the patients make use of the measuring equipment provided by FocusCura. Nowadays, these include the Omron MIT5, which is a blood pressure monitor, and the IHealth lite weight scale, see figure 3. These devices can be connected through Bluetooth with cVitals. When a patient measures their blood pressure or weight, the measurement is automatically sent to cVitals with a Bluetooth connection. Users of the webapp often make use of a Withings blood pressure monitor and weight scale. These devices cannot be connected to cVitals automatically. When users use these devices, a Withings app pops up, users go back to cVitals and the data is automatically imported from the Withings app. Besides blood pressure and weight, the following measurements can be included: pulse, glucose-level, and questionnaires for COPD and heart failure. Although all patients use the same application, different programs exist. COPD patients only complete a COPD questionnaire, while HF patients submit a HF questionnaire, blood pressure and weight.

Figure 3 iHealth weightscale and Omron MIT5 blood pressure monitor

2.1.2 Implementation

cVitals was first implemented in 2013. The development team continuously improves and adapts cVitals based on the newest insights and problems. This varies from small bug fixes to development of a new functionality (such as a freemium model or integration with the hospital’s information system). It is unknown whether these changes of the designers are based upon theory and knowledge about influencing factors of use and acceptance of mHealth. Furthermore, user acceptance and usability of cVitals have not formally been evaluated yet. The product manager, which is lead of the development team, receives comments and ideas about the application from different care organisations. In addition, the product manager visits a client every 2 weeks to gather experiences and ideas. All these ideas and comments are discussed, prioritized and placed on a roadmap. The roadmap is a document where the current adjustments are incorporated and organised by prioritization.

At the end of this study, cVitals had 450 active COPD and HF users of 11 different care organisations. The group of patients with heart failure is the largest user group (60%). The age and gender of these patients is unknown. However, heart failure and COPD mainly occurs in the older population, so most of the patients are older adults. More than 160.000

(17)

16 measurements have been sent to health professionals from the start of implementation until now. The care organisation is owner of this data and they have an agreement with the user. FocusCura has permission to use the data in order to improve the cVitals. However, when the data needs to be analysed and published, permission of users and the care organisation is necessary. For this study permission was given by all involved care organisations.

2.1.3 Users

At the start of this study, two main target groups existed: COPD and HF patients. Both diseases are chronic and usually occur in the elderly population.

COPD is a chronic constriction which is progressive and non-curable. Symptoms of COPD are chronic cough, mucus and shortness of breath. Patients can also have exacerbations. “An exacerbation of COPD is defined as an acute event characterized by a worsening of the patient’s respiratory symptoms that is beyond normal day-o-day variations and leads to a change in medication (22)”. COPD occurs mainly in people aged 55 and older and the prevalence increases with age. The main cause of COPD is smoking behaviour. Other risk factors involve other long diseases, genetics, and environmental factors (air pollution e.g.). To diagnose COPD, a lung-function test, spirometry is necessary. Based on the severity of airflow limitation, a classification of COPD is made. The first stage is mild, the second moderate, third is severe and the fourth and final stage is very severe (22). COPD is a non-curable disease, however treatment can decline the regression of long function. Treatment in the form of lifestyle advices (smoking prevention, increasing physical activity) and pharmacologic therapies (bronchodilators and inhaled corticosteroids) are often used to manage COPD. COPD patients are mostly guided by a specialised long nurse and a general practitioner to manage their disease by reduce symptoms and risks.

HF is a chronic disease which occurs when the heart is unable to pump sufficiently to maintain blood flow to meet the body’s needs. Two forms of heart failure exist: systolic- and diastolic heart failure. Systolic HF happens when the heart does not pump blood out to the body as normal. Diastolic HF occurs when the heart is unable to relax enough, so the left chamber is not able to fill properly during the diastolic phase (23). Symptoms of this condition vary, but can include: shortness of breath, fatigue, swelling, and reduced exercise capacity. The prevalence of heart failure increases with age, most patients are aged above 65. Causes of HF can include: heart attacks, prolonged hypertension, heart valve disorder, and longstanding cardiac arrhythmia. Multiple tests (blood investigation, ECG, Xray, exercise stress test) are used to diagnose patients with HF. Possible treatments are medications, cardiac resynchronisation therapy or a heart transplantation (23). Monitoring is one of the possibilities for guiding patients by health professionals. Based on the severity of the symptoms, a classification is made. Stage 1 reveals no symptoms of HF, stage 2 patients have symptoms by considerable physical effort, stage 3 involves patients who have symptoms by moderate physical effort, and stage 4 includes patients who receive symptoms in relaxation.

HF and COPD are both characterised by instances of a sudden decline which may require hospital admission and results in a lower patient health and quality of life. “These

(18)

17 exacerbations account for a large part of the burden of these diseases and are considered to be partly preventable, or at least reducible, if deterioration could be detected sufficiently early and appropriate treatment provided (24)”. Therefore, HTM for both HF and COPD patients is useful, since a decline of health can be detected in an early stage and thus the appropriate treatment can be provided.

(19)

18

3. Usability and acceptance guidelines and models

3.1 HIMSS Usability guidelines and mHealth framework for

older users

The impact of HTM on healthcare is a double faced phenomenon. On the one hand, HTM applications are promising systems, since benefits of monitoring devices include reducing risk of hospitalisations, healthcare costs, emergency visits, the patients’ health and improving quality of life for the chronically ill (7, 13, 19). Although appropriate technology is available, a very limited spreading of remote mobile health monitoring solutions is evidenced, especially among those users that could especially benefit from it, such as older adults or the chronically ill (25). Drop outs of patients in telemonitoring treatments are mainly caused by usability- and technical problems (7). HTM solutions are often not specifically designed for older adults with chronic diseases, especially since young app developers have difficulties in understanding the cognitive and physical barriers of older adults that might influence elderly’s perceived ease of use of mobile health apps (26). In turn, a systematic review (13) and prominent technology acceptance models such as the TAM (27) and the UTAUT (28) indicate that ease of use as one of the most important factors for acceptance and use of technologies. Insight in the usability problems and the underlying causes may also help in adapting mhealth apps to increase usability. Therefore, investigation of ease of use is of importance.

In this study, usability is defined as (29): “The effectiveness, efficiency and satisfaction with which specified users achieve specified goals in particular environments. Whereby these concepts are defined as followed:

 Effectiveness: the accuracy and completeness with which specified users can achieve specified goals in particular environments

 Efficiency: the resources expended in relation to the accuracy and completeness of goals achieved

 Satisfaction: the comfort and acceptability of the work system to its users and other people affected by its use.”

SUMMARY

This chapter gives an introduction in usability and technology acceptance. The emphasis lies on theories and models assessing usability and acceptance of home telemonitoring devices. It also gives a detailed description of the theories on which this study is

(20)

19 As stated above, usability depends on several concepts (effectiveness, efficiency and satisfaction). Therefore, it is not possible to state one measurement which describes whether an application has “good or bad usability”. Instead, usability of an application should be described with multiple indicators. Indicators for effectiveness, efficiency and satisfaction are the percentage of tasks successfully achieved, completion time and satisfaction ratings (30, 31).

Different methods exist to evaluate the usability of systems. In general two main approaches exist: the expert-based and user-based approach. An evaluator which uses an expert-based approach “inspects a user interface with a set of guidelines, heuristics or questions in mind or by performing a step-wise approach, derived from general knowledge about how humans process through tasks (32)”. Expert-based methods are relatively cheap. Nevertheless, they do not assess actual use of the system and should be conducted by evaluators which have firm knowledge about usability (30). A user-based approach evaluates how real end users make use of the system. This approach is a good approach to provide insights into the underlying causes for system usability problems encountered by use (32).

User performance measurements, cognitive workload assessments, satisfaction questionnaires, log-file and keystroke analysis, and participatory evaluations are examples of user-based methods (32). The first four methods are relatively easy to conduct but do not provide insights in all usability problems because the user does not interact during the session with the system. In participatory evaluations, users interact with the system and are asked to explain what they are doing to “think aloud” or in a retrospective interview. The most widely adopted participatory evaluation is the concurrent think-aloud. “Think aloud is a form of observation where the user is asked to talk through what he is doing as he is being observed (30).” This method is selected for this study because it is a well-recognized method and can provide information about underlying causes for usability problems encountered by older users.

Previous studies (24, 33-44) investigated the usability of HTM systems for COPD and/or HF patients. However, several of these studies only evaluate one specific part of usability, i.e. satisfaction. Six studies (36, 38, 39, 41-43) were found where the think aloud method was used for usability evaluation. Overall, these evaluations took place in the pilot phase of a HTM system, while the think aloud method is especially well suited for both the design phase as when the system is already implemented.

When the usability has been evaluated by any method, mapping of the usability problems is necessary. Several guidelines exist to map usability problems revealed by an usability evaluation of a health information technology (HIT). The Healthcare Information and Management Systems Society (HIMSS) provides a usability guideline based on 9 principles of usability (15):

1. Simplicity refers to everything from lack of visual clutter, to inclusion of only functionality.

(21)

20 2. Naturalness refers to how automatically familiar and easy to use the application feels to the user. Associated factors are terminology and how well a screen flow map to the users tasks and expectations.

3. Consistency is divided into internal and external consistency. Internal consistency has to do if concepts/terms are used consistently throughout. External has to do with how much an application structure matches experiences with other applications.

4. Minimizing cognitive load means presenting all the information needed at hand reduces the cognitive load. Transparency is of great importance as well, the use of an application should not create too much thoughts for the user like: How, What, and Where? 5. Efficient interactions means that users can perform tasks efficiently through the

application. The number of steps to complete tasks should be minimized.

6. Forgiveness and feedback means that a design allows the user to discover the application without fear of a disastrous outcome (e.g. not being able to restore an unintended action in the system). Good feedback supports a user by informing about the effects of the actions they perform and informs a user to recover from this.

7. Effective use of language means that the used language should be adapted to the level of the users. The terms should be familiar and meaningful to the user.

8. Effective information presentation consist of 3 components: appropriate density, meaningfull use of color, and readability.

9. Preservation of context means keeping screen changes and visual interruptions to a minimum during completion of a task.

A systematic review (45) showed that there are some aspects which are mentioned by older patients as usability problems which could not be mapped into the basic HIMMS guideline. Wherefore the researchers extended the HIMMS guideline with 6 principles:

10. Less time consumption: time spent by users on process is kept to a minimum and automation is incorporated in the applications where possible.

11. Discrete transactions: the application supports discrete transations; it allows users to postpone or skip processes.

12. Low burden/high motivation: gamification and social media are incorporated to lower the burden of use and increase the willingness to use.

13. Behaviour change stimulation: the application uses functionalities to stimulate behaviour change.

14. Educatory resource provision: the application contains information regarding topics contained in the app.

15. Data security: the application guarantees that users data is not spread without consent and is protected with a password.

Wildenbos et. al. (16) reported that the HIMSS usability guideline is not sufficient for developing a highly usable app for older adults. Therefore a framework on mHealth for older adults has been developed, see figure 4. This framework can be used to support usability

(22)

21 evaluators and mHealth app designers in understanding the intrinsic aging barriers that are inherent to usability issues encountered by older users (16). This framework is especially helpful when usability problems are detected and underlying causes should be analysed. The framework consist of 4 main categories: cognitive barriers, physical barriers, motivational issues, and perception barriers. These categories are divided in age related concepts. These age related concepts are divided in age dependent abilities.

Figure 4 Framework 'mHealth for Older Users' (16)

3.2 The Unified Theory of Acceptance and Use of Technology

(UTAUT) model

The success of HTM applications is closely related to how well users adopt and use the technologies. Perceived difficulties in use can be a barrier to intention to use these technologies and consequently result into lower adherence (7). Especially older adults are novice users of these new technologies and so the potential benefits are put at risk by these barriers (26). Barriers for elderly in accepting applications are computer anxiety, concerns of data security, lack of perceived usefulness for electronic technologies and concerns regarding the usability of these technologies (46-48). According to Cruz (7), HTM systems should fulfil certain criteria to be successful: easy to use, ensure data security, and operate without any interruptions.

Perceived usefulness and Perceived ease of use (usability) are key constructs of several acceptance models. The Technology Acceptance Model (TAM) (27) is the leading theory in analysis of acceptance. A previous study (49) used the TAM model to investigate the acceptance by COPD patients of an eHealth tool. The author developed a questionnaire based on this model. This study concluded that the TAM model adapted to COPD users was able to explain 58% of the variation in intention to use. However. the latest derivative of TAM is the

(23)

22 Unified Theory of Acceptance and Use of Technology (28), see figure 5. The UTAUT model is better in explaining the variation in Intention to use than the TAM model. The UTAUT model was able to explain 70% of this variation (28).

Figure 5 UTAUT model (28)

The UTAUT model tries to define several constructs that influence use behaviour, which is “the degree to which a person actually uses the system (28)”. The following definitions are used for each construct (28):

 Behavioural Intention: an individual’s motivation or willingness to exert effort to

perform the target behaviour

 Performance Expectancy: the degree to which using a system will provide benefits to the user in performing certain activities

 Effort Expectancy: the degree of ease associated with the use of the system

 Facilitating conditions: the degree to which user believe that resources and support

are available to perform certain activities

 Social Influence: the degree to which an user perceives that important others believe he or she should use the system.

Behavioural intention is found to be a predictor of actual use behaviour of a system (28, 50). Behavioural intention is influenced by 3 core constructs in the model: performance expectancy, effort expectancy, and social influence. Performance expectancy and effort expectancy are the two most relevant predictors for intention to use, because they are derived from perceived usefulness and perceived ease of use introduced in the original TAM model (10). Another

(24)

23 construct, facilitating conditions directly influences use behaviour. In this study, we define acceptance as a concept which consists of these four constructs. The theory also suggests that the effect of these four constructs is moderated by four other variables: age, gender, experience and voluntariness of use. These four other variables are not always incorporated to assess acceptance of mHealth.

Several studies extended the UTAUT model (10, 50). An extended version of the UTAUT (UTAUT2) was developed by the authors of the original model (50). In this model, facilitating conditions also influences behavioural intention to use. Three new constructs, hedonic motivation, habits and price value were added to the model to influence intention to use. Voluntariness of use was deleted from the model because it did not significantly affected intention to use. Another study (10) extended the UTAUT especially for the elderly. In addition, one more study (51) showed that some aspects which are mentioned by older adults as important for acceptance, are not included in the original UTAUT. Therefore an extended version of the UTAUT was developed.

Acceptance of HTM applications for patients with HF or COPD have been evaluated several times (7, 13, 33, 52-62). However, most of these studies lack a strong theoretical background to investigate acceptance, such as the UTAUT model. In fact, none of these studies makes use of the UTAUT or TAM. Different telemonitoring systems, such as mobile phones, computers, or other devices are evaluated. A large proportion of these studies make use of mobile phones to send data, but no smartphones or tablets are included. In addition, acceptance of the measuring equipment, such as the blood pressure monitor and weight scale, are not taken into account. The measuring equipment is part of the whole monitoring process, so it does influence acceptance of HTM applications. Hall (56) reports that although most participants reported monitoring their blood pressure regularly, few monitor their weight daily. Patients are aware of the importance of daily weighing but several factors contribute to unwillingness to weight, such as not wanting to know their weight, broken scale or time.

(25)

24

4. Methods

4.1 Usability methods

4.1.1 HIMSS guidelines use analysis

To investigate whether the extended HIMSS usability guideline is (implictely/explicitely) used when designing cVitals, a structured interview with open-ended questions is carried out with the User Interface (UI)/User Exeperience (UX) designer of cVitals.

The only UI/UX designer of cVitals is contacted by email and he agreed to participate in this study. The interview is based on the different concepts of the extended HIMSS guideline and is performed by the first researcher. With consent of the designer the interviews are audio recorded to collect the data, supported by written notes of the first researcher. At first, the designer is asked if he is aware of the HIMSS guidelines and if he uses this guideline and/or other guidelines. For each principle of the guideline, the designer is asked if he knows what this principle encompasses. The first researcher explains the principle when the designer does not know the meaning. Subsequently, the designer is asked if he incorporates this in designing cVitals and if he is able to give examples of using this principle.

The results of this interview are transcribed and analysed by the first researcher. The first researcher identifies which principles are used and how they are used. Explanations of principles which are not used are described as well.

4.1.2 Think aloud study

Study design

A user based approach by means of conducting a think aloud study is used to identify usability issues. An overview of the think aloud study design is shown in figure 6. The different aspects of the study design are explained in the next paragraphs.

Participants

All care organisations using cVitals were asked whether they agreed to contact their patients for scientific research to investigate usability. When care organisations agreed to contact patients for this study, the main researcher contacted them by phone. Formal ethical approval of a medical ethics committee was unnecessary since the Medical Research Involving Human Subjects (WMO) does not apply to this study. Informed consent of both the care organisation as the patients was performed by telephone.

The population consists of patients who used the HF or COPD programme in cVitals. One care organisation agreed to contact their patients so all participants of this study were monitored by this care organisation.

(26)

25

Think aloud protocol

A user-based approach in the form of think aloud sessions is used for the usability evaluation since it is important to identify underlying causes of problems so that useful recommendations for redesign can be suggested. Additionally, cVitals is already evaluated multiple times by designers who have knowledge of usability but less from the user perspective.

Think aloud study design

an al ys is Th in k-al ou d Pa rt ici pa nt s Start

Webapp Native app

10 users (5 COPD en 5 HF) Differentiate between... HIMSS task mHealth framework Suggestions for improvement Severity level End Think aloud evaluation Usability issues Clustering without HIMSS

Figure 6 Usability study design

The think aloud method is used to identify usability problems by studying two groups that are evaluated: a group of five COPD patients and five HF patients. The number of

(27)

26 participants is stated to be a good trade-off between the amount of time and benefits (63). During the period of performing the think aloud sessions no new versions of cVitals are released. The think-aloud sessions are structured by 9 tasks. These tasks are designed by the first researcher, the second researcher and productmanager of cVitals. During the think aloud sessions, users are asked to think aloud while performing the tasks. The first researcher guides the patients to the session but did not intervene or disrupt the thinking process. Help is only provided when patient explicitly requested this and tried to complete the tasks themselves. The think-aloud sessions are documented by verbal reports. In addition, the completion rate and time to complete task are documented. The completion rate is an indicator for effectiveness and time on task (TOT) is an indicator for efficiency. Completion rate is the percentage of test participants who successfully complete the task without assistance from the first researcher. The time to complete a task is referred to as time on task (TOT). It is measured from the time the person begins the task to the time he/she signals completion or the researcher indicates that the task is achieved. These indicators are recommended by ISO to test these usability principles (30).

After the think-aloud sessions, patients are interviewed about their opinion of cVitals and are asked to fill in the System Usability Scale (SUS) (64). The interviewer takes notes of the interview and defined a summary from each interview. The SUS is used to measure user satisfaction. The SUS is scored on a 5-point Likert scale (where 1=strongly agree). The SUS provides a general measure of system satisfaction (65). Although the SUS is originally developed to measure usability, it is a common used tool to measure user satisfaction (38, 43, 46). The original SUS is English, so it is translated to Dutch. Translations are made based on previous translations (66, 67) and on the judgement of the first and second researcher.

Afterwards, the Six-item Cognitive Impairment Test (6CIT) (68) is conducted to investigate if patients had cognitive impairments. These results are both an indication of the cognitive state of the participants and it can explain encountered issues during the think-aloud sessions. The 6CIT is used because it only takes a few minutes to administer and appears to be suitable to screen for cognitive impairments in older patients at home (68-70).

Data analysis

Firstly, the results of the think aloud sessions are labelled as usability issues or suggestions for improvement. Secondly, the encountered usability issues are clustered among the 9 predefined tasks:

1. Open cVitals

2. Measure and send blood pressure 3. Measure and send weight

4. Fill in and send questionnaire

5. Show how many points you had for the last questionnaire 6. Find out how many minutes you should exercise per day 7. Fill in that you are absent from 10 to 12 October

(28)

27 8. Check whether you have messages of missed measurements

9. Find out if you can change your measurements day as a patient

Task 2 and 3 are only performed by HF patients because COPD patients do not measure weight and blood pressure. See appendix A for the task instructions for the participants. Thirdly, the encountered issues and suggestions are clustered amongst the themes of the extended HIMMS usability Guideline for mHealth. Some issues and suggestions cannot be clustered amongst the HIMSS. Therefore, these issues are clustered among new themes discussed and made by the first and second researcher. Fourthly, usability issues and suggestions are clustered onto the framework of mHealth for older users. Finally, the usability problems are prioritized by means of a severity rating from 0 to 4 according to Nielsen classification of severity (71):

0. I don't agree that this is a usability problem at all

1. Cosmetic problem only: need not be fixed unless extra time is available on project 2. Minor usability problem: fixing this should be given low priority

3. Major usability problem: important to fix, so should be given high priority 4. Usability catastrophe: imperative to fix this before product can be released

The severity rating helps to decide about which problems should be translated to specific design recommendations. Describing and clustering is performed by both the first researcher and the second researcher.

4.2 Acceptance methods

Design

To answer the research questions a semi-structured questionnaire is used. The questionnaire is send to new users at baseline (T0) within four weeks after installation of cVitals and is repeated after 12 weeks (T1), see figure 7.

Figure 7 Study flow patients acceptance

An overview of the data collection schedule can be seen in figure 8. The different aspects of the study design are explained in the next paragraphs.

(29)

28

Acceptance study design

U TA U T pr ep ar at io n U TA U T D at a an al ys is Start Literature orientation UTAUT questionnaire Patient selection Request for participation by email Request for participation by telephone

end _{Fill in questionnaire} by email Fill in questionnaire by telephone* Yes No Yes Document questionnaire in excel Data analysis End end No

Utaut T1 on same patients Utaut T1 on same patients

Figure 8 Acceptance study design

Participants

Informed consent of the care organisations of participating in this research is received by telephone. Formal ethical approval of a medical ethics committee is unnecessary since the Medical Research Involving Human Subjects (WMO) does not apply to this study. When care organisations agree to contact patients for this study, the first researcher calls these patients to request for participation of the acceptance evaluation. The aim of the researcher is to

(30)

29 interview at least 25 patients. This amount is based on the amount of new users which started two months before the evaluation. However, this amount could not be reached when the interviews could only be done by telephone.Telephone numbers of new users are not always available since not all new users received an installation by FocusCura. Therefore, the questionnaires are also emailed to new users for which the telephone number was unknown since mail addresses are always available.

The population consists of new users who used the HF, COPD or the heart problem programme in cVitals for a maximum of four weeks. Patients from three different care organisations in the Netherlands are included. Patients who follow the heart problem program do not have necessarily heart failure. However, these patients were included as HF patients because these patients need to perform similar measurements. The heart problem patients are considered as heart failure patients for further analysis. Only patients who use cVitals for less than four weeks are included. The period of inclusion of T0 is between July and September 2016.

At T1 both patients who are still using cVitals and patients who already stopped are included. A prerequisite for the users who already stopped is that they should have used the system for at least eight weeks. One care organisation stopped using cVitals and as such multiple users quitted. However, these users are nevertheless included because they may still have intention to use cVitals. The period of inclusion of the T0 questionnaire is between October and December 2016.

UTAUT questionnaire

To be able to collect data from multiple new users of the cVitals within the time limits of the thesis, conducting the questionnaire by telephone and email is necessary. To prepare for the telephonic interviews the researcher uses literature on how to perform telephonic questionnaires in a scientific setting. The program Surveymonkey is used to make the questionnaire and to send emails to the patients.

In this study, evaluation of acceptance is based on a theoretical background of the Unified Theory of Acceptance (28). An extended version of UTAUT is used as a tool in the form of a questionnaire to evaluate acceptance of the cVitals. Four studies (10, 28, 50, 51) are used to construct the extended version of the UTAUT. The original version (28) is used as a basic questionnaire. Ultimately, four factors are added to this model: The following definitions are used for these four constructs:

 Hedonic Motivation: the fun or pleasure derived from using the technology  Habit: the extent to which people tend to perform behaviours automatically

because of learning

 Computer anxiety: anxiety to use the system

 Contentment and trust: the extent to which people think the system protects their data and if people feel more sure and connected

(31)

30 The model tested to identify the influencing factors on behavioural intention to use (acceptance) can be found in figure 9. All items of these factors are scored on a 7-point Likert scale (where 1=strongly disagree and 7=strongly agree).

Behavioural Intention to use

Items: BI1, BI2, BI3 Computer Anxiety (CA)

Items: CA1, CA2, CA3 Habits (HA)

Items: HA1, HA2, HA3, HA4

Contentment and trust (CT)

Items: CT1, CT2, CT3, CT4

Effort Expectancy (EE) Items: EE1, EE2, EE3,

EE4, EE5

Social Influence (SI) Items: SI1, SI2, SI3, SI4

Facilitating Conditions (FC) Items: FC1, FC2, FC3, FC4 Hedonic Motivation (HM) Items: HM1, HM2, HM3, HM4 Performance Expectancy (PE) Items: PE1, PE2, PE3, PE4, PE5, PE6, PE7, PE8, PE9, PE10

Figure 9 The model tested to explain intention to use

General questions, like gender, age and experience level, and use of other technologies are asked to have an overview of the characteristics of the study population. The interviews lasts for about 10 minutes. After the first interview, the patients are asked to participate 3 months later. See appendix B for the extended UTAUT questionnaire.

Data analysis

Analysis are performed using the Statistical Package for Social Sciences (IBM SPSS statistics 24). Descriptive statistics are used to describe the study sample, internet experience and severity of disease. An independent t-test is applied to describe whether significant differences between COPD and HF patients exist. A descriptive two paired t-test is performed to search for significant differences between the means scores on factors of both T0 and T1.

To analyze the influencing factors on behavioural intention to use (BI) a principal component analysis should be performed on both datasets T0 and T1. This principal component analysis consist of 3 main steps:

1. Meeting the preconditions: the Kaiser-Meyer-Olkin measure of sampling adequacy

should be greater than 0.6 (72, 73). The data should also be suitable for data reduction. Therefore, Bartlett’s test of sphericity must be significant (<0.05) (72, 73).

(32)

31 Additionally, several correlations with scores above 0.3 in the correlation matrix should exist (74). To fulfil these preconditions, the items of the questionnaire are added one by one. When a new item causes a violation of the preconditions, it is removed from further analysis.

2. Determining number of components: to determine the number of components

which should be included the Kaiser criterion and the Screeplot test are used. According to the Kaiser criterion (75) the number of components are the components which have a eigenvalue of > 1.0 in the Total Variance Explained table. The number of components determined by the screeplot test are the components which are above the ‘elbow’ of the graph (76). As a check, the component matrix is checked on which components score the majority of the items. The number of items per component should be more than two. An oblique rotation, direct oblimin, is used because it is assumed that the (possible) influencing factors are correlated.

3. Data analysis: the analysis is carried out again with the defined amount of

components. The data sets are checked for missing values. When these values incorporate less than 5% of all values and are completely missing at random, they are replaced by the mean for further analysis. Cronbach’s alpha is calculated to test the internal consistency of the different items of each factor of the questionnaire. To determine the influencing factors on intention to use the pattern matrix is used. The highest values of the items determine to which component a factor belongs. When a factor scores on the same component with the items of intention to use it means that they relate to each other and the mean score of the factor is calculated. If items of the same factor are distributed among different components the mean of all highest values on each component are calculated and the factor scores on the component with the highest mean.

(33)

32

5. Results

5.1 Usability results

5.1.1 HIMSS guidelines use analysis

The interview was performed on 25 may 2016. The UI/UX designer was unfamiliar with the extended HIMSS guidelines for usability. However, the designer used implicitly almost all aspects of the extended HIMSS guidelines when designing cVitals, see appendix C for the subscriptions of the interview. Only two aspects of the guideline are not yet taken into account: discrete transactions and behaviour change stimulation. Discrete transactions are not taken into account because, according to the designer: “If you look honestly at yourself then you

should also be honest to yourself that something you have a sickness. I do not think that it would help if I hide the disease from them”. Consequently, patients are not able to postpone

measurements. Behaviour change stimulation is not taken into account yet but the designer would like to incorporate it in cVitals. The designer considers it difficult to incorporate behaviour change stimulation since “you have to have a certain algorithm to account for each

individual”.

The main guideline for designing cVitals is the IOS Human Interface guidelines (77). Consistency and experience with look-a-like apps are main reasons. These guidelines advice on designing IOS apps. The guidelines are detailed, advices such as fonts and specific buttons are described. Reasoning based on experience of the designer and testing with the user are an important factor in the design process as well. Feedback from the users is incorporated but cannot always be incorporated in a new design. The reason is the choice for the IOS guidelines as standard. Some designs might be better for users, but are not incorporated because they don’t comply with the IOS guidelines. In addition, experience with other devices, e.g. android is not taken into account because of the use of IOS guidelines.

5.1.2. Think aloud study

The think aloud sessions were performed from 20 September until 11 October 2016. The think aloud sessions were guided by the first researcher. The product manager and the designer of cVitals attended both one session separately. To not influence the think aloud sessions, these employees were instructed to not interrupt users during the think aloud sessions and ask their questions before or afterwards.

Participant characteristics

In total ten users (five COPD and five HF) were included in the think aloud sessions. Table 1 shows the participant characteristics. The mean age of all participants was 67 years. Three females and seven males were included. All users scored less than eight on the 6CIT, which means that no signs of cognitive impairment were detected. The average SUS score was 85 on a scale of 100, which is above average. HF patients (81.5) scored lower on the SUS than COPD patients (87). Four users (2 HF, 2 COPD) started with cVitals in 2014 and therefore used the

(34)

33 webapp. Six started this year and used the native app. The latest version during the think aloud sessions was 2.8 for the webapp and 1.9 for the native app. Nevertheless, it was left to the user whether they updated cVitals to the latest version. As a consequence, some users used an older version of cVitals. Some users did not performed any updates at all or did not know that they should accept updates. Different measuring equipment was used among the HF patients. Two HF patients used a Withings bloodpressure monitor and weightscale. The other three HF patients used the Omron MIT 5 bloodpressure monitor and the iHealth weightscale. Two users declared always using cVitals together with their partner. The reasons to do this were a reduced vision of the main user and technology anxiety. Accordingly, the choice was made to include the partners as users by the identification of usability issues as well.

Table 1 Participant characteristics

Heartfailure N=5 COPD N=5 gender male female 3 2 4 1 age (M) 66 67 Experience Unexperienced Reasonable experienced Experienced very experienced 3 1 1 0 1 3 0 1

Use of visual aid

reading glass glass lenses lens no aid 2 2 0 0 1 1 3 0 1 0 State of hearing Good Reasonable

Bad, use of hearing aid

4 0 1 3 1 1 6CIT score (M) 2.8 1.2 SUS score (M) 81.5 87 Usability issues

In total 51 usability issues of cVitals were identified. Table 2 shows the issues by task clustered among the HIMSS guidelines. Table 3 shows the encountered usability issues clustered without the HIMSS guidelines. Appendix D incorporates the average time on task and completion rate for each participant.

Table 2 Encountered issues by task Task AVG time

on task (min:sec)

Completion rate (%)

Usability issue HIMSS mHealth

framework for Older Users

Severity

1 0:27 100% No usability issues - - -

2 6:39 80% Uncertain what it means when the light of the blood pressure monitor (withings) blinks. Forgiveness and feedback - 4.3 Motivational issues- computer literacy 2