Article Text

other Versions

PDF

Original article
Development and potential of a clinical resource and educational iPhone and iPad App based on patient testimony
  1. Andy Levy,
  2. Dominic Alder
  1. Henry Wellcome Labs, University of Bristol, Bristol, UK
  1. Correspondence to Professor Andy Levy, Henry Wellcome Labs, University of Bristol, Dorothy Hodgkin Building, 1 Whitson Street, Bristol BS1 3NY, UK; mdal{at}bris.ac.uk

Abstract

This paper describes the development and production of an educational and reference app designed for the medical and pharmaceutical industries as well as interested members of the public. ‘Clinically Speaking’ consists of edited, impromptu movies of clinical interviews of patients with common and rare conditions archived over two decades. Each is indexed by subject area and disease, identified by short title and thumbnail and accompanied by a description and Wikipedia URL. The app design allows users to scroll down through subject areas such as cardiology and gastroenterology, then across to access several hundred edited cases. For users with Windows, Google or Android rather than iOS devices, a similar experience is delivered through a web-based version. The almost complete market penetration of smart phones and computers allows educational database material to be distributed very economically. An opportunity to partially recoup design and coding costs and cover on-going expenses of upgrades, server charges and annual developer fees is afforded by app store sales. Distributed worldwide via the app store, ‘Clinically Speaking’ constitutes a versatile vehicle for addition of further cases and provides experience of conditions that are infrequently seen in practice. Apps and mobile devices capture the current zeitgeist and constitute potentially powerful new vehicles for delivering reference material and clinical medical education. The development process, described herein, is slow, expensive and complex. The market is saturated and the rationale for adding further apps needs careful consideration.

  • Medical Apps
  • mHealth
  • Assistive Technology

Statistics from Altmetric.com

Introduction

Clinical medical education has traditionally centred around patient contact and is therefore largely confined to the wards, consulting rooms and out-patient departments of hospitals and surgeries. Rather than the day or two before a planned intervention, patients are now often admitted only hours before even major elective surgery. Postoperative patients join acute medical admissions and those undergoing day case procedures in convalescing at home. The on-going trend to minimise admission duration coupled with increasing restrictions such as ‘protected meal times’, extended visiting and working time directives, have eroded opportunities for students and trainee doctors to interact directly with patients in the clinical arena. Additional difficulties for students, junior and indeed senior doctors alike come from separation of patients into super-specialist silos almost immediately after admission. The consequence is that the body of experience that facilitates clinical pattern recognition from content, context and body language is increasingly difficult to acquire and reinforce. As learning how to manually record blood pressure and pulse, respiratory rate and temperature now seems less important than knowing how to interpret an observation chart, perhaps it is timely for new technology to substitute for some of the elements previously assimilated through clinical experience. Good quality online learning material is also valued for its underlying consistency, which contrasts with the varied teaching and teaching styles experienced at the clinical interface.1

The rationale for developing this clinical resource was to provide undergraduate medical students and qualified doctors, members of the pharmaceutical industry and interested lay people with an easily accessible source of first person testimony about medical conditions and the effects of treatment. The library would include information and insights that certainly for rare conditions, might not be available from any other source.

Distribution of some of the same material had been tried several times before. Edited video case-based interviews were originally distributed in compact discs bound with various publications,2–4 and as a weekly illustrated case uploaded to a dedicated website. The latter, which also hosted therapeutics quizzes and a question and answer column, ran for several successive years and was coded to collect quantitative usage data. Despite emulating very closely the content of very high stake assessments and positioned to feed into revision, the site attracted at most only 50% of the student corpus in any 1 week, and often no more than 10–15% of students. Similar case-based material was subsequently used for a prototype internet medical school ‘MedSchool.com’ that similarly failed to gain traction at the time. By providing more powerful and much more portable platforms, smart phone and tablet computers might encourage increased usage of case-based video material. The causes of the discrepancy between perceived utility and actual use of an electronic resource, however, remain complex.5 Worthwhile feedback about an app concept at an early design stage is extremely difficult to obtain, and empty enthusiasm that will not translate into usage is almost inevitable. It is analogous to being asked whether the food you have been served in a restaurant is to one's satisfaction. Even if exactly the opposite is the case, the damage-limiting response to a poor dining experience is not to confess the truth at the time, but to eat elsewhere in future. Thus a central issue is not whether an icon is intriguing, an interface compelling, or the time and effort expended to collect content heroic, but whether anyone would wish to avail themselves of the information in the first place.

The stakes are quite high. In the app ‘Clinically Speaking’, the model for this article, the content library, consisting of movies of patients describing their medical conditions and experiences, took more than two decades to collect. The time from the first concept to commissioning and publication, annotating, editing and designing and coding, took 5 years. Coding itself took several months. The central function of the app was extended to become more than a video library of cases classified by specialty area, but an authorship vehicle to allow users to become producers and co-owners of the content. As an activity-based medical education strategy, knowledge and insights brought by students from different cultural backgrounds and educational experiences can be incorporated and enjoyed as a contribution for all.6

Methods

The strategy described above was used to archive patient interviews first using Hi8, then DV tape and subsequently cameras recording at high definition onto built in memory or SD-card solid state memory as new technologies became available. Content is critical for an app based on patient testimony and establishing a rapport is central to that. Patients have to be relaxed and confident that they will not feel foolish if their conversation founders. Encouragement can be communicated to the patient with nods and smiles rather than words, and brief silences after answers—although a little unnerving socially—are very useful editorially. Almost all information is carried in the sound rather than video, and the ambient clatter in out-patient departments and on wards is important to add atmosphere and context. As background images contribute little, tight facial close-up is most effective at making the best use of the medium and helps to prevent auto-exposure systems from silhouetting subjects against a brighter background. Even seemingly innocuous illumination from one side such as overcast daylight through a window can lead to deep shadows and reduce image quality unless some care is taken with seating arrangements. A camera light source may be useful on occasions to fill in the shadows but is increasingly unnecessary as technology improves. A consequence of extreme close-up is that to seamlessly track patient movements the camera must be hand held rather than fixed to a tripod. All cameras now have excellent steady shot technology, so very little skill is required to record smooth images provided the temptation to zoom and pan are resisted. During editing, no transition effects other than the occasional cross fade were used, as what might be moderately entertaining on first viewing becomes distracting and irritating by the second. Consent is important as patients are not named but remain recognisable. The content of the consent form was clarified for each patient by explaining the implications of ‘moral rights’, but also by making it clear that the footage would not be used if at any time the patient had second thoughts about appearing. The text of the consent form that was used throughout the two decade long video collection period is shown in table 1. It is important to note that content for the app was being collected for many years before an appropriate vehicle was developed to distribute it. There is no quick fix for absence of content, and the effort and time required to collect it should not be undervalued or underestimated.

Table 1

The text of the consent form used for video recording

Results

The icon for an app is believed to be important as a first impression and might have the potential to significantly affect usage. It is the image on which assumptions about functionality and quality may be built. It has to work at small scale and be clear yet distinctive. It needs to be simple, memorable and somehow convey the purpose of the app or at least some sense of its utility. It should aspire to the impact of a memorable commercial or print advertisement. It has to complement other app icons already displayed on the device that the user will access it from and needs to flow in colour and architecture seamlessly into the screens that follow. Iconography also has to match current fashion. Skeuomorphism, that is, faux leather and fabric texturing, ring-binding, shadow and pseudo-3D effects now looks dated. Borders and embedded text should also be avoided as both are added by the system and may subtly change. Examples of different icons explored for the patient video app are shown in figure 1, with the final version created with professional design assistance.

Figure 1

Different icons considered for the app, some of which incorporate early names for the app.

The name is a brand that needs to be unique, short, descriptive and above all memorable for good or even bad reasons. Avoiding existing names is important, particularly if the purposes of the products are similar, and it is important to ensure that a matching web domain and email identity are available. A series of names were considered for the app, including Vox, Patients’ Voice and Patient Voices, before ‘Clinically Speaking’ was chosen.

The interface not only needs to be compelling and intuitive to use, but has to be accessible to people with dyslexia or those who might not be able to hear or see clearly. It should contain suitably positioned and sized buttons that users are familiar with and while helping them to understand and interact with the content, do not distract from it. Particularly for an app that depends on repeated downloading of large files, the interface should control responses with loading indicators to throttle down repeat key strokes as users become frustrated or wonder what might be going on. The key considerations to ensure that users have a high quality and familiar experience are described in the iOS human interface guidelines.7 A core element of the design process was to carefully review and evaluate large numbers of published apps, both medical and non-medical, for their utility and usability.

Once key elements had been considered, and the design of many extant apps reviewed, specific design ideas were sketched on paper as a storyboard. KeyNote was then used to trial various behaviours, design ideas and interface transitions. By conveying the impression of scrolling or sliding between pages, animations indicate how various elements relate to each other and enhance the sense that the user is in control. They direct the attention of the user, keep them orientated and demonstrate destinations within the app and how to navigate home, as well as imparting a pleasing and unique character and experience. The structure needed to be accessible and self-explanatory so that users would know how to use it at the outset, and versatile so that additional content categories could be added.

For ‘Clinically Speaking’ a structure very similar to the one used for the original BBC news app seemed the most appropriate for the archive of case reports (see examples of pages from an interface mockup in figure 2). It allowed users to scroll vertically through the systems (cardiology, respiratory, neurology etc) then horizontally to select from thumbnails of over 300 cases. Each case consists of a short explanatory text (about 100 words), a link to the Wikipedia URL related to the subject and symptom key words linking cases to each other. A similar format was used for the web version, operating without touch screen and ‘swipe’ facility.

Figure 2

Examples of one of the early interface mockups drawn in Adobe Photoshop and Adobe Illustrator. The thumbnail ‘buttons’ were shaded from 30% to 90% vertically with a grey 0.4 point thick stroke surround with 1.5 mm radius rounded corners. The panel behind the ‘clinical voices’ logo at the top is 90% grey fading abruptly to 5% grey at the top and bottom, giving the impression of lines. The subject areas, cardiology, respiratory medicine and general (internal) medicine are in 7 point Helvetica bold at 5% grey on a back panel that is 90% grey. The video thumb nails are 20 × 15 mm (4 × 3 ratio) within a 20 × 20 mm box, the bottom half of which has an overlay filled with a vertical gradient from 5% grey at the top to 100% at the bottom with transparency at 100% ‘darken’. The text description of each case is Helvetica bold 6.5 point spaced at 8 point with AV (kerning, ie, Character line adjustment) set at zero and hyphen-free. The horizontal lines separating the thumbnails into horizontal rows is 0.5 point black above 0.5 point 90% grey, 4 mm above the top of the box below and 1.5 mm below the row of boxes above.

‘Rent a coder’, ‘PeoplePerHour’, ‘Fiverr’ and ‘Elance’ or innovation centres such as those attached to educational institutions are useful sources to obtain an idea of costs and what might be involved in your project when you pitch it online. Ultimately it is still invaluable to meet the people you will be working with, see their previous work and start the process of project design, structure and negotiation. Multiple quotes might be useful, although there is always a risk that inappropriately low quotes that do not allow the project to be completed will be made in order to secure the business. A strategy to avoid this may be to explain to potential coders that you will request several quotes and most strongly consider the one nearest the mean, rather than the one that is the lowest. As coding time is expensive and work schedule time limited, the features and functions of the app need to be agreed with as much clarity as possible at the outset to avoid or at least minimise so-called scope creep or function creep leading to cost over-run.

The tools for coding theoretically make app design and construction accessible to non-experts, but the process is complex, time consuming and requires considerable skill, patience, attention to detail and innate aptitude. Often derived from the ranks of computer science, physics or mathematics graduates with particular mindsets, coders have to take into account the changing freedoms, demands and constraints of the delivery platforms such as the development of images for high definition displays, and requirement to scale content to device size. There are also very high expectations that the interface is not only intuitive but comfortable and compelling to use. This can be far from straightforward when large files are being handled and instant results expected. The app must survive being shut down at any point during its operation, and depending on its intended function, may need to resume at the same point. There must also be commitment to keep up with the continual stream of refinements and developments in coding languages and systems, and to be able to exploit the App Store algorithm to move their products up the charts.

Submitting apps to the app store is complicated. The iOS Dev Center has a provisional portal which is required for creating development certificates and provisioning profiles. iTunes Connect with the Apple Application Loader utility, is then used to prepare the app for submission to Apple.

Costs

Financial reward is unlikely to be a priority in producing medical educational apps, but it is important to appreciate that the initial outlay in design, programming and various registrations is followed by on-going costs for updating the software, and hosting the domain and content. There will be recurrent annual fees for domains and ‘developer’ charges by Apple, even if you consider yourself ‘a customer’ of a coding company rather than ‘a developer’ in your own right. It is all too easy to assume that the available, qualified and target market will translate into purchases. Add on to this the cost of developing a user-friendly ‘back end’ if the information delivered by the app or on its accompanying web site need to be frequently edited or augmented, and it is very likely, overall, that the investment will not be returned.

The first requirement is to write a business plan focusing on the rationale, then examining the market, including undergraduate and postgraduate students, and those seeking further medical education and revalidation. Additional markets might be pharmaceutical industry staff, patient support groups and the general public who for personal reasons might be interested in finding out more about particular medical condition. It is, however, important to bear in mind that a wealth of information is provided free, and increasingly, the expectation is exactly that. Expecting users to pay even a modest fee to download an app may be unrealistic. Funded from in-app advertising is a possibility, but it may make the user feel uncomfortable and will not provide a realistic revenue stream unless usage figures are unattainably high.

Marketing

Educational apps are the third most popular category in terms of downloads, with games and business apps first and second. Medical apps, of which there are roughly 43 000 currently available, account for just over 2% of downloads. At the time of writing there are roughly 1.5 million apps in the app store and new ones are being added at over 1000 a day. In such a crowded market, standing out is almost impossible irrespective of the novelty or usability of the content and function, although word of mouth and promotion through the active medical education community across social media platforms can make an app more noticeable. A great idea, some money and programming skills is unlikely to be enough to join the elite ranks of the small band of developers who recoup the cost of their investment. Listing and carefully considering the strengths of the idea, its weaknesses, the opportunities it affords and the threats to success at the outset, is a useful conceptual tool but rarely a reality check. The cost of creating apps tends to be prohibitive and even if the content is instantly engaging, the user conversion rate, defined as the gap between considering and actually completing a purchase—is wide.

The whole concept needs to be very carefully thought through with a clear evaluation of the potential utility and versatility of the app and a candid appraisal of the fundamental objectives of creating it in the first place. The competition that needs to be considered is not limited to the internet and electronic media, but includes other calls on the time potential purchasers’ feel able to devote to education.

Within educational establishments, the provision of portable platforms for distribution of curriculum information is becoming more common. For institutions where such plans remain as yet unfulfilled, care must be taken to ensure that content is universally accessible, irrespective of the particular platforms that the target audience have at their disposal. Android and iOS applications use different programming languages and frameworks so that coding costs will be effectively doubled by developing an app for both platforms. A more economical strategy involved developing a complementary website with content accessible on computer, tablet or mobile phone by subscription.8

Conclusions

Making movies of patients describing their experience of disease widely accessible through the App Store is a potentially exciting advance in medical education. Accessing the very crowded development and marketing world of apps is expensive. It requires the services of designers and coders able to interpret a brief and produce a product that is useful and compelling for end users. It also has to have a simple and accessible ‘back end’ for the client to edit and upload information, and be deliverable within a budget and agreed timeframe. Coders also need to provide prompt support in the longer term and manage the interface with the various hosting and distribution organisations. As part of a platform for higher educational establishments, the content must be accessible to all students, irrespective of their personal choice of platform and any disability that they might have. The expectation of users is that educational material is compelling and free to use, and accessible at a time and location that suits. Finally, it is important to be mindful that new concepts and fashions in medical education and the content and structure of assessments can quite abruptly change learning expectations and the perceived utility of teaching resources.

Acknowledgments

Andy Caddell (andy@mccallanmarketing.com) very generously came up with a series of names for the app including ‘Clinically Speaking’, which was the clear winner

References

View Abstract

Footnotes

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.