Medication Management Support for Vision Impaired Elderly

Scenarios and Technological Possibilities

Minna Isomursu, Mari Ervasti, Vili Törmänen VTT Technical Research Centre of Finland

Oulu, Finland Abstract—This paper describes scenarios and discusses technological possibilities for implementing a mobile application to support medication management of elderly vision impaired people. The scenarios have been defined in workshops and discussions between technology providers, elderly care personnel, representatives of associations for blind, and pharmacy professionals complemented with observation of vision impaired elderly. Technological discussion evaluates the possibilities for implementing a medication support application with near-field communication (NFC) technology. The goal is to develop a mobile application concept that allows vision impaired elderly to manage their daily medication autonomously.

Keywords-component; near-field communication, ambient assisted living, medication adherence

I. INTRODUCTION The proportion of older people in developed countries is

rapidly increasing. As people age, they experience declines in many abilities that impact on various aspects of their everyday lives, resulting in a growing need for more support in carrying out their daily tasks and activities. Different degrees of vision impairments are inevitable results of growing old, as the physiology of our eyes changes with time when the eye tissues loose their flexibility and suffer from damages caused by everyday life, different health conditions (such as diabetes or blood pressure) and gravity. Our possibilities to prevent vision impairment with technology are very limited, but there are promising possibilities to support the visually impaired elderly in better managing their everyday lives with the help of modern information and communication technology. Older people desire independence [14], and successful independent living involves the ability to carry out daily living activities [15].

The elderly experience more troubles with their health, which appear as symptoms and diseases that also require medical treatment. The elderly use a substantial amount of drugs; between four and six medications on a daily basis [9], [10]. A subgroup of the elderly (frail elderly with multiple co-morbidities) takes an average of nine medications [11].

An important aspect of medication use is the medication management (MM). MM has been defined for example in a following way: “managing medications is an instrumental self-care activity that requites cognitive and functional capabilities to coordinate and carry out the associated tasks” [12]. Beckman Gyllenstrand [13] describes the process of MM to include the

following five steps: 1) receiving a prescription, 2) filling the prescription, 3) storing the drug, 4) preparing a dose for administration, and 5) administering a dose. In her study Beckman Gyllenstrand found that in a Swedish population aged 77+, a large proportion (66%) of the elderly people was unable to perform rather simple tests related to medication management. In addition, she discovered that a substantial proportion (60.5%) of the elderly who reported that they could manage their medication were not able to perform all the tests of MM.

The work discussed in this paper explores the possibilities of using modern mobile communication devices in supporting elderly vision impaired in medication management.

II. BACKGROUND As people get older, the risk factors associated with

medication non-adherence become more common [3]. Non-adherence can be caused by unintentional or intentional reasons [3]. Unintentional non-adherence can happen because of forgetfulness or other cognitive reasons, or physical problems, such as difficulty of reading instructions or opening medication packaging. The patient may also intentionally decide not to take prescribed medicine as instructed.

In the five steps of the MM process described earlier, Beckman Gyllenstrand [13] found several functional components that are crucial for the whole process of MM. These components are vision, memory and comprehension, hand function and mobility. Beckman Gyllenstrand found significant correlations between memory and vision and knowing what dose to take, as well as between hand function and opening of containers.

In Europe, there is regulatory pressure for making medicine instructions accessible for vision impaired. The directive 2004/27/EC [18] requires that package information leaflet must be made available in formats appropriate for the blind and partially-sighted. Also, the same directive states that name of the medicine product must be printed in Braille format on the packaging. The problem with Braille format is that not all people who have problems with seeing can read Braille. For example, aging often causes visual impairments, and learning to read Braille at an old age might be difficult.

Here, we will address only the problem of unintentional non-adherence, and specifically on non-adherence and

978-1-4244-4641-4/09/$25.00 ©2009 IEEE

inconvenience caused by declined vision and difficulties of properly seeing and reading the medicine information and related instructions.

A. Ambient Assisted Living Concepts for Supporting Medication Adherance A wide range of educational, behavioural and technological

interventions to improve medication adherence have been developed and trialled. There is evidence that pharmacist medication reviews with individualized education and instructions can be effective in increasing medication adherence with elderly [3], even though it is time consuming and therefore difficult to implement in practice for large amounts of users.

Technological solutions presented in the literature include, for example, solutions based on medicine storage boxes or cabinets with embedded intelligence [4][5], electronic reminders [4][8] and behaviour or activity monitoring and modelling [5]. The solutions based on special storage boxes or cabinets may be costly to purchase and install. In addition, the medicine administration may be difficult and time consuming. Even high-functioning elderly have been reported to have major problems in placing medicines correctly in a medicine box [6]. Also, specific storage devices rarely take into account nor support the existing non-technology based methods for managing medications, such as spatio-temporal arrangements and temporal routines [7]. There is not much research available that provide strong evidence on the effectiveness and impact of technical solutions to medication adherence, and practically no evidence with elderly.

B. User Interface Solutions for Elderly Vision Impaired Ishihara et al. [2] found out that problem with resolution,

i.e. difficulties in reading small print, is the most prevalent visual problem with elderly. The space limitations of the surface available on the medical packaging are tight compared to the amount of information that is needed to present. This inevitably leads to small resolution text, and difficulties for elderly to effortlessly read the textual information.

An alternative for text would be to provide medicine information in audio. Audio interface requires a technical computing device. A mobile phone provides a promising platform for audio interface, as it is relatively cheap and commonly used both by vision impaired and elderly. The small keypad and screen-intensive user interfaces may cause difficulties for elderly with declined vision and hand-eye coordination. However, encouraging results of mobile phone use as service interface has been achieved with mobile interfaces based on touch paradigm [16].

The interface solution adopted in the solutions described in this paper are based on technology enabler called Near field communication (NFC); a technology based on Radio Frequency Identification (RFID) technique. With NFC it is possible to write and read data in electronic passive tags which can be attached to medicine packages. NFC reader can be embedded in a mobile phone. This allows building physical user interfaces that can be used simply by touching a tag with a mobile phone.

III. RESEARCH METHODS In this paper, we present the results of exploratory research

that aims at constructing scenarios on how a mobile application based on touch-based user interface could support vision impaired elderly in their daily medication processes. As a result of the exploratory process, we present alternative scenarios that can later be evaluated through user trials.

The constructions are based on service co-creation process [19] that has involved discussions, workshops and concept evaluations. The process has been carried out during year 2008 and 2009. The development work was actively started when the initial concept description won the yearly NFC application competition in the WIMA conference in year 2007. The following interest group representatives have contributed to the process resulting in construction presented in this paper:

• Technology providers. As our goal was to use mobile technology and touch-based interfaces, representatives of companies developing mobile phones and RFID technology were involved.

• Elderly care personnel. Both public and third-sector elderly care providers contributed with their knowledge and experiences related to common problems with elderly and medication, and provisioning and managing elderly care services.

• Representatives of associations of blind. The associations for blind provided their knowledge on practical constraints and difficulties related to vision impairments, and current availability and adoption of technological aids for vision impaired.

• Pharmacy professionals. Pharmacy professionals have contributed with their knowledge of relevant, practical and legally required data that is processed and needed in managing medicines, and in planning the integration with existing medication logistic systems.

• Elderly vision impaired people. Four elderly people participated through the process of contextual inquiry [20]. The oldest subject was 92 years old, and the youngest one was 83.

The actual scenario construction and analysis has been carried out in several successive workshops, and in discussions between different actors. The process has been coordinated by a group of researchers specialized in mobile applications, ICT supported services and human-computer interaction.

In this paper, we evaluate the scenarios solely from the technical and practical viewpoints, and leave the analysis of use outside the scope of this paper. Research shows [3] that methods that have been proposed and found effective for solving problems related to medication non-adherence often are difficult or impossible to use in practice. Reasons for that are related to, for example, requirements of increased amount of manual work by professionals who already are too busy and overloaded with responsibility. Therefore, it is important to evaluate not only the impact of intervention to non-adherence, but also from the practical feasibility of the scenario.

IV. SCENARIOS The technical enabler triggering the scenarios is Near field

communication (NFC) technology. In both scenarios an NFC enabled phone and NFC tags are used in physical interfaces used for providing audio information about medication for those who have difficulties in seeing and reading. The focus of this paper is on exploring how this technical enabler could be best utilized in a medication management service concept. The scenarios presented here are the results of cooperative service design process described in the previous chapter.

Two alternative scenarios were constructed. Both use NFC technology for creating the link between the medication package and the audio information presented through the mobile phone. The first scenario is based on an idea of providing the information currently available in written format in the patient package inserts through audio. The second scenario combines medication recognition with reminders and a background system monitoring medicine consumption.

A. Scenario 1: Audio interface to medicine information In the first scenario (Figure 1), the NFC tag containing the

medicine information is attached into the medicine package in the pharmacy. The computer used for printing patient package inserts is equipped with a NFC tag writer, and the same text that is printed in the patient package insert is stored as text on the NFC tag. In addition, the pharmacist writes the URL pointing to the information of the specific medicine in question in medicine database to the tag.

The NFC tag is then attached to the medicine box or packaging insert by the pharmacist. Now, when the tag is touched with a NFC enabled mobile phone, the speech synthesizer can be used for transforming the text stored in the tag into audio. In the case the user wants to continue receiving more detailed information about the medicine, the phone accesses the URL containing the detailed description of the

medicine. A small application is needed in a phone, as the information stored in the tag includes several structural parts (i.e. the package insert text and the URL) and the information needs to be processed in a specific way (i.e. speech synthesizer needs to be used).

B. Scenario 2: Integration of medication management with a background system The second scenario (Figure 2) is heavily based on a

background system managing the medicine consumption through pre-programmed medication administration.

The background system stores information about medication that needs to be taken according to defined schedule. The background system can be administered, for example, by a home care service provider, who sells medication support service for their clients. The personnel of the home care provider review the medication prescribed to the client, and creates a medication schedule which is stored into the background system. The medication schedule contains

Figure 2. Illustration of Scenario 2.

information about which medication need to be taken, required dosage and recommended timing of medication consumption. For each medicine included in the medication schedule, the home care employee inserts a NFC tag identification number corresponding to the tag attached into the medicine package. This can be done, for example, by

Figure 1. Illustration of Scenario 1.

equipping the computer used for medication schedule creation with an NFC reader.

When the background system identifies a scheduled need to take medicine, it sends a reminder to the NFC enabled mobile phone of the client. The reminder includes two parts. The first part is a text message which is processed by the speech synthesizer into audio message telling the name and dosage of the medicine that needs to be taken. The second part of the reminder includes the identification of the NFC tag that is attached into the medicine packaging. Elderly often need to take multiple medications. Reminders for all medications that are scheduled to be taken at the same time are sent to the phone at the same time. However, only one is activated by speech synthesizer, and the next one is activated after one medication has been successfully taken.Upon receiving the reminder, the user touches the medicine package to indicate medicine consumption. If a wrong medicine package is touched, the mobile phone tells the user that this medication is not needed at that point of time. When the user touches any medicine that needs to be taken at that point of time, the speech synthesizer indicates that the correct medicine was found, and repeats the dosage. This is repeated as long as all needed medication is consumed. If the background system does not receive a confirmation of correctly identified medication, it repeats the reminder after predefined period of time, and stores the information about possible non-adherence.

In case of medication that is not needed regularly, no reminders are sent by the background system. However, if the medication is tagged, and the user touches the tag when s/he takes such medicine, the mobile phone sends a message to the background system. This can be used in monitoring and reviewing medication usage.

V. EVALUATION The scenario evaluation is done with the following

evaluation framework:

• Effectiveness of adoption. How effective the scenario is in making the service available as widely as possible?

• Flexibility. How the scenario is able to integrate or support the personal procedures and methods created or adopted by the users for medication management.

• Cost model. Who would cover the costs?

• Availability of needed components. Are all the components that are required available at the time of writing the evaluation?

A. Scenario 1: Audio interface to medicine information

1) Effectiveness of adoption: In Finland, all medicines are processed through licensed and highly regulated pharmacies. Both over-the-counter and prescription medicines are sold only through pharmacies. Therefore, if pharmacies would play a central role in offering the NFC based medication identification, this would provide a distribution channel that would cover the whole country, and would be available for all

medication sold. Having pharmacies and trained pharmacists in the service process and integrating the tag writing with existing pharmacy processes would likely result in high trustworthiness in the quality and validity of the information. The same quality assurance procedures could be applied for written and electronic medication information. Finnish community pharmacies are privately owned small pharmacies with the average staff of six pharmacists and serving an average of 6500 patients [1]. Computerized systems for managing written patient information and for printing patient package inserts are used [1]. The pharmacies are connected through a common association which plays a central role in ICT development and information system integration for all pharmacies. Therefore, it is rather simple to integrate the scenario into existing technical infrastructures in a pharmacy.

This scenario can also be used for all medication needed. It can be used for prescription and over-the-counter medication. It can also manage regularly administered daily medication and medication that needs to be taken only occasionally when needed, e.g. in the case of an asthma attack or sudden head ache. As the medicine does not need to be stored in a specific storage place, it can manage medicine in different formats, such as liquid, powder or pills, and medicine can be stored in refrigerator if needed.

2) Flexibility: The scenario itself does not require the user to change the practices s/he has followed in medication management. The scenario does not require the user to store the medication in a specific way, nor does it force the user into predefined schedules or activity procedures. If the user has created specific methods for recognizing the medication, s/he can continue to use them.

Of course, the benefits of the scenario diminish, if the user does not use the mobile phone for listening to the medication information. If the user thinks s/he recognizes the medication correctly and trusts to remember the dosage required, there is a danger s/he does not use the technology as described in the scenario, and this might lead into situations of misidentification or wrong dosage. As the scenario has no mechanism for detecting non-adherence, it cannot react in such events.

3) Cost model: The direct costs caused by the scenario are the following: (1)NFC tags, (2) NFC tag writers in the pharmacy and (3) NFC enabled phones.

In addition, there might be some costs related to integrating the system with existing pharmacy information systems. However, this is considered here to be normal information system development work, and is left outside of discussion.

In Finland, there exists a healthcare framework for covering the expenses of special auxiliary devices needed by vision impaired. This framework can cover, for example, costs of buying mobile phones, as they can be justified as auxiliary device. However, mobile phones are becoming common also with elderly, so many have them already. NFC enabled mobile phones are still rare, but expected to get more common in forthcoming years.

There are several possibilities for covering the costs of the NFC tags. Firstly, the cost can be added to the cost of the medication, and charged from the user. If the user gets added value from the service, s/he might be happy to cover the cost of the tag (some ten cents). Also, the cost can be covered by the pharmacy. The pharmacy might see the service as an added-value service that can be used to improve their customer service, if they can justify the added cost by increased sales and better customer satisfaction. However, the service could also form an accessibility service that would be covered by the healthcare framework just as auxiliary devices needed by people with special needs. In this case, the payment structure could, for example, be that healthcare framework would pay the pharmacy a fixed sum from each NFC tag added into the medicine delivery. In this case, the payment model could boost up service adoption, as the pharmacies would be motivated in instructing their users in adopting the service, as it would improve the service they provide and thus could improve customer satisfaction.

The NFC tag writer could be part of the information system investment of the pharmacy. Compensation from public healthcare framework might be possible, if the public authorities would like to push the adoption of the service. However, as the purchase of the tag writer is a one-off payment (some hundred euros), the costs caused by daily use of NFC tags and related added working time are probably more significant from the service adoption viewpoint. If the service would indeed improve medicine adherence, this would probably have a positive effect on public healthcare expenses. Therefore, it might be justified to cover the costs of the service from public healthcare framework.

4) Availability of needed components: The NFC tags, writers and mobile phones are already commercially available. However, the selection of NFC enabled mobile phones is limited in most parts of the worls. Only in Japan, mobile phones with NFC functionality are relatively widely available. In Europe, the only NFC enabled mobile phone available is the Nokia Series 40 6212 classic phone. Unfortunately, there are no speech synthesizers available for Series 40 platform. According to the estimation of Finnish and Spanish associations for blind, most vision impaired use S60 platform phones because the availability of speech synthesizers and other supporting applications. However, as the penetration of NFC into mobile phones and other devices is expected to increase within a few years [17], it can be predicted that NFC functionality and speech synthesizers will be available in same platform in near future.

The scenario requires specific software functionality both in the pharmacy information system, and in the mobile phone. The research project described here has implemented one version of the mobile phone side software application. This is available only for research purposes. However, distribution and exploitation of the software will be one task of the research project towards its ending.

To optimize the functionality and time needed at the pharmacy, the software functionality should be integrated into the information system used at the pharmacies. The

functionality is simple, and the required modification would most probably be rather small. In Finland, there are two leading pharmacy chains that both have their own pharmacy information systems. Therefore, the modification would be needed only in two information systems. The most critical decision for that to happen is if the pharmacies decide to adopt the service or not. The work described here is one step in a research initiative aiming to provide pharmacies justification whether or not to adopt the service.

The information contained by package information leaflet is accessed through an URL pointing to a medication information database. Such databases do already exist and are freely accessible in Finland. However, this might not be the case in all countries and for all languages.

B. Scenario 2: Integration of medication management with a background system

1) Effectiveness as distribution channel: The second

scenario requires a service provider that would be interested in providing the medication support service for its clients. In Finland, the service providers are mostly small companies or third sector organizations with little or no co-operation with each others. They each have their own information system infrastrctures and service provisioning strategies. Therefore, the adoption of the service desribed in the scenario 2 would be dependent on thousands of small service providers all over the country.

From the user point of view, the service would be available only for those who can access the services of a service provider offering this service for their customers. This would require that the user lives on the operating area of the service provider. Even though it would be technically possible to provide the service from a distance, it would complicate the service interface for the end user.

2) Flexibility: The scenario expects that the user will follow the predefined schedule for taking medications. It is not very flexible with timings. This might be problematic in cases, where user cannot or does not want to follow predefined schedule, for example, because is travelling.

However, the scenario does allow the user to store the medications as s/he wishes. Even though the reminder part of the scenario works only for medication that is needed regularly, it can help also in monitoring medication intake with other types of medication. However, it can be speculated if the user is motivated to use the service only for providing information for monitoring, or is it reasonable to expect that the user would remember to use the service in such cases, when there is no immediate benefit for the user, and the usage pattern differs from the normal routines of using the service.

3) Cost model: The second scenario is based on service provisioning. The service provider will have costs for producing the service, and would then charge the end user of the service. The service provider would need to attach NFC tags into medication packaging. When making the link

between the physical package and the medication information in the background system, the service provider needs a NFC tag reader which can read the ID number of the tag. However, this can be done with a normal NFC phone. Therefore, the technical parts are not too expensive. The most part of expenses would be caused by the work needed to create medication schedules in the background system. This requires manual work from service provider employees. Also, quality assurance needs to be strict.

In this cost model, it is critical that the costs caused by the service can be covered with pricing level that is still acceptable for the end user. Also, the total revenue of the service provisioning needs to be profitable for the service provider. This would require both covering the costs and having a big enough customer base.

4) Availability of needed components: The NFC tags, tag writers and readers are already commercially available. However, as explained in the analysis of the previous scenario, speech synthesizers are not available for NFC phones currently commercially available in Europe.

Most elderly care providers in Finland already use computerized information systems for managing customer information, and service process management systems are also widely available. Several systems also have components covering medication information of the clients. However, integration with NFC marked medication packaging has not been implemented and is not readily available in commercial information systems targeted for elderly care providers. Also, the heterogeneity of the information systems is great. There are lots of small information system providers and competing and complementing information systems and used in organizations using elderly services. Therefore, wide adoption of functionality described in the scenario 2 would require changes in lots of information systems.

VI. CONCLUSION This paper describes the results of exploratory research that

created service scenarios for medication management for elderly users. The scenarios were constructed through co-creation process that involved all actors of the service process. The scenarios were analyzed from technical and practical viewpoint. The goal was to analyze how the scenarios could be implemented in practice, and what changes or interventions would be needed to make them practically feasible. We will continue this work by evaluating the scenarios in field settings, and analyzing the effect of the proposed scenarios on medication non-adherence and perceived satisfaction.

ACKNOWLEDGMENT This work was done in the HearMeFeelMe project

which has been partly funded by Tekes, the Finnish Funding Agency for Technology and Innovation. HearMeFeelMe is a project within the AAL (Ambient Assisted Living) joint programme.

REFERENCES

[1] K. Vainio, M. Airaksinen, T. Hyykky and K. H. Enlund, ”Effect of Therapeutic Class on Counseling in Community Pharmacies,” The Annals of Pharmacotherapy, vol 36, pp. 781 – 786, May 2002.

[2] K. Ishihara, S. Ishihara, M. Nagamachi, H. Osaki and S. Hiramatsu ”Independence of older adults in performing instrumental activities of daily living (IADLs) and the relation of this performance to visual abilities,” Theor. Issues in Ergon. Sci., vol 5, pp. 198 – 213, May-June 2004.

[3] K. George, R.A. Elliott and D.C. Stewart. “ A Systematic Review of interventions to Improve Medication Taking in Elderly Patients Prescribed Multiple Medications,” Drugs Aging, vol. 25, pp. 207 – 324, 2008.

[4] E. Becker, V. Metsis, R. Arora, J. Vinjumur, Y. Xu and F. Makedon. “SmartDrawer: RFID-Based Smart Medicine Drawer for Assistive Environments”in Proceedings of Pervasive Technologies Related to Assistive environments (PETRA’09). Corfu, Greece. June 9-13, 2009.

[5] M. Hoogendoorn, M. Klein and B. Mosch, “Online Application for Simulating Intelligent Support for Medicine Intake,” in Proceedings of Pervasive Technologies Related to Assistive environments (PETRA’08). Athens, Greece. July 15-19, 2008.

[6] B. Windham, M. Griswold, L. Fried, G. Rubin, Q-L. Xue and M. Carlson, “Impaired Vision and the Ability to Take Medications,” JAGS, vol. 53, pp. 1179-1190, 2005.

[7] L. Palen and S. Aaløkke, “Of Pill Boxes and Piano Benches: “Home-made” Methods for Managing Medication,” in Proceedings of Computer Supported Cooperative Work (CSCS’06), pp. 79 – 87. Banff, Canada. November 4-8, 2006.

[8] S. A. Safren, E. S. Hendriksen, N. Desousa, S. L. Boswell and K. H. Mayer, “Use of an on-line pager system to increase adherence to antiretroviral medications,” AIDS Care, pp. 787 – 793, Volume 15, Issue 6 December 2003.

[9] M.S. Giron, H.X. Wang, C. Bernsten, M. Thorslund, B.. Windblad, and J. Fastbom, “The appropriateness of drug use in an older nondemented and demented population,” J Am Geriatr Soc, 49(3), pp. 277-283, 2001.

[10] P. von Heideken Wågert, J. Gustavsson, L. Lundin-Olsson, K. Kallin, B. Nygren, B. Lundman, A. Norberg, Y. Gustafson, “Health status in the oldest old. Age and sex differences in the Umeå 85+ Study,” Aging clinical and experimental research, 18(2), pp. 116-126, 2006.

[11] R.A. Elliott, “Problems with medication use in the elderly: an Australian perspective,” J Pharm Prac Res, 36, pp. 58-66, 2006.

[12] J.M. Ruscin, and T.P. Semla, “Assessment of medication management skills in older outpatients,” Ann Pharmacother, 30(10), pp. 1083-1088, 1996.

[13] A. Beckman Gyllenstrand, “Medication management and patient compliance in old age,” Thesis for doctoral degree (Ph.D.), Karolinska Institutet, Stockholm, Sweden, 2007.

[14] M. Marshall, “A brighter horizon,” Community Care Supplement: Inside Dementia, 29, pp. 2-3, 1998.

[15] B. Sheehan, E. Burton and L. Mitchell, “Outdoor wayfinding in dementia,” Dementia 2006, 5(2), pp. 271-281, 2006.

[16] M. Isomursu, J. Häikiö, A. Wallin, and H. Ailisto, “Experiences from a Touch-Based Interaction and Digitally Enhanced Meal-Delivery Service for the Elderly,” Advances in Human-Computer Interaction. 2008.

[17] ABI research. ABI research, “Twenty Percent of Mobile Handsets Will Include Near Field Communication by 2012.” Posted to: http://www.abiresearch.com/abiprdisplay.jsp?pressid=838. 2007.

[18] Directive 2004/27/EC. available from http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:32004L0027:EN:HTML. Accessed on 11th September 2009.

[19] B. Mager, “Service design as an emerging field,” In: Designing Services with Innovative Methods. University of Art and Design Helsinki. Finland. 2009.

[20] H. Bayer and K. Holzblatt, Contextual design: Defining customer centered systems. San Francisco: Morgan Kaufmann Publishers. 1998.