According not reach. For example HIV patients, who have

According to EGV 98% of Estonian population have some kind of mutation in genetic variation for drug treatment. It means that depending on the individual there is always some kind of medicine that will not work for the person depending on the situation or disease.

Adverse drug reaction (ADR) is unwanted effect caused by taking the medication. In  Estonia there were 257 ADR reports created and sent to the Estonian Agency of Medicines or directly to the distributors in 2016. The ADR reporting has been trending over the years and not only in Estonia. In the U.S. ADR morbidity rate was reported to be 0,32% in 1998 and causing the systemic problem to be a 4th cause of deaths during that time. In Estonia, the numbers are hard to collect as reported ADR incidents are not the total number of unexpected incidents and intoxication statistics are often reported along with accidents.

Accidents and intoxications are among the top leading causes of deaths in Estonia (15%), following cancer (25%) and cardiovascular diseases (40%). When looking separately statistics for intoxications in 2015, the morbidity percentage from the drugs were 25% cause from all the intoxications. Although we do not have detailed statistics for ADR in Estonia (home and hospital), we can compare the general statistics and assume the majority of the problem. Author of this paper has also personal experience with ADR incidents in Estonia which supports the interest to the topic and it will be used in examples at user-case scenarios. In general ADR has been handled by clinicians in Estonia.

Celsius healthcare is offering a warning system Inxbase and database Riskbase for patient’s medication prescribing. The solution is used by all Estonian doctors and pharmacists and funded by Estonian Health Insurance Fund. While this solution manages the highest possible risk for ADR cases in the controlled environments, like hospitals and pharmacies, there are still several cases where Inxbase and Riskbase can not reach. For example HIV patients, who have requested to not show or reveal their disease status to their general practitioners (GP) and so the possible medication mismatch for drugs is not known: antiretroviral (ARV) medication combined with antidepressants (AD). It is not possible to control the ADR incidents at home environment, where the doctors and pharmacists can not reach to offer their consultation on what medication can or can not be used together within one family’s home pharmacy. There is a consultation phone number “1220” for instant access to the GPs, but this number is typically used when trouble has advanced, not beforehand. As patient-empowerment is emerging among well educated communities, the need to self control and prevent ADR incidents is also rising. Our persona – 35 year old mother in Estonia is well educated and has adopted digital solutions very well, including smartphone apps.

As she is mother for young children, she is still active in her career and might be that in some point the father of the children took the opportunity for parental leave. Still the family health and home pharmacy management is on the shoulders for the mother and at times delegated to the father. In this paper the author is going to introduce a solution to improve ADR control for the whole family and the general public by mHealth solution – MediKeep. Types of adverse drug reactions.According to the Merck Consumer Information, there are three types of adverse drug reactions (ADR). Starting with the dose-related adverse drug reactions, where there may be so called drug side effects and they are typically predictable but sometimes unavoidable.

Drug-drug interactions (DDI) are also important topic here, where for example another drug slows the metabolism of the first drug and thus increases its level in the blood. Dose related side effects are usually not serious but are very common. The second type of ADR is called allergic reactions.

They are not depending on the dose but  need a prior exposure to the medication. It can develop when the body’s immune system creates inappropriate reaction to a drug. They can be predicted with allergic tests. Idiosyncratic adverse drug reactions are the least ones to be understood and are largely unpredictable. They are the most serious but do not happen that often. Here pharmacogenomics (PGx) can come to help to give dosage and metabolism suggestions based on the individual DNA.

MediKeep – the biohacking tool for smart devices.Behind every smart solution there is a complex system.  In the digital era, technology and digital services are the providers for those complex systems. Machine learning and artificial intelligence can offer personalised solutions. According to the Pew research in U.S., more than 51% of smartphone users have been looking for health information on their device when connected to the internet.

In the 2014 study by Nielsen, it is reported that about one third of the smartphone users were using fitness and health apps. The author of this paper can predict those numbers rising as mobile acquisition is still trending and the markets are not so different in Europe or U.S. Smartphone is a personal and individual device, just as is personal and individual – our health.

The job-to-be done for the mother in Estonia is to take care of the family while there are health events happening, considering everyone’s personal differences. She needs health information, tracking tools, to be reminded and delegate if necessary. She also does not want to have several apps to manage bits for her problems. Health is very much time depending and can vary from time to time – but in general this information needs to be logged into one place to be able to predict and prevent based on the previous information. The person oneself is responsible for the health outcomes but is not possibly the best person to make decisions. As GP’s are loaded with work in Estonia and can not handle detailed health information for every patient – technology comes to help. MediKeep simplifies everyday health management with affordable science (DNA test) and technology (smartphone, smart device), not only by offering a place for digital home pharmacy inventory and medication reminders, but possibility to analyse the substances adverse reactions as well.

Not leaving out pharmacogenomic interpretation (PGx). It offers simple and effective solution for smaller problems like finding the nearest pharmacy (also when travelling globally) and making personal health notes while creating a history of events on the timeline (health diary log within the app). MediKeep offers DNA genotyping together with personal recommendations in a language that everyone understands – from one person to another – in pharmacogenomics, food metabolism and for the prevention.User cases and job-to-be-done.Let’s take a look at user case where the mother needs to keep track on her own health and medication. Typically the family medication box with small children is quite voluminous.

According to the MediKeep’s own research in 2016, family with 4 children (ages 2-8) had to throw away 31 medication boxes within one year due to the expire dates, while another 30 were kept. The box contained around 60 medication packages in general. Typically over the counter (OTC) medication do not have serious adverse reactions, but the ADR risk increases with combination and amount of the medicines taken regularly for the chronic diseases.

 Nevertheless, even in family mother who is taking birth control pills may not be aware of adverse reactions or DDIs. Father may not be aware of medication interactions with his blood pressure. It may be hard to track the amount and frequency of fever medication given to the small children during their cold and flu periods.

To reduce the risks described in ADR section of this paper, she has to add (can be simplified with scanning the barcodes from the packages and link them with the drugs in the MK database) all the family medication to the digital inventory within the app. It is important that the active substance in present. As the database includes also drug information leaflets in PDF’s, it is easy to read the ADR options any time (if the leaflet is missing). Every time the medication is added to the digital inventory, the MK system is checking for possible DDI’s. DDI system can be any local or international system like Inxbase and Riskbase or similar.  Now the family caretaker has full overview of their home medication, separate medication leaflets and possible drug-drug interactions.

MediKeep also includes pharmacogenetic (PGx) interpretation service to offer personalised medicine information and where digitalisation of the medicine cabinet is not a prerequisite but adds extra value. Let’s say that the mother has done genetic test some time ago and now can download her DNA raw data for her own use from the portal. MediKeep offers an online platform to upload the data and link it with the account in MediKeep and the person. Now using another service for PGx interpretation the mother can check if there is a suggestion or alert for any active substance she is taking or planning to take in the future. Same can apply for any member of the family as the app allows to create different profiles and upload different DNA raw data files.

The digital inventory then can be filtered by:Family member profile.Use of the drug (hypertension, cold and flu, pain, chronic disease etc).The app offers 3 levels of ADR services:Single medication information leaflet for side effect information.Drug-drug interaction analysis.PGx interpretation based on individual DNA.Marketing.

MediKeep solution is very much location dependent, because of the different laws by countries and medications used, also databases available. Medication management itself is not a popular topic among majority of the people and the healthy ones. Thus the company’s team has predicted the PGx solution and genomic interpretation to be the most desireable. Along with the pharmacogenetic and food metabolism interpretation the app can send personalised articles within the dashboard feed.

It means that the app can engage people by sending out also “fun facts” about the persons what people consider very interesting and entertaining, still learning about themselves on the go (food taste preferences, caffeine consumption, restless leg syndrome, sleep preferences, food and nutrition metabolism, stress handling, athletic performance, resistance to HIV, baldness, sense of pleasure, sensitivity to anesthetics, dependences and carvings and many more). New phenotypes are discovered by the scientists daily. Typically the reports generated by different competitors are too long to read or hard to understand at once. MediKeep offers a subscription plan for the users to opt in for learning course about self. Once or twice a week, the app will send one feed article for the user to read, learn and consider, about topic related to the individual, while important facts about PGx and metabolisms are sent out at once.

This is the main differentiation between MediKeep and its competitors, while the retention of the app is predicted to rise.   MediKeep DNA interpretation service will be also accepting DNA raw file formats from other services like, and more, as there is no standard file format on the market yet. MediKeep aims to be one of the first companies to offer DNA raw data download in VCF format (popular among academic institutions) along with the MediKeep own .

txt format. Purchasing methods.As in Estonia the public health purchasing is managed by Estonian Health Insurance fund there is probably no options to offer the solution along with the insurance service. For example in Netherlands, where people need to choose extra insurance from the private insurance companies, the solution could be offered as side tool for cheaper insurance expenses for the consumer.

While the user has proved to be dedicated user of the app, the insurance plan could have a smaller fee. Same could apply in Estonian private insurance providers but as they are not used commonly, another method needs to apply. In Estonia the OOP purchasing method is considered for the applications revenue model. It means that the consumer has the option to buy services directly from MediKeep as a private company. There are two options to do that. Monthly or annual subscription plans or “pay once lifetime fee”. MediKeep can also offer genotyping service in collaboration with a laboratory in Tartu who are working with the latest Illumina GSA chips that has a revenue stream of its own and needs to be considered separately from the application services.

In the Table 1, author of this paper offers in-app monetization structure, where based on the subscription model, the content and value is added accordingly.Table 1. MediKeep in-app monetization structureUSER TYPEFEEDAD free contentPGxFree account or guestPersonalisedarticles–Free account with DNA raw data filePersonalised articles-?Premium account Plan APersonalised articles, no ads? -Premium account with DNA raw data filePlan BPersonalised articles, DNA interpretation? ?Risks.The main risks for mHealth applications are the security issues, as personal health information can be attractive for thieves and hackers. According to the Keith Katz, app developer at Kony, Inc., this kind of data “can be used in fraudulent financial transactions”.To manage the risks, they provide four-step risk analysis by Health Insurance Portability Accountability Act (HIPAA). Where at first the application developers should evaluate likelihood and impact of potential risks to protected health information (PHI).

Secondly to implement appropriate security measures to address those risks and document those security measures. For the last the developers should maintain continuous, reasonable and appropriate security protections. The table 2 shows different kind of risks and how they could be managed:Type of risk in increasing order of severityMain stakeholder affectedSample scenario where this risk could ariseWhat can be done to manage this riskLoss of reputationProfessional/organizationApp displays sensitive performance data about professional or serviceGood securityLoss of privacy (patient confidentiality)PatientPoor security of patient dataEncryptionLose phone holding patient dataAvoid holding patient data on mobile devicePoor quality patient dataPatient/professional/ organization (eg, financial data)App allows bad data to be entered into patient record or retrieved from it at handoverData validation on entry and retrieval from authenticated sourcePoor lifestyle or clinical decisionPatient/professionalBad patient data used in risk calculation algorithmCheck correct data retrievedBad knowledge or search toolCheck algorithm properly codedBad advice or algorithmUse proven health behavior change methodsPoor risk communicationInappropriate but reversible clinical actionPatient/professionalPoor medication adviceTest quality of advice on sample dataProvide facility for user feedback and respond to thisInappropriate and irreversible clinical actionPatient/professional/ organization (liability exposure)Bad algorithm controlling insulin pump, surgical robot, radiotherapy machine, etcAdopt safety critical software design and development methodsExhaustively check design and test algorithm & user interfaceConclusion.Despite of the risks that mHealth apps bring, they have also potential to transform healthcare and achieve the triple aim of improving the pa