I track my exercise – but then what?

By F. Daniel Nicolalde, Project HealthDesign - PhD Student - UW-Madison.

Like most cyclists, I own a small bike computer that allows me to keep track of my speed, cadence, and distance.  Unfortunately, tracking my exercise can still be a rather difficult task as I still need the motivation, payback, and technology to translate this data into meaningful information.

Bad
Last season I used a time consuming paper-based bicycle journal to keep track of my rides. I managed to have three entries during the whole biking season.  The only tangible data I walked away with was a running count of kilometers that the bike computer stored – over 1,000. 

Better
Before this year’s biking season started, I purchased a heart rate monitor that allows me to maintain an online profile of my exercise routines.  All the data that the system stores is primarily quantitative - taken straight from the heart rate monitor and the bike computer. Unfortunately, the system has no automated upload so I have to enter all the information manually.  Thus far I have entered more than twenty entries.  While the system has limitations, it displays the data in a couple of ways.

Hopefully the future holds the best
It was my goal to share my exercise data with my personal health record. This task hasn’t proven to be straightforward.  The online system exported an XML file but my PHR wasn’t able to interpret the content, so now I have an XML file in my PHR with all my exercise information on it.  Storing health information in proprietary information system creates silos of information, which can only be interpreted by the system that generated it.  As a biker, a consumer of health technologies, and exercise gadget owner, I’d like to echo what Patti Brennan said in her last post, we should ask for true interoperability and not be bonded to one particular vendor.

If you keep track of you exercise routines, I’d like to know how and what motives you to maintain the record. Should these types of observations of daily living be part of our health records? And if so, with whom should we share them?

So many phones! Which one to choose? How can I tell?

Selecting a device for monitoring patient activity and collecting patient data

By Jay Levinsohn, BreathEasy Technical Lead, RTI International.

Read all the BreathEasy posts.

The Android platform
The BreathEasy project has been in the process of determining what kind of devices would be appropriate and useful for collecting health information in a patient population.  Early in the process it was determined that a smart phone was the device most likely to be useful.  In today's environment it quickly became apparent that selecting among the emerging Android based smart phones would provide a set of very able technologies. 

The general Android platform and development environment was selected for several reasons:

• it is an open system with a growing development environment
• there are many vendors for both hardware and software
• the hardware platforms have adequate screen size for data collection
• it supports multitasking
• the software development tools are either free or very inexpensive

Specifications and more specifications
Given all the available choices in the Android marketplace, we quickly realized that we needed to determine which specs would be most important to the project. We suspected it would be impossible to get everything in one phone. After conferring with the team, we concluded that the most critical function of the phone would be data collection.  All of our criteria are detailed in our phone specifications table.


Testing our first phone
Given these dimensions the candidate phone that was selected for full scale testing was the Google Nexus One. This phone met our criteria and has a very fast and capable processor.  The device has a large screen with high resolution and runs under Android 2.1.  It has a removable 1400 mAh battery that provides 5-6 hours of internet use time.  It has good memory and external storage capability (512 MB ram and a 4-32 GB microSD card). Our developers are now busy testing the phone’s performance in many different settings. Stay tuned and we’ll let you know what worked well for us, and what left something to be desired.

An Example of Our Regulatory/Assurance Function in Action

By Bob Belfort, Manatt Phelps & Phillips, LLP.

Over the past few weeks, my Manatt colleagues and I have met by phone with each of the Project HealthDesign grantees to review how existing state and federal laws governing the privacy and security of health information impact their projects, and what steps they should take to ensure compliance with those laws.

We talked about things like minimizing security risks associated with loss or theft of, or improper access to, portable devices and ensuring transmission security of information between portable devices and other computer systems. These requirements generally stem from the HIPAA Security Rule, which requires covered entities to employ administrative, physical, and technical safeguards to protect the confidentiality and integrity of protected health information maintained or transmitted electronically. 

To Encrypt or Not to Encrypt:  Text Message Transmission Security
We had a very interesting discussion with one grantee about health care providers’ use of text messages to send reminders and alerts to patients participating in their project.  It highlights the type of judgment calls that are often necessary when interpreting one’s responsibilities under HIPAA and state privacy laws.

Under the HIPAA Security Rule, a covered entity has discretion to adopt security measures within specified categories based on the entity’s size, complexity, capabilities and resources.  In addition, while certain security measures are required, others are “addressable,” which means that a covered entity has the flexibility, through a formal security risk analysis, to assess whether the measure is “reasonable and appropriate” in its particular environment and, if not, to adopt an alternative reasonable and appropriate measure.

As many of you know, there are obstacles to encrypting text messages.  Because the security measure that protected health information sent over an electronic communications network be encrypted is an “addressable” one under the HIPAA Security Rule, we engaged in a mini risk analysis during our call, weighing the risks of sending certain types of information through unencrypted text messages and considering how best to mitigate these risks.  One mitigation strategy we discussed was limiting the nature of the information transmitted. We talked through the risks associated with texting “Don’t forget to take your medication today” as compared to “You forgot to take your Lipitor today,” and discussed a variety of other potential scenarios.

The bottom line is that there is simply no bright line to guide a grantee’s decision whether to encrypt or not to encrypt, and, if not to encrypt, what types of alternative measures to adopt to ensure the security of the information being sent via text.  For example, while limiting text messages to “pointers” (e.g. “Check the Carrot.com for a note from your health coach.”) is probably the most conservative way to minimize the risk of unauthorized access to protected health information or other information sent via text, it may not be the most effective way to motivate patients to take control of their health.

This is one of the reasons we at Manatt and our partner Deven McGraw at the Center for Democracy and Technology are here:  to talk these types of questions through with grantees and assist in finding solutions that minimize legal risks but also work in practice.  We’ll be addressing these types of risks during standing monthly calls. We’re looking forward to some interesting conversations.

You’re a PHR vendor? Really?

By Patricia Flatley Brennan, Director, Project HealthDesign.

From the HIT tradeshow floors to the halls of HHS, we’re hearing LOTS about PHR vendors—what is their product? Can it be certified? Should it be HIPAA protected?

But what is it? At Project HealthDesign, we see PHRs as suites of services - platforms, programs and data stores – united in a way that helps individuals manage their health and health care, and perhaps form the very core of crowd-sourced health as in the Asthmapolis.

Vendors of HIT products and PHR tools will play an increasingly important role in mapping out the landscape of personal health information management – but nowhere on the terrain do I see a single vendor for a freestanding PHR taking hold!

It’s time to stop looking for the turnkey PHR and start realizing that the idea of a single PHR as the sole point of intersection for all data, services, and visualization tools needed by a lay person to manage his or her health is so 1999.  It’s time to recognize that the suite of tools needed by one person to manage asthma or handle diabetes might well be quite unique to that person’s life process, health state and personal preferences, and quite different than that needed by another.

All is not lost! Think for a minute why the analogy of a PHR as a great big binder of data is so pleasing – it allows for a single point of entry, it enables extension of known privacy and security approaches, and it most likely tethers patients to a care provider.  All of these advantages are still possible in a distributed model of PHRs, in which data from various sources are integrated on the fly into a single view to help a person to know what action to take now. Privacy and security tools can follow the data, not become the responsibility of the data holder.  Visualization tools can become enablers, not end points.  And the link to the clinical provider is fostered by the value of the care choices enabled, not the ties that bind the data.

A PHR vendor? So 1999 - let’s party like it's 2010 and invite everyone in!

One Size Doesn’t Fit All

By Gillian Hayes, University of California, Irvine, FitBaby Principal Investigator.

Read all the FitBaby blog posts.

Personal health records (PHRs) are systems maintained by patients and their families to track health-related information and share that information with clinicians and other interested stakeholders.  They  are central to the way we might get observations of daily living from patients and their families into clinician hands.

However, in the United States, only 7 million adult patients make use of PHRs. So, one of my MS students, Leslie Liu, set out to understand what combination of usability issues and functionality concerns were limiting people’s ability to use PHRs and their willingness to adopt them. In this work, we conducted both in-lab usability evaluations and interviews with medical professionals and patients surrounding three PHR systems:  Microsoft HealthVault, Google Health, and WorldMedCard.

 
Our results suggest that substantial usability challenges and the complexity of PHR systems can lead people to believe they “may not actually save time” by using them or that they would only be useful if a patient has a “serious health condition”.  Furthermore, the current state of these systems includes minimal flexibility and customizability for the specific needs of the different stakeholders.

What’s more, even physicians in favor of patient-managed data in our study were worried about the balance of burden and benefit. For example, one physician told us in an interview:

“…most of that information is really pretty easy to get on a first visit …. if they’re…conscious. …the only advantage would be if a patient kept it up to date... But [the PHR systems shown] almost seemed too simplistic…”

The results of our research suggest that a big part of the problem is that these systems have primarily been treated as monolithic solutions capable of serving a wide variety of stakeholders without substantial personalization or customization. Everyone is trying to solve the problem of managing records for everyone and every condition. The problem is, this model does not always allow for full consideration of the complex set of issues that emerge in health care and the disparate interests of the patients, family members, friends, clinicians, and staff members involved in their use.

As PHR systems evolve, there continues to be substantial need for addressing patient-provider communication, collaboration, and information sharing through adaptable and flexible collaborative systems to effectively advance collaboration and to improve health care delivery and outcomes.

In the FitBaby project, and in all the Project HealthDesign projects, we will be using PHR systems to store and manage the massive amounts of data that will come out of collecting observations of daily living. The PHR is essential to our ability to collect, store, and access these data. The question for us is, how can we develop visualization and alerts on top of these systems to allow for the kind of customization and flexibility that the results of our research on PHR suggest we need.

We’d love to hear from anyone out there using PHRs (or not!)… tell us how things are going?  What can we do to make these systems fit you the individual and your needs?

The Clock is Ticking

By Amireh Ghorob, iN Touch Program Manager, San Francisco General Hospital.

Read all iN Touch blog posts.

Imagine Maria, a 19 year-old patient, on the exam table at a routine 15-minute physician visit. Maria has been recently diagnosed with diabetes and is struggling with her weight. Her last two blood pressure readings were high. She is worried that she may have a sexually transmitted disease because she just discovered her boyfriend has been cheating on her and has been feeling very down lately. Assuming she was able to express all these concerns, her primary care physician must decide which issues to tackle now and which must wait until the next visit. In today’s health climate, it is impossible for the 15-minute visit to address acute, chronic, and preventive care, build meaningful relationships with patients, and manage multiple diagnoses. 

Put Me in Coach!
Enter the “health coach”: a member of the health care team other than a physician, nurse-practitioner, or physician’s assistant. At the Family Health Center in San Francisco, health coaching was launched in 2006. In 2008, the Teen and Young Adult Clinic at the Family Health Center implemented Takin’ Charge, matching health coaches with 16-25 year old patients at risk for cardiovascular diseases. 

Health coaches work to help patients become active creators of their care plans. They are trained to provide self-management support, health education, assistance with behavior change and medication adherence, and between-visit care. They interact with patients in the clinic (before, during and after a visit)  and have demonstrated that members of the health care team can successfully engage patients by using phone calls, texting, email, and in-person coaching sessions between-visits.

Game Time
A person spends a very small part of their life being a “patient.” Consequently, the health care team gets only a glimpse of people’s lives. Health coaching expands the 15-minute visit and engages the patient more. But coaches face the same challenges as the rest of the health care team:  difficulty getting in touch with patient, gaining the trust of patients, and missed appointments. Collecting ODLs will give the coach more insight about the person who is their patient.  But how will collection of ODLs impact patients? Will it increase patient engagement, foster more trust and improve adherence? 

Ready. Set. Go.
Health coaches rely on patient recall when assisting with behavior change action plans during their weekly or biweekly interactions. By asking patients to upload ODLs pertaining to exercise habits, dietary choices and stress, we want to capture real-time lifestyle choices and the patient’s experiences as they occur.  We hope this will portray a more accurate picture of the patient’s life as they are living it. Will this help patients and health coaches collaborate in the creation of manageable and successful action plans? What other advantages may occur?  And are there disadvantages we can anticipate and plan for?    

Is there a relationship between Observations of Daily Living and Activities of Daily Living?

By Monica Tentori, Post-Doctoral Scholar, FitBaby
Read this post is Spanish

In the FitBaby project, we are working on developing a comprehensive solution that will show relevant observations of daily living (ODLs) for both clinicians and parents of pre-term infants. As part of this effort, we started talking about how activities of daily living (ADLs) are related to the concept of ODLs and where there might be some tension between these two ideas.

In eldercare, a common goal is measuring and understanding ADLs: The things we normally do in daily living, including any daily activity we perform for self-care. The beauty of ADLs is that they allow us to assign a functional score on a common scale. ODLs on the other hand, are sentinel observations that lack an explicit tie between indicator and function.

What follows is a typical response from the mother of a low birth weight infant describing an average day:

My girls usually get up about 7:30 and play around on the floor for maybe half an hour to an hour - reading books and playing with toys. Breakfast then takes about 30 minutes because they’re starting to get into finger food. They eat their yogurt or cereal and then they play with the puffs. They play for another hour at home, usually by themselves at that point, in their room. And then they take a nap for about an hour. After that, they have their lunch.  After lunch, we go for a walk to the park for about two hours. When we come back, they get their snack and milk and take another nap. In the evening we play with their toys, read books and take a bath. They then go to sleep.

 
What you see here is a lot of discussion about what we are classifying as baby ADLs (eating, bathing, playing, and so on) and almost nothing about what we might typically call ODLs (sleep quality, mood,  appetite and so on). Indeed, when analyzing our data from interviews and focus groups over the last several months, we found out that from 275 occurrences of indicators actually related to function associated with the tracked ODLs by moms, 75% include babies’ ADLs.

Our team isn’t the only one exploring connections between ADLs and ODLs. The Embedded Assessment project of CMU is observing things like opening a pillbox, making tea, and other standard activities one might measure to detect cognitive decline in elders. In many ways, tracking data on how well an elderly patient performs these activities, or at least how consistently, is both an ADL and a sort of meta-ODL that captures several ODLs at once.

To me, it seems as though ADLs can be used as an instrument to monitor some ODLs while ODLs can provide context and meaning to caregivers about ADLs.
 
What are your thoughts?

Sensing Technology Can Observe it All - Almost

By Anind K. Dey, Carnegie Mellon University, Embedded Assessment Principal Investigator.

In our project, we are focused on sensing activities of daily living from which we want to create measures of observations of daily living. One of the activities we are trying to sense is how well our subjects take their medicine. This involves almost every stage of the process:

• picking up the pillbox
• opening the appropriate slot for the current day of the week
• taking the pills out of the slot
• getting some water
• using the water

We are interested in the quality of the performance of each of these activities. However, it should be pretty obvious that the one thing missing from the list and that we can't sense without using invasive video (which our subjects would not accept) is whether the pills are actually consumed. In searching for a solution, we have discovered some very interesting technologies. While they aren’t available today, one day, in the not-too-distant future...

Proteus Biomedical makes smart pills that are powered by the stomach's gastric juices. Once powered, a digestible microchip embedded inside a pill reports out an id. The id is picked up by a microelectronic receiver that is worn as a skin patch or implanted under the skin. The receiver can then send the information out to a cell phone.

Researchers at the University of Florida have developed a similar system using pills with a microchip and antenna, and created the company eTect. The antenna receives energy from a transmitter worn, say on a wrist, and sends it's unique id, along with data collected from the microchip like the time and the type of pill that was ingested. The transmitter can then communicate that out to an external device for recording.

Researchers at Georgia Tech have developed MagneTrace, a necklace that measures the strength of the magnetic field in the neck area. In this system, tiny magnets are embedded in pills, and when a pill is swallowed and travels down the esophagus, the necklace measures the change in the magnetic field and determines which type of pill was swallowed and how many. The necklace then sends that information to a smart phone or nearby computer. An alternative form factor for the sensing device is a simple patch. It was 94% effective at correctly detecting the magnets in a lab evaluation.

Very cool technology, but they will all require, arguably small but time-consuming changes in the ways that pills are manufactured. In other words, it’s exciting to follow the advances, but they won’t be available for our work. What will we do? Unfortunately, we may just have to skip that aspect of pill taking, as we don't have a solid technology with which to track actual pill ingestion.