Apple's healthy CareKit revolution

The CareKit framework enables developers and doctors to collect copious amounts of data -- and make sense of it all.
  • Ryan Faas (Computerworld (US))
  • 15 April, 2016 20:02

Around three years ago, I was talking with the CIO of a regional hospital about how mobile apps could improve hospital and post-discharge experiences for patients. The conversation came up largely because of the Affordable Care Act, which included new rules to lower Medicare reimbursements to hospitals with high rates of readmissions for people with specific conditions. The goal: Encourage better discharge and follow-up care that would mean healthier patients and lower medical costs. We specifically talked about how a discharge app could be a major improvement in the discharge process and ensure better follow-ups for patients.

Anyone who's ever been, or has had a loved one, discharged from a hospital (surgery or not), can relate. The do's and don'ts, the packet of prescriptions and lists of providers for follow-up care can be daunting for even the most simple of hospital stays. An app is a great option to improve medical outcomes.

I remembered that conversation earlier this month as I watched Apple introduce CareKit and the Postsurgical care app created with it by Texas Medical Center. The app is similar to what that regional hospital had hoped to create, but had never been able to build from scratch.

Apple COO Jeff Williams spelled out why CareKit and the apps it will enable are important at the same March 21 event during which Apple unveiled the new iPhone SE.

"One of the most important things affecting the outcome of surgery is what you do in the recovery process, yet we go from being monitored by a team of highly trained specialists using leading-edge technology to being discharged with a single sheet of paper," Williams said. "This is your list of things to do, not to do, which days to do them on. Adherence to this is notoriously very poor."

jeff williams carekit 100652382 orig

Apple COO Jeff Williams.

CareKit provides a solid framework for any healthcare organization to build upon. Apps are based around a series of modules -- four at present -- designed by Apple. Developers can start with these modules and build off them, customizing the data collected as needed. (This is very similar to how ResearchKit apps are developed.) The apps can get data from the sensors in an iPhone and/or Apple Watch as well as from connected devices like blood pressure cuffs, but they can also acquire information much more simply from user input.

This modular approach means that developers don't have to reinvent the wheel for common tasks or functions, reducing the resources required to create care-related apps. (Not surprisingly, the hospital officials I was talking to in 2013 are planning to take a serious look at CareKit.)

That tactic not only provides a level of consistency across CareKit apps, but also allows them to exchange information with each other or with ResearchKit apps and means they can interact with HealthKit data -- as long as Apple ensures that users allow such data sharing.

CareKit's four modules

CareKit offers developers just four modules for now. That may seem like a small number, but there is significant flexibility in how they can be used individually and integrated together. Although all four modules are available, it appears that Apple is allowing the use of only one or two if that's all that's needed for an app to accomplish its mission.

The Care Card module is essentially a treatment plan or medical to-do list. It can serve as a reminder to take meds, perform physical therapy, change bandages, perform self-tests (such as measuring heart rate/blood pressure, check range of motion, etc.), record information about conditions and let users note they've performed the tasks. It can also provide additional information like warning signs or things like food, medication or activities to avoid. Data in the Care Card module can also be updated by healthcare providers during an office visit or based on data shared by the app itself. This is a major component of Texas Medical Center's Postsurgical care app.

Care Card Apple

Think of the Care Card module as essentially a  medical to-do list.

The Symptom and Measurement Tracker does exactly what its name implies: It allows users to record information about symptoms and other metrics related to a given condition. That's a broad description, but tracking different conditions can mean tracking very different things, including mood for users with depression; range of motion for those recovering from surgery or musculoskeletal injures; pain levels; vital signs like blood pressure or lung function; body temperature; and even photos of wounds or skin abnormalities. Additional relevant data could also be captured, either by being manually entered or recorded on an iPhone. (An asthma or COPD-related app could, for example, automatically record air quality data based on a user's location.)

Capturing this data allows users and healthcare providers to see trends over time and to build a much more accurate picture of a person's progress than you'd get from just periodic office visits.

The Insight Dashboard is probably the most significant module because it makes sense of the data recorded by the first two modules. At its most effective, it can combine those data sets and provide a better understanding of how treatment is working. The act of tracking adherence to a treatment plan or to medication needs is helpful in its own right. So is the act of tracking symptoms. Put them together and you can get a much fuller picture of the effectiveness of treatment and/or medication. Is a given pain med providing relief? Do the prescribed physical therapy exercises improve movement and/or relieve pain? Is an anti-depressant or anti-anxiety medication actually effective?

Finally, the Connect module provides simple but key functionalities; it shares the information the app collects. Typically, this will be with a healthcare provider or care team, but it can also include a partner or spouse, parents, or anyone else a person trusts and who needs to know how that person is doing.

CareKit and HealthKit

Like Apple's HealthKit and ResearchKit, CareKit isn't a product in itself. It's a set of frameworks that enables a clear and consistent approach to tracking and managing data related to specific conditions. HealthKit, however, offers a single dashboard -- the Health app -- where users can track all the data collected by HealthKit-enabled apps and devices.

Although CareKit apps will almost certainly be able to read and write data from the Health app, Apple doesn't appear to be creating a single dashboard for CareKit-related information. This makes sense given that each CareKit app can, and almost certainly will, contain its own Insight Dashboard module specific to the conditions and metrics it tracks.

One advantage of HealthKit is that it facilitates information sharing with healthcare providers' systems. That's largely because of HealthKit support in apps from various manufacturers of electronic medical records systems and/or patient portals. They allow HealthKit data (once authorized by a user) to be passed through an app to the associated system where it becomes part of a user's medical record. (Data can also be passed from such apps into the Health app.)

It isn't clear how much data, particularly Insight data, will be shared this way, though any data a CareKit app writes into the Health app could be shared automatically. Exactly what level of automatic sharing is allowed and whether it's processed through HealthKit or directly communicated to a provider's systems remains to be seen; it will likely vary from app to app.

Page Break

Provider vs. third-party commercial apps

One of the interesting things about CareKit is that apps will be created by two broad categories of organizations. The first is healthcare providers themselves -- hospitals, medical groups, insurance companies, physician practices, etc. Texas Medical Center's Postsurgical care app is an excellent example (as is that discharge app concept I discussed early on). And then there are third-party developers, companies that produce health-related apps for the general public but aren't associated with a specific provider. Both will provide value, but there will be some notable differences.

Apps developed by a provider will more likely allow a user to share data with and receive updates from a care team. Since these apps will be designed to work with a specific hospital, medical group or practice, they will tie into the backend systems used by those providers (including, but not limited to, electronic health records, patient portals and internal communications systems). The shared data will also likely be mapped to a provider's clinical workflows, meaning that doctors, nurses and others will be expecting to receive the data, have processes in place to review it and have a plan for how to respond if needed.

These apps will prove to be immensely valuable to users and clinicians because they will be tailored to a provider's processes, specialties and the needs of patients. And healthcare interactions will be richer and more useful because the apps are more tailored to specific needs based on patient conditions or the specialties of the providers.

They will, however, have a narrower focus because the data and interactions will be centered around specific health issues. Sharing data with other providers (such as a primary care physician) may not be as easy as with the hospital-based care team or specialty practice. The audience for these apps will also be limited to the patients of a specific provider.

Apps developed for more general use will obviously have a broader audience since they're designed for a mass market, available to anyone that needs or wants to track a given condition. This makes them more accessible, but it also means that more user setup will be needed to customize the app for specific needs and treatment plans. There will likely also be less direct interaction with a provider or care team, meaning it will be up to users to get more involved in their own care. That could involve simply printing out the data generated by an app; some existing health-tracking apps that are planning to integrate CareKit already rely on this method for sharing data.

This won't make these apps any less useful. In fact, broad availability could be an advantage since individual medical practices or doctors in smaller healthcare groups will be able to suggest apps to a range of patients and may even be able to walk them through initial setup during an office visit. The mass market could also allow apps to be tailored tightly to specific conditions, particularly rarer ones, than provider-generated apps.

Obviously, broad availability also means that individuals will be able to select and use the apps without input from any healthcare provider while still offering useful data and insights that can be shared as needed (and with any/all healthcare professionals rather than just a subset). That process, however, will likely be less automated and directed more by the user than by a physician or care team.

CareKit's challenge -- making use of the data

Without a doubt, CareKit provides a lot of insight into a user's health and conditions, particularly those that are chronic or progressive. It offers a simple way to take a wide swath of information from a variety of sources, correlate it and present it in easy-to-digest formats. In short, it makes all this data meaningful. The trick is making it actionable. That's where the ability to share data and discuss it with clinicians is key.

The challenge here is that physicians and other providers need to be ready and willing to go over these insights with patients. For apps developed by a healthcare provider, this is pretty much assumed. For mass-market apps, this may not be the case. There remains a debate in the healthcare community about the value of patient-generated data as well as concerns about being able to absorb and analyze large sets of data.

The brilliance of CareKit is that it is designed to ensure accuracy by relying on sensors and connected devices to record data and to make that data meaningful by correlating multiple sets of information, including subjective information like how a patient feels. This provides more useful information than multiple separate sets of raw data.

An overall challenge will be getting healthcare providers to actually use those insights beyond simply looking at the data itself. This may very well be an area where patients need to educate healthcare professionals about what CareKit is, the apps available related to their specific needs and the insights these apps can provide. Sharing the information from an app's Insight Dashboard during an office visit would be a very good starting point.

Why CareKit is revolutionary

One of the remarkable things about CareKit is that it provides a consistent, template-based approach to creating apps that eases the development process and offers users an extremely shallow learning curve -- figure out how to use one CareKit app, and you'll likely find others familiar and accessible.

Another is that the framework allows for immense versatility. Some of the apps already announced use the framework to track a wide range of health issues, from postsurgical care to depression (focusing on the effectiveness of various treatments), pregnancy, diabetes, Parkinson's Disease and other chronic conditions. It's hard to imagine any physical or mental healthcare issue that CareKit couldn't help monitor and manage.

A third important point is that CareKit, like HealthKit and ResearchKit, isn't limited to just patient responses to questions or reminders. It has access to the various sensors of an iPhone as well as connected devices and additional apps that are HealthKit-enabled. That provides a great range of objective data as well as subjective data. By being able to track and correlate this data over time, CareKit apps provide a unique living view of a person's health, progress and issues that impact a given condition.

Ultimately CareKit is one of the most revolutionary technologies introduced for the overall health ecosystem because it enables developers to connect so many sources of information and make sense of them. Broadly adopted, it will allow many people greater understanding of their unique health issues and needs, particularly those with chronic or difficult-to-treat conditions. It offers an accessible path to more personalized medicine by building on Apple's existing health initiatives and the work of a disparate range of developers, device makers and healthcare providers.