Category: Application Development

A Peek at CareKit

The ability to interact with patients in multiple ways is an important facet of quality care.   The new Comprehensive Primary Care Plus tracks acknowledge this. Track 2, in fact, makes monitoring of patient outcomes a requirement.

Document and track patient reported outcomes CMS is evaluating a patient reported outcome survey instrument that will be sent to CPC+ Track 2 patients to identify specific care needs requiring intervention/management by the CPC+ practice site team. CMS plans to use the data collected from the patient-reported outcome survey to develop a patient-reported outcome performance measure that may be included in CPC+ measure set in the later years of the model. The modes of administration are yet to be determined.

  1. The health IT tool should provide the care team with the ability to administer the survey, store and track patient responses, and score results longitudinally for each patient surveyed.
  2. The practice should be able to review the patient responses/results in their EHR or other health IT tool and, as appropriate, establish care plans /interventions for positive findings.

 

CMS CPC+ Track 2 Requirement

CMS Plans to track outcomes and expects Track 2 practices to be able to do fairly sophisticated things with the data captured.   Ideally, the outcomes data should be seamlessly integrated into the EHR or other HIT system. This is not a simple request.   One of the biggest challenges is how to administer the instrument in a way that is secure and convenient for the patient.   Here is where CareKit could come in handy.

CareKit is an application development framework for iOS. Using it, vendors can create clinical survey instruments that users could run on an iPhone or iPad.   Data entered can be encrypted and stored on the phone and shared with practices.   Aside from giving developers a jump-start in creating patient-facing data collection tools, CareKit might be useful to developers exploring clinical UI design and app feature sets. The sample apps included are excellent examples of good design thinking.

CareKit is modular, which makes adding new data or features relatively straightforward.

MainCareKitModules
CareKit architecture (courtesy of Apple, Inc.)

From this diagram, one can see the separation in functionality between the data store and user interface.

Care Card holds the patient care plan—the series of interventions that will occur.  When an intervention has been completed, the patient uses Care Card to indicate when the event (e.g., performed an exercise, took a medication, etc.) occurred. Care Card presents the full list of required interventions, which allows the clinician and patient to have a complete view of what is expected and when.

Symptom and Measurement Tracker (SMT) provides the tools needed to collect detailed data from patients about their status. Data can be subjective (surveys, questionnaires, pain scales) or objective (readings from instruments or sensors).   Using SMT features, practices can gather detailed information about how patients are progressing and spot potential problems faster. For example, SMT would be a great way to catch adverse drug events or side effects. Every drug dose would be logged in CareCard, and SMT would display an instrument (survey, form, radio buttons) that queried about possible drug-related signs and symptoms.

Insights is basically a report module. It can be used to display charts, graphs, or other output that would allow patients and clinicians to view the care plan’s progress.

Contacts holds all those with whom care plan results might be shared.

From a developer perspective, I like the way the Care Plan Store functions. It allows storage of interventions, assessments, and results. Four types of care plan data are stored: intervention activity, intervention events, assessment activity, and assessment events.   As might be expected with a well-designed framework, activities and events can be added without coding. This makes it possible to create a library of activities or events and download them to the app as desired. Further, CareKit generates events automatically for any activity added to the care plan. Thus, if the activity is taking one’s temperature before each dose of medication, adding a “take temp” activity to the care plan will result in CareKit generating a “took temp” event on CareCard that the patient would use to indicate the temp has been done.

I see numerous possibilities for CareKit apps. Since activities and events are not hardcoded, there is significant potential for standardization and sharing.   It makes sense that CMS, professional societies, or simply practices that work together would create activity libraries.   De facto care plan standards could arise simply from a sufficient installed base of apps. One could imagine care plans for common diseases being shared among practices.

From a primary care standpoint, CareKit apps are exactly what I wished for while practicing.   Every flu and cold season, I would be inundated by patients wanting to be seen for URIs. Having a large number of patients show up with colds helped no one, and usually gave colds to the staff and me. It would have been so wonderful to have a URI app that patients could use to share their symptoms without having to come for a visit or go to urgent care.   Alerts for temps over 101, vomiting, diarrhea, shortness of breath or other symptoms that might herald serious cases, would have allowed me to see those who were really sick while reassuring and managing the rest. Essentially, less worry with better care.

CareKit has a lot of potential, especially given the push by CMS.   CareKit will make patient-facing clinical apps easier to build, and I hope easier to integrate into backend clinical care systems.   It points the way to next-generation clinical care delivery and is a much more friendly way to interact with one’s primary care clinician. Now, if only Apple would bless a workflow engine or buy a BPM company…

Facebooktwittergoogle_plusredditpinterestlinkedinmail