Standardized exercise codes towards database integration

Question

This is a submission I'm about to make to The National Center for Advancing Translational Sciences.  HealthVault appears to be very relevant, although I'd appreciate comments from forum users regarding database integration and standardized exercise codes - how relevant are they to integration?  I'm not an IT person, I just see various entities developing their own tools but seemingly without regard to the ability to integrate their data with other databases, which would be beneficial for research.  My interests are in exercise oncology but this issue applies to other areas of healthcare where exercise is a factor and where it is being recorded.  Thank you!

RE: NOT-TR-13-001 Request for Information (RFI): Enhancing Community - Engaged Research through the Clinical and Translational Science Awards (CTSA) Program

Regarding information technologies (IT), healthcare, and the possibility of database integration (1,2), there does not seem to be agreed upon standards for recording exercise or physical activity measures into electronic health records (EHR) or into other healthcare IT tools.  This will make utilizing database integration problematic for exercise advocates and will limit research in an era when data is being generated and stored daily by insurers, researchers, institutions, foundations, and by general public IT users via smart phones, apps, and exercise training devices such as Garmin, fitbit, and Microsoft's HealthVault, among many others.  To miss, or too delay, the opportunity to smoothly gather exercise data should not occur because of something seemingly as simple as establishing standard codes for exercise or physical activity.

The USDA uses Ainsworth (3) codes in their new Super Tracker device, which LIVESTRONG also appears to have incorporated into their individual user training logs.  ExeRxcise is Medicine, an initiative of the American College of Sports Medicine, has been involved in establishing an exercise "vital sign" (4) in collaboration with Kaiser Permanente Southern California to record minutes per week of moderate to vigorous exercise.  The Nurses' Health Study (5), uses questionnaires to record walking, running, biking, swimming, tennis, or aerobic exercise machine use, what their database codes are I don't know.  From personal conversations I understand that United HealthCare has a user-entered exercise tool but their codes I do not know either.  The American Institute for Cancer Research has a Continuous Update Project that endeavors to keep recommendations on physical activity, among others, current with new scientific evidence, however, these recommendations are dependent on interpreting each new research study's exercise measures.  There is a lot being recorded among just these stakeholders but standardized recording is lacking, which can limit translation of evidence as mentioned by Campbell (6).

Other stakeholders are developing healthcare IT tools that unfortunately lack exercise fields, while other tools have already been developed that lack exercise or physical activity fields.  Rectifying missing exercise and physical activity fields, and/or more so, standardizing the codes for these fields when they are included in new IT tools, should be an important initial step in a translational infrastructure.  This will require collaboration among stakeholders to establish codes that can then be given to IT developers for incorporation into healthcare tools.

(1) Studying Cost, Efficacy of Cancer Care. Cancer Discovery, September 2012, 2;757.

(2) Digiulo S. The Survivorship Care Gap: Delivery Models-What Works, What Doesn't. Oncology Times, September 10, 2010, Vol. 34, No. 17, pp. 11-13.

(3) Ainsworth BE, et al. Compendium of Physical Activities: an update of activity codes and MET intensities. Med. Sci. Sports Exerc., Vol. 32, No. 9, Suppl., pp. S498-S516, 2000.

(4) Coleman KJ, et al. Initial Validation of and Exercise "Vital Sign" in Electronic Medical Records. Med. Sci. Sports Exerc., Vol. 44, No. 11, pp. 2071-2076, 2012.

(5) Nurses' Health Study [Internet]. Harvard Medical School. Questionnaire available from:http://www.channing.harvard.edu/nhs/questionnaires/pdfs/NHSI/2010.pdf

(6) Campbell KL, Neil SE, Winters-Stone KM. Review of exercise studies in breast cancer survivors: attention to principles of exercise training. Br J Sports Med 2012;46:909-916.

Answer 1

Hi,

The HealthVault data model defines an Excercise data type. You'll notice the activity type is defined as a "codable-value", meaning the field can be coded with values from vocabularies. HealthVault does provide a built-in vocabulary for Excercise Activities. Of course, applications are not restricted to coding this data using only our vocabularies. Where there are industry standard vocabularies, those can be used as well to allow for better semantic interop of the data. I hope this is useful information for your submission. Happy to discuss furthur if you have any questions or comments.

Answer 2

Thank you for your reply and for the embedded links within it.

I still have a few questions that I hope do not seem redundant to you, as I posted, I'm not an IT guy, just an exercise advocate trying to move forward the integration of exercise data from as many sources as possible for the benefit of researchers and eventually the public.

It seems that already existing exercise IT tools, and those currently being developed outside Microsoft's HealthVault framework, will not be able to integrate unless they by chance are using the same 'name', and other codes, as HealthVault.  Is this correct?  If so, would they then need to write some program in order to merge their data into HealthVault?

Or, should other exercise researcher stakeholders decide to use different codes for exercise, such as F.I.T.T. (Frequency, Intensity, Time, Type) but coded as numbers, acronyms, and so forth, then HealthVault data would not be able to merge with them without HealthVault writing programs to translate their data into the other stakeholders' codes, correct?

How likely then are other stakeholders to want to use HealthVault codes as their standard, and does HealthVault want to be THE database for exercise in healthcare?  Or, is HealthVault willing to translate backwards into standard codes different than what they are currently using?  Who did HealthVault consult with regarding exercise codes for the database?  What other healthcare stakeholders were involved in the process?

Or, is all this moot, in that HealthVault could easily incorporate new standardized exercise codes into what appears to me to be quite an expansive tool, or vice versa for exercise stakeholders outside HealthVault?

Of concern to me is that there seems to be many new exercise IT tools but everyone is recording the data differently, and without knowledge of what others are doing, making future integration problematic.  Of course this is a research viewpoint not a capitalistic one.  However, how difficult would it be for all stakeholders to standardize their F.I.T.T. measures yet each present them to users in whatever form they deem marketable?

Does this better explain the dilemma I see, or have you actually answered it but I don't recognize it amidst the IT jargon?  I would like to see exercise stakeholders from insurance, medicine, research, institutions, the public, and IT all involved in agreeing on basic measures and standard codes for exercise, and then moving forward with them embedded in each new healthcare tool.

I look forward to any reply you may have.

An example of some of Compendium codes for those who haven't checked the references:

1993 2000

compcode METS compcode METS heading description

01009 8.5 01009 8.5 bicycling bicycling, BMX or mountain

01010 4.0 01010 4.0 bicycling bicycling, <10 mph, leisure, to work or for pleasure (Taylor Code 115

01015 8.0 bicycling bicycling, general

01020 6.0 01020 6.0 bicycling bicycling, 10-11.9 mph, leisure, slow, light effort

01030 8.0 01030 8.0 bicycling bicycling, 12-13.9 mph, leisure, moderate effort

01040 10.0 01040 10.0 bicycling bicycling, 14-15.9 mph, racing or leisure, fast, vigorous effor

01050 12.0 01050 12.0 bicycling bicycling, 16-19 mph, racing/not drafting or >19 mph drafting, very fast, racing genera

01060 16.0 01060 16.0 bicycling bicycling, >20 mph, racing, not drafting

01070 5.0 01070 5.0 bicycling unicycling

02010 5.0 02010 7.0 conditioning exercise bicycling, stationary, general

02011 3.0 02011 3.0 conditioning exercise bicycling, stationary, 50 watts, very light effor

02012 5.5 02012 5.5 conditioning exercise bicycling, stationary, 100 watts, light effor

02013 7.0 02013 7.0 conditioning exercise bicycling, stationary, 150 watts, moderate effor

02014 10.5 02014 10.5 conditioning exercise bicycling, stationary, 200 watts, vigorous effor

02015 12.5 02015 12.5 conditioning exercise bicycling, stationary, 250 watts, very vigorous effor

02020 8.0 02020 8.0 conditioning exercise calisthenics (e.g. pushups, situps, pullups, jumping jacks), heavy, vigorous effort

02030 4.5 02030 3.5 conditioning exercise calisthenics, home exercise, light or moderate effort, general (example: back exercises), going up & down from floor (Taylor Code 150

02040 8.0 02040 8.0 conditioning exercise circuit training, including some aerobic movement with minimal rest, general

02050 6.0 02050 6.0 conditioning exercise weight lifting (free weight, nautilus or universal-type), power lifting or body building, vigorous effort (Taylor Code 210

02060 5.5 02060 5.5 conditioning exercise health club exercise, general (Taylor Code 160

02065 6.0 02065 9.0 conditioning exercise stair-treadmill ergometer, general

02070 9.5 02070 7.0 conditioning exercise rowing, stationary ergometer, genera

02071 3.5 02071 3.5 conditioning exercise rowing, stationary, 50 watts, light effort

02072 7.0 02072 7.0 conditioning exercise rowing, stationary, 100 watts, moderate effor

02073 8.5 02073 8.5 conditioning exercise rowing, stationary, 150 watts, vigorous effor

02074 12.0 02074 12.0 conditioning exercise rowing, stationary, 200 watts, very vigorous effor

02080 9.5 02080 7.0 conditioning exercise ski machine, general

02090 6.0 02090 6.0 conditioning exercise slimnastics, jazzercise

02100 4.0 02100 2.5 conditioning exercise stretching, hatha yoga

02101 2.5 conditioning exercise mild stretching

02110 6.0 02110 6.0 conditioning exercise teaching aerobic exercise class

02120 4.0 02120 4.0 conditioning exercise water aerobics, water calisthenics

02130 3.0 02130 3.0 conditioning exercise weight lifting (free, nautilus or universal-type), light or moderate effort, light workout, genera

02135 1.0 02135 1.0 conditioning exercise whirlpool, sitting

03010 6.0 03010 4.8 dancing ballet or modern, twist, jazz, tap, jitterbug

03015 6.0 03015 6.5 dancing aerobic, general

03016 8.5 dancing aerobic, step, with 6 – 8 inch step

03017 10.0 dancing aerobic, step, with 10 – 12 inch step

03020 5.0 03020 5.0 dancing aerobic, low impact

03021 7.0 03021 7.0 dancing aerobic, high impact

03025 4.5 03025 4.5 dancing general, Greek, Middle Eastern, hula, flamenco, belly, and swing dancing

03030 5.5 03030 5.5 dancing ballroom, dancing fast (Taylor Code 125)

03031 4.5 dancing ballroom, fast (disco, folk, square), line dancing, Irish step dancing, polka, contra, country

03040 3.0 03040 3.0 dancing ballroom, slow (e.g. waltz, foxtrot, slow dancing), samba, tango, 19th C, mambo, chacha

03050 5.5 dancing Anishinaabe Jingle Dancing or other traditional American Indian dancing

04001 4.0 04001 3.0 fishing and hunting fishing, general

04010 4.0 04010 4.0 fishing and hunting digging worms, with shovel

Answer 3

You might find the following blog post about FitnessSyncer useful: http://blogs.msdn.com/b/familyhealthguy/archive/2012/11/18/welcome-fitnesssyncer-aka-why-ecosystems-are-awesome.aspx. I think it addresses a number of your questions with a practical example.

In this case the blog describes, there are a number of fitness services/apps each storing fitness data in their own representation, but only one organization needed to decide to map these individual fitness formats to a common format in order to have a single representation of the data. This single representation can now be consumed by applications without concern as to the source of the data to infer the format.

In this particular case the N fitness services include Garmin Connect, MapMyFitness, and RunKeeper and the common format is the HealthVault Excercise format.

Until there's critical mass for a single fitness data standard, you're likely to see a similar integration pattern where data from multiple sources is transformed as needed to a common format for consumption.

Ali

Answer 4

Ali,

Thank you, that appears to be a/the solution.  Now to address the critical mass problem, establishing fitness data standards can be initiated by a group, or collectively among groups, such as FitnessSyncer, HealthVault, or the American College of Sports Medicine (among the many stakeholders), otherwise establishing standards will be slow and there may develop a reluctance to collaborate once groups have put all their eggs into their own data baskets.  Getting on the same page shouldn't be too difficult if collective marketing of something like FitnessSyncer can begin.

All interested stakeholders on here please consider moving standardization forward.

This can be a part of reducing healthcare costs and contributing significantly towards research.

Thank you!