Privacy for the patient, convenience for the patient and medical providers, and access to anonymous data for medical researchers are all important and in many ways conflicting goals.

I think one aspect of a good solution is for identifiable patient data to be stored on a USB key or similar device. That way the patient must be present and must approve of all data access, and the patient can also access the data himself at any time. Some/most/all access could also involve an online service which stores non-private patient info and anonymized patient info.

The device could also be used to securely transmit data between providers. For example your doctor might put a cryptographically signed prescription on your key, which you take to the pharmacist to get filled. The doctor (or you) could also transmit the prescription electronically. When you get to the pharmacy, your key confirms that the prescription is for you.

I think a big problem in this space is liability, as a medical provider you have legal protections that help defray some of the liability associated with breaches in data, and more importantly subpoena power from the feds. Any data a user gives to you has to be given freely by the patient, bypassing HIPAA, and thus making that data available to parties it would not otherwise be available to.

This is the real reason only the big guys like google and MS are really trying this out IMHO, because they've got a big checkbox to face the potential legal ramifications of this problem.

You should try contacting people with regular health problems like people with obesity, diabetes, aids.

Another group that can provide you good ideas is high-performance athletes.

You definitely need something that is a device or is installed on a mobile so people can add information without having to move to a website.

On the website you could show progress, suggestions and recommendations based on the aggregated data.

good advice. you might also check some existing tracking apps in your area of interest.

we consulted with some Swiss marathon champion and diets specialists when worked on running and body tracking apps

You should start reading up on HIPAA regulations if you will be hosting any of your users' private medical data.

This is one of the big reasons that only Google and Microsoft are dueling it out over PHR standards--they have major legislative and regulatory affairs departments and fleets of lawyers to make sure they are covered legally.

(disclaimer: I'm working on an app in this space)

If you're really serious about building a better Google Health, why not apply at Google and try to actually build a better Google Health?

Unfortunately, there's nothing simple about modeling and analyzing useful medical data. This is a very hard problem to solve.

You might find your efforts better used in joining someone who's already made a lot of progress, like we have at PatientsLikeMe.

http://www.patientslikeme.com/about/careers

So this space is good enough for you to be in, but not good enough for new competitors to be in? Jeez.

If we both care about solving the same problems, I'm definitely happy to extend a hand.

The largest problem with Google Health for many people is that their medical provider does not share data with Google so it is very difficult to get your data in. I think this would be difficult for a solo start up so you might look into providing easy ways for the user to create data, such as a mobile app like Instant Heart Rate.