Apple has been making headlines again, which is not unusual for a company that sets trends and moves markets. This past June Apple announced that iOS 8 would have a built-in app to collect all of your health info in one place. Then there was the announcement of the partnership with Mayo Clinic and Epic Systems.

This generated even more headlines. I saw all kinds of opinion pieces published, ranging from predictions that this will be the 'game changer' that mHealth has been waiting for, to those who said, "Not so fast. What is really novel here?"  I let the dust settle, processed all of this, and came up with the following reflections.

It's impossible to give an educated opinion because all we've seen so far are some hints at what the software will do. We have no idea what the hardware play will be and we don't yet know all of the planned software capabilities. However, there are some things I believe we can count on.

Apple has an amazing track record of creating superbly designed, intuitive software and beautiful, flawlessly integrated hardware. By contrast, virtually all software created for health care users (providers, patients and administrators) is poorly designed. If Apple offers some breathtaking software and very sexy hardware to help us stay healthier, it could make a difference, or at least point the way for others as the iPod and iPhone did.

Also, hats off to Epic and Mayo Clinic for showing us the way on the integration of patient-generated data into the electronic health record. If the platform gets traction with Mayo clinicians and/or if they can show patient-generated data being used to improve health outcomes or lower cost, this could be a very big deal.

I can't help but be optimistic. But I also can't resist playing armchair quarterback and offering some advice as the effort goes from screen shots to reality. With so much opportunity to be transformative, I hope they take up that mantle seriously and don't repeat the mistakes of recent connected health history. Here is my wish list for Apple as they launch HealthKit.

1. Please don't think that a health data repository on a mobile device will be transformative. Google discontinued its health data repository and Microsoft's has gotten very little use. We (society, health care experts, providers) have not given consumers/patients enough of a reason to make the effort to store all of their health data on one platform. People don't feel compelled to take ownership of their health data.  The only compelling use case is the traveler who gets ill and that just isn't enough. Yes, HealthKit will make the collection of health data mobile and most likely passively captured. But overcoming these consumer barriers will not be enough to assure widespread adoption, I fear.

2. Please don't make it just about seeing your doctor's notes/lab data, etc., on your mobile device. This would be under-imagining the potential of this type of platform. I'm assuming that with the wave of wearables and Apple's interest in health tracking, we will see heavy integration of patient-generated data. But, assuming is dangerous, so this is my explicit plea to Apple to do lots with patient-generated data, both collected via a device and self-reported.

3. Please do employ analytics on all of the data streams to feed insights to users in order to help us to improve our health. This seems obvious, but I worry. At the Center for Connected Health, we have accumulated lots of evidence that personalized engagement messaging is what will make something like HealthKit sticky over the long run. I don't see evidence that Apple has done this.

Apple does OK at best, while companies like Google, Amazon, Netflix and Facebook live and die on their analytics ability. Other than iTunes, can you think of a software application that Apple created that is superior? Apple excels at many things, but any time they've been challenged to use analytics to target messaging and personalization, it has not gone so well. Just look at the example of Ping, Apple's attempt at a social network, or how iTunes Radio stacks up to Pandora. We know there is no comparison to great machine learning as it pertains to keeping individuals engaged. I don't see this as a core competency at Apple.

For me, it boils down to this: Will easy-to-use, intuitive, engaging software and beautifully designed hardware be enough to bring people into an environment like HealthKit and keep them there? Or will it take a killer app with analytics to drive personalization to keep people engaged (a la Netflix recommendations, Pandora's predicting songs for you or Google knowing exactly what your searching for after only three keystrokes).

If you are in the former camp, you predict Apple will be remembered for changing the game in connected health.  If you are in the latter camp, you may be seeing HealthKit go the way of Ping or Google's health data repository.

Joseph C. Kvedar, MD, is the founder and director of the Center for Connected Health, part of Partners HealthCare. The center works to create a new model of healthcare delivery, by developing programs and innovative strategies to move care from the hospital or doctor's office into the day-to-day lives of patients. This commentary was first published on the center's CHealth Blog.