3 Comments

1. 

   David C. Kibbe, MD MBA on April 13, 2009 at 10:21 am

   In all honesty, John, I think the morass is not Google Health, but the data sources that our Dave was provided as feeds to Google Health. This is an important time not to blame any of the messengers, but to coolly assess why Dave didn’t get the patient-centered outcome he wanted.
   Regards, DCK

2. 

   Gilles Frydman on April 13, 2009 at 11:04 am

   I absolutely agree with you, David.
   This is a fundamental time and we must bring all the skeletons out.

3. 

   David C. Kibbe, MD MBA on April 13, 2009 at 3:59 pm

   Gilles: So, let’s talk about what happened to e-patient Dave. He wanted an accurate, up-to-date, summary of his health care experience to be transferred from a hospital where he had been treated, to his Google Health PHR account.

   Kudos to both Google Health for creating such a service, and to the hospital for creating the interface to make the data transferable. And to both of them for making it work, even if imperfectly.

   Here’s where we run into trouble. Instead of relevant information in nicely organized groups of data – such as diagnoses, a list of medications, his allergies, procedures, and pertinent lab results – Dave got a data dump that was largely incomprehensible. It was also clearly not filtered so as to be current, accurate, or well-organized.

   In short, he was amazed, because the data was pretty much useless in the form that it was rendered in after passage to Google Health.

   Where’s the responsibility? How might this be done better in the future, so as to surprise and delight e-patient Dave, rather than surprise and dismay him?

   First, I think the hospital leadership should have studied the model that Google Health is using to create personal health records, and committed to filtering their own data stores to comply more cleanly and fit more easily into the receptacle/container that Google Health has built. If they didn’t like that container, then why partner with Google Health in the first place?

   Second, I think Google Health ought to use better judgment when it partners with hospitals, to assure that the data isn’t simply dumped helter skelter into its app. At the very least Google Health (and I love these guys) has been somewhat naive about interfacing sets of data between a hospital’s system and its own. This has created an uproar of criticism against Google Health, which is unfair. But this was bound to happen. Google Health was the paper cup at the end of the shute from a 75 foot high data silo. They need to find a way to avoid this happening with other partners.

   e-patient Dave’s doctors might have held his hand while performing this exercise the first time. It was bound to disappoint Dave, right?

   To me, the issue of coding systems is a red herring. There’s nothing wrong with using ICD-9 codes for problems and diagnoses, even if these are generated from a billing system.

   However, if there is an up-to-date list of clinical codes and descriptions available somewhere in the records that is reconciled by the discharging physician, or a nurse that is appointed this task, then this ought to be the data passed to Google Health or any PHR, in preference to the billing information. Right?

   I’m sure there are more skeletons here. But the real issue is how to provide Dave and patients like Dave a patient-understandable summary health document in both machine and human readable formatting, which can be passed on to the next provider, to a PHR, to the cell phone.

   Regards, DCK