More communication = better blood sugar in Type 2 diabetes

150 veterans with type 2 diabetes in Pittsburgh were part of a randomized trial recently reported in Diabetes Care in which half of them got basic disease management (monthly calls from a nurse practitioner) and half got that PLUS home telemonitoring of blood glucose, blood pressure and weight. All the patients improved over 6 months, but the exchange of information between the patient and the provider resulted in a bigger drop - over 1% more of HbA1C. That's more than you usually get from adding a second or third medication.

Once again, the more the communication between the patient and the provider, the better the outcomes. Hats off to Drs. Stone, DeRubertis and company!  Now, the trick is to find the most efficient and effective way to do it.

Using Health Information to Find Bad Medicines Quicker and Cheaper

Here is a very valuable use of personal health information aggregated to a population level. John Brownstein and colleagues from Harvard used a network of electronic medical records systems to gather up over 34,000 patient records and confirm that the diabetes medication rosiglitazone (sold as Avandia among other trade names) is associated with an increased risk of heart attacks. Of course, that's not really news. What is even better is that they showed that if they had been doing this back when the drug was introduced (1999), they would have found the association in late 2000, rather than 2007, when the reports first appeared.  

Rapid identification of myocardial infarction risk associated with diabetic medications using electronic medical records by Brownstein, J. S., Murphy, S. N., Goldfine, A. B., Grant, R. W., Sordo, M., Gainer, V., Colecchi, J. A., Dubey, A., Nathan, D. M., Glaser, J. P., Kohane, I. S. Diabetes Care Journal publish ahead of print articles

Finding out years earlier that a commonly used drug has an important adverse effect is a very good thing and ought to be a big part of the business case for more exchange of information.

Contractual gag orders in EMRs - a petition

MedInformaticsMD and the good folks at Health Care Renewal  have launched an interesting and important electronic petition drive about Transparency and Openness in Electronic Patient Records and Other Healthcare Information Technology Systems. The key issue is that some manufacturers of electronic medical record systems and related technology put clauses in their service contracts that customers may not publically report or discuss bugs or problems in the system. Even truly disastrous problems that result in patient deaths or institutional mayhem are subject to these gag orders, limiting the ability of the medical and health IT communities to mitigate, repair or avoid them.


I share the belief that these kinds of restrictions are bad for patients, bad for providers, bad for the country, and in the long run, bad for the manufacturers themselves. So, I have signed the petition online, and suggest that you consider doing so, too.

Vermedx highlighted in VermontBiz.com

We're so proud! VermontBiz .com interviewed our own Charlie MacLean and did a great profile on Vermedx, so check it out!

Diabetes Registries for Public Health Surveillance

We missed it back in 2008, when it first came out, but this article from Washington state is worth going back for. The good people in the state health department linked together a bunch of clinic-based diabetes registries to get a bird's eye view of the quality of diabetes care across the state. This is an approach we used before in New York City, San Antonio and Vermont and we're glad to see it being used in the Evergreen State.

The difficulty with the clinical registry product that they used is its heavy reliance on manual data entry. To cover a broader segment of the target population at much lower cost, we use automatic data feeds from clinical laboratories. But, the idea is the same and its a good one - exchange clinical information with public agencies to improve their ability to manage the health of the state's population.

Exploring the feasibility of combining chronic disease patient registry data to monitor the status of diabetes care.

Prev Chronic Dis. 2008 Oct;5(4):A124. Epub 2008 Sep 15.

Kemple AM, Hartwick N, Sitaker MH, Harmon JJ, Clark K, Norman J.

Chronic Disease Prevention Unit, Washington State Department of Health, PO Box 47855, Olympia, WA 98504-7855, USA. angela.kemple@doh.wa.gov

INTRODUCTION: To provide direction and to support improvements in diabetes care, states must be able to measure the effectiveness of interventions and gain feedback on progress. We wanted to know if data from multiple health clinics that are implementing quality improvement strategies could be combined to provide useful measurements of diabetes care processes and control of intermediate outcomes. METHODS: We combined and analyzed electronic patient health data from clinic sites across Washington State that used the Chronic Disease Electronic Management System (CDEMS) registry. The data were used to determine whether national and state objectives for diabetes care were met. We calculated the percentage of patients that met standards of care in 2004. RESULTS: The pooled dataset included 17,349 adult patients with diabetes from 90 clinics. More than half of patients were above recommended target levels for hemoglobin A1c testing, foot examination, hemoglobin A1c control, and low-density lipoprotein cholesterol control. Fewer patients met recommendations for nephropathy assessment, eye examinations, and blood pressure control. In terms of meeting these standards, rates of diabetes care varied across clinics. CDEMS rates of care were compared with those reported by other data sources, but no consistent pattern of similarities or differences emerged. CONCLUSION: With committed staff time, provider support, and resources, data from clinical information systems like CDEMS can be combined to address a deficiency in state-level diabetes surveillance and evaluation systems--specifically, the inability to capture clinical biometric values to measure intermediate health outcomes. These data can complement other surveillance and evaluation data sources to help provide a better picture of diabetes care in a state.

Tweeting for Health

Recently, we've posted about collecting data from patients via brief text messages (SMS) over their cell phones. Today, we have communication in the opposite direction: sending prompts to patients to encourage certain behavior.

Text-Message Reminders to Improve Sunscreen Use: A Randomized, Controlled Trial Using Electronic Monitoring
April W. Armstrong, MD; Alice J. Watson, MD, MPH; Maryanne Makredes, MD; Jason E. Frangos, MD; Alexandra B. Kimball, MD, MPH; Joseph C. Kvedar, MD
Arch Dermatol. 2009;145(11):1230-1236.

This group sent the local weather report along with a reminder to wear sunscreen to 35 adults via their cell phones. After 6 weeks, sunscreen usage was 56% vs 30% among 35 control subjects (P<.001). The NNT (Number-Needed-to-Text) was a very low 3.8, suggesting that about 1 in 4 subjects responded to the messages. This is remarkably high success rate for such a low cost intervention and suggests that text messages might have broader applicability in health behavior change.

What other areas would benefit from a periodic cell phone reminder?  Where won't it work?