Article Title

Barriers to Implementing and Disseminating an Intervention to Improve Hypertension Control With Home Monitoring and Uploading of Data Into an Electronic Health Record

Publication Date



hypertension, electronic health record


Background/Aims: The CONtrolling Disease Using Inexpensive Technology (CONDUIT) study tests an approach to monitoring and managing hypertension that could be easily and widely disseminated and scalable to self-monitoring of other conditions.

Methods: The CONDUIT randomized control trial tests whether an intervention consisting of self-monitoring of blood pressure (BP) and a feedback loop involving nurses and primary care providers improves control of hypertension in patients with uncontrolled hypertension. Rather than embedding the intervention in a proprietary electronic health record (EHR), we used the free Microsoft HealthVault platform to receive participants’ BP readings electronically and developed an interface to transmit HealthVault data into the EHR. To further maximize dissemination potential, participants who could not upload BP data from home could upload data at their clinics.

Results: Substantial effort was required to develop the HealthVault-EHR interface, including custom programming to poll HealthVault for BP data and periodically send messages to nurses summarizing the patient-uploaded data. We encountered barriers to system implementation at multiple levels, often because vendors did not consider use cases similar to ours. For instance, BP monitors lacked unique device identifiers; neither the devices, HealthVault or the EHR validated date/time data; software changes by any of several entities caused data flows to break and required frequent revision of patient instructions; and protected health information protection in clinic-based uploads proved challenging. Patients and staff expressed satisfaction when the system worked, but had limited tolerance for software failures. Most clinicians supported the system, but would have greater enthusiasm if patient-generated BP readings were considered in HEDIS scoring.

Discussion: When the CONDUIT system worked as designed, it was well-accepted by patients and providers, but the various “moving parts” under control of different organizations led to multiple challenges and frustrations. For similar interventions to be successful, hardware and software vendors must consider a wider range of use cases in their design processes.




April 1st, 2015


April 28th, 2015