Automatic Population of Data. In both the discussion groups and observation activities, beneficiaries consistently reported that the more data included in the PHR, the better. Although the population disagreed on the end limit of how much data was enough (e.g. whether beneficiaries would want to see MRIs even if they could not interpret them), the overwhelming consensus was to include as much data as possible in as clear a manner as possible. Similarly, entering data into the PHR proved difficult, as some beneficiaries were not computer-savvy, some of the terminology in the PHR was unfamiliar and the method for entering data was at times complicated. Beneficiaries found that it was time consuming and frustrating to enter the data themselves, and although some persevered and entered their data because they perceived the value of having access to the information to be greater than the loss of having to enter it, all beneficiaries would benefit from having more data pre-populated in the PHR and less self-entry.
Simple Login Process Necessary. A significant barrier to beneficiary use of PHRs was successful login. Many beneficiaries were not able to log in without assistance the first time, and many reported difficulties logging in each time. A login name and password that beneficiaries generate themselves would help beneficiaries to remember these items and enter them correctly. Additionally, while security and privacy were important to beneficiaries, they would prefer being able to keep their passwords for a longer period of time, if not permanently, rather than replacing the passwords every 90 days for security.
Address Inaccurate or Missing Data. Future claims-based PHR tools will need to be developed in consideration of the claims data process. While some claims are submitted on time and are free of error, many are submitted late and include inaccurate diagnosis codes. Some beneficiaries were confused by some of their diagnoses, while others did not see any data in the PHR because their claims had not yet been submitted. In order to ensure data in the PHR are accurate and is provided to beneficiaries in a timely manner, PHR developers should consider importing clinical data as an alternative to claims data. Because of the complicated diagnosis coding language used for medical billing and as a result of the fact that physicians may code things differently, claims data are highly prone to inaccuracies,. For instance, claims can contain incorrect, incomplete or outdated diagnosis and provider information. As a result, many physicians have raised concerns about the use of such data in PHRs because of the potential serious ramifications of acting on imprecise information,.
Strong Technical Support. A strong technical support team is instrumental for beneficiaries to be able to easily use the PHR. Therefore, such technical support should be included up-front as well as on an as-needed ongoing basis. Many beneficiaries in this pilot depended upon the telephone support of QSSI to guide them through logging in and navigating the system. Perseverance in using PHRs despite the evident barriers may be a factor of the level and dedication of technical support services offered. Although not all PHR vendors will be able to provide this support, NORC recommends developing PHRs using usability design and considering the end user first and foremost.