The results of future analyses to evaluate PHRs could be vastly improved by the inclusion of certain types of additional data. The effectiveness of outreach efforts could more easily be measured if, when registering, beneficiaries were asked for a referral source. This could take the form of a general survey question offering multiple choices for source of referral. Alternatively, specific outreach events could provide numerical codes corresponding to that event, which a beneficiary would then use to register for MyPHRSC. Customized URLs or referral links could be used in a similar manner.
Additional demographic data may be useful in conducting multivariate analyses to predict usage of PHRs. In particular, data on household composition, income, assets, or beneficiary race and ethnicity, for example, could reveal predictive variables that may better explain beneficiary use of PHRs. Information on utilization of health care services may also be linked to usage of PHRs. An examination of beneficiary utilization of health care services will be included as part of the evaluation of the Medicare PHR Choice pilot program.
Data on page views that are aggregated daily or perhaps even by session can provide a significantly richer information source than page views aggregated quarterly. Ideally, a user's navigational path through the system would be captured so that beneficiaries’ activities could be compared to look for differences in usage inside the PHR product.
Finally, a satisfaction and/or usability survey could provide valuable information on usability or other challenges, and could potentially explain patterns of use and nonuse highlighted in this analysis.