Seventy-two percent of primary care providers reported that electronic referrals improved overall clinical care compared to prior methods. Several factors, such as perceptions of better communication as manifested by improved tracking and increased access manifested by decreased wait times, may have contributed to this impression. We found that nearly 60% of participants reported that access for non-urgent patient issues had improved, supporting our hypothesis that electronic referrals would improve access to care. Participants reported decreased wait times; this perception was born out by administrative data (data not shown). Features that may have enhanced communication were the improvements in tracking, standardized referral templates, and the centralized review. These may have improved the quality and timeliness of information available to subspecialty reviewers through standardization of communication and equitable access to reviewers. Improving the quality and availability of clinical information has been shown to improve the likelihood that the subspecialist responds to the clinical question.14,26,27
Providers had better guidance of the workup, which may have improved effectiveness of the consultation. Improvement in the ability to track referrals may have improved accountability and prevented duplication of efforts.
We were concerned that electronic referrals would disrupt workflow, which could have impaired the perception of improved care. Workflow challenges present important barriers to adoption and sustainability of health IT.28
We found that a significant proportion of Consortium clinic providers reported taking longer to submit electronic referrals than other providers; the length of time for submission was independently associated with lower satisfaction with overall clinical care. The lack of available computer terminals, multiple-step procedures to access the secured electronic referral portal, and frequent disconnections likely contributed to this increased time spent submitting referrals. We did not find that Consortium providers reported less affinity for newer information technology, increasing the likelihood that the finding of increased dissatisfaction was related to structural barriers.
We found evidence of the impact of the “digital divide” among safety-net health systems on acceptability.29
We are attempting to address these barriers by garnering additional resources to improve computer access and connectivity in Consortium clinics and by encouraging those who are able to submit quickly despite the barriers to share their expertise. Yet, the findings within our study point to a larger phenomenon of “digital disparity” in access to and usability of health IT. Our findings are consistent with prior studies showing that interventions using computerized decision support, electronic referrals, or electronic health records have had variable uptake in part due to the heterogeneity of patients, practices, and resources especially among safety net health systems.30
Our study had important limitations. First, we used web-based questionnaires to ask participants about electronic referrals, which may have selected participants who had greater willingness to use IT. However, we mailed paper versions of the questionnaire to non-responders. Second, we relied on primary care provider recall of the referral system prior to electronic referrals, which may introduce recall bias. Third, this study focuses solely on the referring providers’ perspective, and relies on their subjective reports of time spent submitting and wait times. We did not weight responses by clinic setting because clinic setting was not the only way to categorize the variability in our participants. We may, however, have underestimated the discontent among non-respondents, as those from the Consortium were both less likely to report that electronic referrals had improved clinical care and were less likely to respond. Because we were interested in the acceptability to individual participants, we did not adjust responses for panel size or number of referrals per participant. Fourth, we analyzed results from providers referring to a single hospital in a safety net health system, including clinician investigators and residents who, while working full time, spend only a few half-days a week of engaged in direct outpatient care. However, because we included primary care providers from three different health systems, our findings may generalize to other community- and university-based providers who work in safety net settings.
Electronic referrals offer the opportunity to improve the effectiveness and efficiency of subspecialty care without inappropriately restricting access. Integrated health-care delivery systems that do not rely on fee-for-service reimbursement may be fertile ground for attempts to improve the efficiency of subspecialty care, thus improving access.31,32
We found that digital disparities among safety net health systems depend less on individual provider preferences than on access to and the usability of health IT. Future interventions should incorporate primary care provider and referring clinic experiences in evaluating the adoption and spread of electronic referrals. In our study, safety net primary care providers reported that electronic referrals improved not only access to specialty care, but also the overall clinical care of patients. Our findings suggest that IT innovations can be instrumental for safety net settings in attaining the “holy grail”33
of ensuring access to timely, high-quality subspecialty care.