Enhancing Fulfillment Data in Community Practices for Clinical Care and Research

In ambulatory care, there are two major forms of prescription data. Prescribing data represent what clinicians have prescribed for patients—ideally, the intended medication regimen. Fulfillment data represent what patients have received from the pharmacy—the actual medication regimen. Nearly all current medication research uses payer-based medication fulfillment billing records. With the rapid expansion of e-prescribing, electronic health record (EHR) fulfillment data may be more relevant for clinical care and research; however, many questions remain about data accessibility, quality, and utility.

This project evaluated the accessibility and utility of EHR fulfillment data within the Distributed Ambulatory Research in Therapeutics Network (DARTNet). DARTNet is a practice-based network that includes 32 independent and geographically dispersed organizations encompassing more than 1,700 clinicians and 4 million patients.

The specific aims of this project were as follows:

• Use surveys and interviews to assess the actual status, organizational plans, and barriers for full e-prescribing, capture of fulfillment data, and clinician use of fulfillment data at all DARTNet organizations. 
• Assess the data’s comprehensiveness and clinical utility in five DARTNet organizations receiving fulfillment data through the e-prescribing-based process, the consent-based process, or both. 
• Develop and pilot test a patient-level report using clinical, prescribing, and fulfillment data to improve the management of hypertension during the clinical encounter, with subjective assessments of utility by survey and group interviews of clinicians in one DARTNet organization capturing fulfillment data. 

A Web-based survey found that medication fulfillment data was not available at most practices. Interviewees cited the cost of obtaining and integrating medication fulfillment data into the EHR as a barrier. Data extraction identified significant data quality issues originating from a third-party data repository. Lack of confidence in the completeness data severely limited the ability to make conclusions from the available data. Despite barriers to accurate data extraction, a clinical decision support (CDS) tool that alerted physicians to potential lapses in adherence to anti-hypertensive medication was piloted at three DARTNet sites. Although the CDS produced a high number of false positive alerts, providers found the drug adherence decision support pilot valuable and worth expanding.