Optimizing Referral Pathways for Patients With Hematuria and Moderate-Severe Proteinuria - Phase 2: A Quality Improvement Project

Adult patients receiving care from a Geisinger PCP within 2 years from the start of the study who underwent urinalysis and had high-risk glomerulonephritis features (positive test results for hematuria and proteinuria) will participate in the study. Patients will be randomized 1:1 at the patient level, to intervention arm or usual care arm. The Geisinger primary care provider caring for the participating patient will then, based on the patient's randomization status, receive or not receive one-time EHR alert that will fire during patient clinic visit prompting referral to Nephrology. Index date is the date of the PCP visit during which the EHR alert will fire. The analysis plan will evaluate the potential impact of EHR alert. The primary outcome will be proportion of adult patients with high-risk features for glomerulonephritis with a nephrology referral order within 1 month of the PCP visit.

Exploratory outcomes will include:

• Comparing patient demographics (i.e. age, sex, insurance status, race), clinical (i.e. BMI, blood pressure, creatinine, eGFR, protein/urine ratio, comorbidities) and PCP practice characteristics (i.e. rural vs. urban, patient volume, nephrology on site vs off site, PCP years of practice, PCP age group, PCP sex, physician vs. advanced practice practitioner).
• Proportion referred to nephrology within 3 and 6 months of index PCP visit.
• Proportion with completed nephrology appointment within 3 and 6 months of index PCP visit
• Median time to nephrology referral
• Proportion of patients with dipstick hematuria and concomitant proteinuria (2+ or 3+) who underwent quantitative albumin or protein testing (albumin/creatinine ratio or protein/creatinine ratio or 24-hour urine albumin or protein) within 6 months of index PCP visit
• Proportion of patients who had completed a repeat urinalysis and microscopy lab test within 6 months of index PCP visit between the intervention and usual care group
• Median time from referral to nephrology visit
• Assess implementation outcomes of the electronic alert intervention: Acceptability; Adoption; Appropriateness; Implementation Costs; Feasibility; Fidelity; Penetration; Sustainability

Analyses will employ Generalized Linear Mixed Model to evaluate the intervention effect on the primary outcome, after adjusting for patient-level and provider-level characteristics as fixed-effects and allowing for dependencies of patient-level observations within providers by incorporating these as hierarchical random-effects.