Evaluating Follow-Up Metrics in an AI-Assisted Telemedicine Screening Program for Diabetic Retinopathy in Primary Care Clinics After Hiring a Patient Care Navigator

Abstract

Purpose : We aim to evaluate the impact of hiring a patient care navigator on patient follow-up in the artificial intelligence (AI) diabetic retinopathy (DR) screening program via Temple Health System. We compared the rate of and time to follow-up in patients with referable screening diagnosis before and after the navigator.

Methods : A retrospective chart review of patients 18 years or older with diabetes were screened for DR with Eyenuk's (Los Angeles, CA) EyeArt software at 8 primary care clinics between 5/12/21–6/30/22. Results were reported as “referable” for more-than-mild DR (International Clinical Diabetic Retinopathy (ICDR) 2-4 +/- evidence of macular edema) or uninterpretable images, or “not referable” (ICDR 0-1 and no evidence of macular edema). Before 01/2022, patients were responsible for calling and scheduling follow-up appointments. After 01/2022, a half-time study navigator was hired to contact all referable patients and schedule them for a follow-up. Follow-up outcomes were collected before and after the navigator and compared. Chi-square and t-test stastical analyses were done.

Results : 43 patients were screened between 5/12/21-12/31/21. 415 patients were screened between 1/1/22-6/30/22. 23 patients in the pre-navigator phase and 245 patients in the post-navigator phase received referable diagnoses by AI. Of the pre-navigator referable group, 7/23 patients (30%) scheduled a follow-up, and 6/23 (26%) patients completed the follow-up. Of the post-navigator referable group, 139/245 patients (57%) scheduled a follow-up (p=0.02), and 85/245 (35%) of patients completed the follow-up (p=0.7). Prior to the navigator, the average time from screening to scheduling a follow-up was 46 days, and the average time from scheduling to completing the follow-up was 158 days. After the navigator, the average time from screening to scheduling a follow-up was 8 days (p=0.002), and the average time from scheduling to completing the follow-up was 67 days (p<0.001).

Conclusions : Hiring a patient care navigator significantly improved the of rate of patients scheduling a follow-up and times to scheduling and completing the follow-up.

This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.

 

Figure 1. Rate of follow-ups scheduled (A) and completed (B) pre- and post-navigator.

 

Figure 2. Average time from screening to scheduling follow-up and scheduling to completing follow-up pre- and post-navigator.