Diabetes is a common cause of blindness in the US and much of this blindness is preventable by early detection and treatment. Although VA quality criteria allow some individuals with diabetes to have biennial examinations, current practice continues to emphasize routine, annual dilated eye exams for most patients. Studies suggest that the current "one-size fits all" method wastes resources while also ignoring an opportunity to improve out-comes for high-risk patients since patients referred for diabetic retinopathy screening exams require different follow-up then those referred for surveillance exams. Therefore, we are conducting this translational research project to examine the impact of a coordinated and targeted system-level intervention, the Progressive Reminder and Scheduling System (PRSS).
The study's primary objective was to determine whether the PRSS improves the optimal timing of photocoagu-lation in diabetic patients in VA. Secondary objectives included assessing if the program: (1) leads to improved compliance with retinopathy screening and surveillance visits; (2) improves patient and provider satisfaction with VA diabetic eye care; (3) reduces eye care visit rates among diabetics receiving eye care at VA; (4) decreases health care resource utilization; and (5) improves the cost-effectiveness of eye care for patients with diabetes.
This quasi-experimental study was designed to examine the effect of the PRSS on eye care for diabetic patients at three VAMCs (Ann Arbor, Cleveland and West Los Angeles) compared with usual management at three VAMCs (Denver, Hines and Houston) which together provide care for 25,000 diabetics. The intervention was be evaluated using historical controls (pre-post analyses) and by comparison to the control sites. This was to be a pragmatic study design (i.e., the study is designed in the current clinical setting but the intervention will be evaluated and refined during the study). Data has been collected from 3 sources: 1) VISTA is being be used to assess resource use, patient demographics, co-morbidities and medications; 2) chart review of patients referred for eye care is being used to assess if the patient underwent a timely examination or it was sub-optimally timed (e.g., the patient already had a major retinal bleed at the time of the procedure); and 3) surveys of patients regarding their attitudes and satisfaction toward VA eye care and surveys of VA health care providers regarding diabetic eye care services are being used to evaluate the program.
The study received IRB approval at all six VAMCs and findings will be divided by specific areas of interest.
Patient Survey: The baseline "patient satisfaction and identification of barriers to eye care in the VA system" survey has been completed at the six sites. We mailed 600 surveys, 100 per study site. 152 veterans from the original sample were determined to be ineligible (due to one of the following: deceased, too ill, not living in the community (in a nursing home, in patient hospital), did not use the VA for the primary source of care/eye care, only receive their medications through the VA). Of the remaining 448 participants where eligibility may have been unknown, there were 265 completed and eligible surveys returned (a response rate of 60%).
Chart Abstraction: We developed an electronic health summary form to produce information for the chart review (e.g., drugs, hemoglobin A1c level, appointments; and Primary Care, Endocrinology and Ophthalmology notes). The form was successfully used at all six sites. We reviewed approximately 830 charts from the six sites and that data is currently being key-punched in preparation for analysis.
Organizational Assessment: A series of concepts was identified to be studied during the organizational assessment. Site visits were conducted to all six sites and approximately 122 interviews were conducted with personnel related to diabetic eye care. These interviews have been transcribed and are currently being analyzed. In addition, a formal written survey was also conducted at each site to supplement the qualitative interviews. Marked organizational issues (e.g., management by optometry or ophthalmology) and technological challenges (e.g, failure of computer systems to talk to each other; inability to upload data into CPRS for groups of patients) have been noted at each of the three sites visited.
PRSS intervention: We have developed a stand-along Access data base for managing the eye care of patients with diabetes. Unfortunately, due to VA computer systems the PRSS cannot be currently deployed at any VA due to the inability of the VA's scheduling system and CPRS clinical system to communicate with each other as they were developed separately. We have pilot tested the feasibility of using this system for two months in the Ann Arbor VA health system. Analysis of that data is currently underway. In addition, to test the ability of the VA system to cope with the unmet need of Veterans who have not been receiving eye care, we sent a mailing to all Veterans at the Ann Arbor VA healthcare system for whom we had no record of their receiving eye care and recommended they come in for an evaluation or notify us if they had an eye examination outside the VA healthcare system of which we were unaware. Approximately 2000 letters were mailed and patients have responded by contacting the health system to arrange appointments. The overall impact of this pilot study is currently being analyzed.
Our project represents an ideal opportunity to translate recent research findings into practice on a broad scale in a timely fashion, without compromising rigorous evaluation methodologies. Findings from the project can impact recommended eye care for veterans with diabetes by examining the feasibility of increasing surveillance eye examinations, decreasing unnecessary screening examinations in low risk patients, optimizing timing of photocoagulation, and improving communication among providers.
- Krein SL, Bernstein SJ, Fletcher CE, Makki F, Goldzweig CL, Watts B, Vijan S, Hayward RA. Improving eye care for veterans with diabetes: an example of using the QUERI steps to move from evidence to implementation: QUERI Series. Implementation science : IS. 2008 Mar 19; 3:18.
- Krein SL. Step-by-Step through the QUERI Process: Improving Diabetes Eye Care. Paper presented at: AcademyHealth Annual Research Meeting; 2005 Jun 1; Boston, MA.