The Effect of Using a Shared Electronic Health Record on Quality of Care in People With Type 2 Diabetes

Systems of managing people with type 2 diabetes through an early goal-orientated strategy with intention to treat to target through frequent visits with a community-based health care provider with specialist training are known to lead to reductions in HbA1c levels. ¹ Community-based diabetes care has been delivered through innovative models of care developed and delivered locally. ²

We have previously intervened successfully in bridging gaps between in patient care and primary care through electronic discharge communications. ³ We sought to overcome communication gaps in outpatient care may by introducing prompts in an electronic medical record as clinical decision support ⁴ and shared electronic health records in community-based care.

We report a 12-month prospective observational study reviewing the service against HBA1c as an outcome and National Institute of Clinical Effectiveness (NICE) 2011 quality standards ⁵ as process and a patient experience questionnaire. We studied 100 patients with type 2 diabetes (mean age 60.4, SD 13.8 years, 51 males) with type 2 diabetes all having HbA1C > 7.5 referred to the service from general practitioners. Glycemic control was prospectively monitored, every 3 months, using HbA1c. HBA1c targets were assessed at baseline, 6 months, and 12 months.

Mean HBA1c fell from 9.5 (SD 1.4%) at baseline to 8.9 (SD 1.5%) at 6 months (P = .004, 95% CI 0.18-0.94) and 8.6 (SD 1.4%) at 12 months (P = .001, 95% CI 0.5-1.2). Quality targets of < 7.5 HbA1C were achieved at 6 months in 16 (16%) and 22 (22%) in 1 year. Patients with HbA1C < 7.5 (8%) where discharged back to primary care within the year.

NICE quality standards were achieved in 100% for structured education annual review and access to ongoing education, receiving personalized nutrition and physical exercise education within an education program, documented agreement on personalized HBA1c target, advice on hypoglycemia, for annual assessment of risk an screening for diabetes complications, women of childbearing age receiving preconceptual counseling but 0% for patients to be assessed for psychological problems and managed appropriately.

We have demonstrated that the service delivered a high quality of care in terms of process and outcome.

The influence of shared records cannot be presumed to be the sole reason for improved control. No intervention occurs in isolation in a real-world setting. Hence unlike in a well-designed and executed randomized clinical trial, it cannot be assumed that all causes of bias have been eliminated. However using a clinical microsystems approach ⁶ to improving quality through service redesign and reevaluation, we have attempted to “successfully juggle multiple improvements while taking excellent care of patients and continue to develop an enhanced sense of ourselves as a system.” ⁶

We conclude that, subject to allowing for bias from confounding variables as outlined above, based on evidence from this real-world nonrandomized, noncontrolled observational study, sharing electronic medical records contributed to the success of integrated intermediate care delivery for people with diabetes and should be considered an essential component of service improvement in integrated care systems.