Skip to content

How can we best implement and scale shared-care models for chronic eye disease?

Belinda Ford on introduced task-sharing models of care to help manage patients with chronic eye diseases.
Published: 10.03.2021
Belinda Ford Postdoctoral Research Fellow
The University of New South Wales

Lessons from existing models in United Kingdom, Finland and Australia

Vision loss and blindness associated with chronic eye diseases, such as glaucoma, can be avoided with timely detection, monitoring, and access to medical/ surgical interventions. But the growing burden of these diseases places high demand on eye care services, causing access blocks for patients.

As a solution, some health systems have introduced task-sharing models of care (known as ‘shared care’ or ‘collaborative care’) to help ophthalmologists in tertiary hospitals diagnose and manage patients with chronic eye diseases. In task-sharing, traditional ophthalmologist-led tasks are conducted by non-medical eye care teams. Multiple studies have showed that task-sharing is not only reliable and safe, but also improves efficiency and capacity for hospital ophthalmology services. Thus, there is rationale to find ways to broaden the reach of these models of care, to further capitalise on these gains.

To explore the key features needed for expansion and adoption of these models of care into new settings, we conducted a qualitative study – using the realist evaluation framework – in health systems with task-sharing in place (Finland, United Kingdom and Australia). Drawing on the experiences of health system stakeholders (clinicians, managers, administrators, policymakers), themes relating to the sustainable implementation and scalability of these models of care were identified.


The impetus to initiate new models of care came from clinicians who were concerned about long delays to access services and the consequences for patient safety and vision loss. However, health system buy-in was gained by redistributing existing health care resources – since this could optimise skill sets with minimal investment. For example, when ophthalmologists provided supervision of low-risk patients through direct or virtual oversight, their time was released to focus on management of more complex or advanced patients. Buy-in was also supported by generating local evidence on service efficiency, safety, productivity and the acceptability of these models of care.

Interdisciplinary trust is essential for successful implementation, since it allows ophthalmologists to shift the clinical responsibility to nurses or optometrists, and in turn, led to better decision making. Trust can be established through formal training and/or daily interactions. Standardised protocols, pathways, and proformas also supported implementation by ensuring that patients were streamlined to appropriate levels of care and the quality of care could be monitored.


Investment from health systems is needed to support sustainability and scalability. Financial incentives may motivate clinicians to participate. These could be justified through cost-savings accrued in task-sharing. However, current financial models are not well-established. Better integrated information technology systems are needed to support information transfer, monitor quality, and to efficiently scale these models. However, specific issues were the use of multiple programmes, emails and/or paper/ scanned files. Often programme governance and processes were adapted to suit the local context. But formal governance processes were necessary when patient volume increased or more partners became involved. These may cover legal, regulatory, accreditation, safety and quality monitoring, and national policy and guidelines. Finally, current evidence assumes that improved capacity and access will lead to better health outcomes for patients. However, there is a need to shift the focus to understand the longer-term health and economic outcomes of these models of care.

Disclaimer: The views, ideas, technologies or policy positions in these blog posts belong to the authors and do not necessarily describe IAPB’s position or views on these matters.