An overseas study has suggested that autonomous AI-based screening can be more cost-effective and efficient than eyecare provider (ECP)-performed eye exams..
As reported by Ophthalmology Europe, a study by the Section on Biomedical Informatics and Data Science, Johns Hopkins University, Baltimore, concluded that AI brought “greater cost-effectiveness for smaller health systems and cost-saving for larger ones”.
These findings might have some relevance in Australia, especially in the regional, rural and remote regions where access to eye tests and eyecare professionals and facilities is sometimes difficult.
The study, led by first author Ms Mahoor Ahmed, MS, pointed out that autonomous AI had already been shown to be safe and effective for screening patients for paediatric diabetic retinal disease (DRD) and could potentially enhance health equity and clinician productivity.
According to Ophthalmology Europe, the researchers pointed out that cost, among other factors, may deter patients from screening, especially in rural and low-resource settings.
“With DRD screening rates as low as 20% nationwide health systems and medical practices are considering alternative methods to increase screening access and adherence,” the researchers said.
“Therefore, considering the financial expenditures associated with AI screening is an important step in determining if and how to implement these systems.”
In the study, Ahmed and colleagues examined the cost-effectiveness of an autonomous AI strategy versus a traditional ECP strategy during the initial year of implementation from a health system perspective in order to obtain some solid financial data. The incremental cost-effectiveness ratio was the main study outcome measure.
The bottom line, according to Ahmed and colleagues, was that “The base-case analysis showed that the AI strategy resulted in an additional cost of $242 [A$388] per patient screened to a cost saving of $140 [A$224] per patient screened, depending on the health system size and patient volume, compared with the ECP strategy.”
The investigators also reported that the base-case analysis showed that the anticipated AI strategy cost was between $19,368 and $133,900 (equivalent to A$31,077 and A$214,850, respectively), compared to that of the ECP strategy, which is between $8,927 and $357,072 (equivalent to $A14,323 and A$572,943, respectively).
“Thus, excluding insurance reimbursements, the AI strategy results in additional costs of up to $10,441 [A$16,753] and potential cost savings of up to $240,972 [A$386,654] depending on the size of the health system.
“Regarding effectiveness, the AI strategy consistently results in more patients being screened, ranging between 43 and 1,724 additional patients screened annually, depending on patient volume.”
More reading:
Research proves AI accuracy in reading fundus photographs
Artificial intelligence: eyecare’s friend of foe?
Optometrists back Aussie-developed AI system
