Effect of no eyeglasses sales on the quality of eye care: an experimental evidence from China.

Background: Eye examinations and eyeglasses acquisition are typically integrated into a cohesive procedure in China. We conducted a randomized controlled trial using incognito standardized patient (SP) approach to evaluate the impact of separating eyeglasses sales on the accuracy of final prescription.

Methods: 52 SPs were trained to provide standardized responses during eye examinations, and undergoing refraction by a senior ophthalmologist at a national-level clinical center. SPs subsequently received eye examinations at 226 private optical shops and public hospitals in Shaanxi, northwestern China. The visits were randomly assigned to either control group, where SPs would typically purchase eyeglasses after refraction, or treatment group, where SPs made an advance declaration not to purchase eyeglasses prior to refraction. The dioptric difference between the final prescriptions provided by local refractionists and expert in the better-seeing eye was determined using the Vector Diopteric Distance method, and the completeness of exams was assessed against national standards. Multiple regressions were conducted to estimate the impact of no eyeglasses sales on the accuracy of the final prescription of local refractionists, as well as the completeness of examinations.

Results: Among 226 eye exams (73 in public hospitals, 153 in private optical shops), 133 (58.8%) were randomized to control group and 93 (41.2%) to no eyeglasses sales group. The inaccuracy rate of final prescriptions provided by local refractionists (≥ 1.0 D, experts' final prescription as the reference) was 25.6% in control group, while 36.6% in no-sale group (P = 0.077). The likelihood of providing inaccurate final prescriptions was significantly higher in no-sale group compared to control group (OR = 1.607; 95% CI: 1.030 to 2.508; P = 0.037). This was particularly evident in private optical shops (OR = 2.433; 95% CI: 1.386 to 4.309; P = 0.002). In terms of process quality, the no-sale group performed significantly less subjective refraction (OR = 0.488; 95% CI: 0.253 to 0.940; P = 0.032) and less testing SP's own eyeglasses (OR = 0.424; 95% CI: 0.201 to 0.897; P = 0.025). The duration of eye exams was 3.917 min shorter (95% CI: -6.798 to -1.036; P = 0.008) in no-sale group.

Conclusions: Separating eyeglasses sales from optical care could lead to worse quality of eye care. Policy makers should carefully consider the role of economic incentives in healthcare reform.