Abstract: Objective: To explore the value of handheld non-mydriatic fundus photography in fundus examination of patients with
gestational hypertension. Methods: A total of 111 patients with gestational hypertension were enrolled in the study. They underwent
desktop non-mydriatic fundus photography, handheld non-mydriatic fundus photography, and direct ophthalmoscope fundus
examination. The sensitivity, specificity and accuracy of the other two fundus examination methods in identifying retinal lesions in
patients with gestational hypertension were compared with the results of desktop non-mydriatic fundus examination as the gold
standard, and Kappa tests were performed. Results: The detection rates of retinal lesions by handheld non-mydriatic fundus
photography and direct ophthalmoscope fundus examination were 80.18% and 76.58%, respectively. There was no statistically
significant difference compared with the detection rate of desktop non-mydriatic fundus photography （82.88%）. The sensitivity,
specificity, and accuracy of handheld non-mydriatic fundus photography were 94.57%, 89.47%, and 93.69%, respectively, with a
Kappa value of 0.79. The sensitivity, specificity, and accuracy of direct ophthalmoscope fundus examination were 88.04%, 78.95%, and
86.48%, respectively, with a Kappa value of 0.58. Conclusion: Handheld non-mydriatic fundus photography can objectively and
truthfully reflect retinal lesions and can be used as a fundus examination method for patients with gestational hypertension, especially
in bedside fundus examinations that require continuous monitoring.

Key words: desktop non-mydriatic fundus photography, handheld non-mydriatic fundus photography, direct ophthalmoscope fundus
examination, gestational hypertension, retinal lesions