UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Accuracy and Utility of Self-report of Refractive Error

Cumberland, PM; Rahi, JS; Chianca, A; (2016) Accuracy and Utility of Self-report of Refractive Error. JAMA Ophthalmology , 134 (7) pp. 794-801. 10.1001/jamaophthalmol.2016.1275. Green open access

[thumbnail of Article]
Text (Article)

Download (480kB) | Preview
[thumbnail of Key points]
Text (Key points)
Cumberland_OPH15-2259_Key points.pdf

Download (174kB) | Preview
[thumbnail of Figure 1]
Image (Figure 1)

Download (876kB) | Preview
[thumbnail of e-Figures]
Text (e-Figures)

Download (582kB) | Preview


Importance Large-scale generic epidemiological studies offer detailed information on potential risk factors for refractive error across the life course, often lacking in ophthalmology-specific studies. However, ophthalmic examination to determine refractive error phenotype is challenging and costly thus, in that context, refractive status is commonly assigned using questionnaires. In a population survey there is often only scope to include a few condition-specific self-reported questions so it is critical that the questions used are effective in both ‘ruling in’ those who have the trait of interest and ‘ruling out’ those without it. Objective We determined the accuracy of identification of refractive status using self-reported age and/or reason for first wearing optical correction. Design UK Biobank study: cross-sectional epidemiological study. Setting Six regional centres in England and Wales. Participants 117,278 participants, aged 40–69 years in 2009/10. Main outcome and Measures Subjects had autorefraction measurement of refractive status. Spherical equivalent (SphEqu) on the more ‘extreme’ eye was used to categorise myopia (SphEqu ≤-1diopter) and hypermetropia (SphEqu ≥+1diopter). Sensitivity and specificity of reason for optical correction were assessed, using autorefraction as the gold-standard. ROC curves assessed the accuracy of self-reported age of first wearing optical correction and incremental improvement with additional information on the reason. Results Of those reporting using glasses/contact lenses, 92,121/95,240 (97%) gave age at first use and 93,156 (98%) the reason. For myopia, sensitivity of reason for optical correction was 89.1% [88.7, 89.4], specificity 83.7% [83.4, 84.0] and positive and negative predictive values were 72.7% [72.2, 73.1] and 94% [93.8, 94.2] respectively. The area under the curve (AUC) was 0.829 [0.826, 0.831], improving to 0.928 [0.926, 0.930] with combined information. By contrast self-report of reason for optical correction for hypermetropia had low sensitivity (38.1% [37.6, 38.6]) and the AUC with combined information was 0.71 [0.709, 0.716]. Conclusions and Relevance In combination, self-report of reason for and age at first use of optical correction are accurate in identifying myopia. These findings indicate an agreed set of questions could be implemented effectively in large-scale generic population-based studies, to increase opportunities for integrated research on refractive error to develop novel prevention or treatment strategies.

Type: Article
Title: Accuracy and Utility of Self-report of Refractive Error
Open access status: An open access version is available from UCL Discovery
DOI: 10.1001/jamaophthalmol.2016.1275
Publisher version: http://dx.doi.org/10.1001/jamaophthalmol.2016.1275
Language: English
Additional information: Copyright © 2016, American Medical Association.
Keywords: Refractive error, UK Biobank, Self-report, Accuracy
UCL classification: UCL
UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences
UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Population Health Sciences > UCL GOS Institute of Child Health
UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Population Health Sciences > UCL GOS Institute of Child Health > Population, Policy and Practice Dept
URI: https://discovery.ucl.ac.uk/id/eprint/1477606
Downloads since deposit
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item