Comparing Interrater reliability between eye examination and eye self-examination 1

Resume Objective: to compare Interrater reliability concerning two eye assessment methods. Method: quasi-experimental study conducted with 324 college students including eye self-examination and eye assessment performed by the researchers in a public university. Kappa coefficient was used to verify agreement. Results: reliability coefficients between Interraters ranged from 0.85 to 0.95, with statistical significance at 0.05. The exams to check for near acuity and peripheral vision presented a reasonable kappa >0.2. The remaining coefficients were higher, ranging from very to totally reliable. Conclusion: comparatively, the results of both methods were similar. The virtual manual on eye self-examination can be used to screen for eye conditions.


Introduction
Eye disorders are usually related to chronic health conditions and present multi-varied etiologies. The onset of eye conditions is insidious so that people often do not notice symptoms and, for this reason, their search for specialized care services is delayed, resulting in complications or irreversible visual loss (1) . The incidence of blindness may affect 2 million people every year. Estimates show that if sufficient resources are not invested in prevention, the number of cases may double in the next 10 years (2) .
The share of the population composed of individuals with moderate or subclinical eye disorders is unknown.
Figuratively speaking, it is assumed that these individuals represent only the tip of the iceberg, that is, the population with eye disorders is much larger, so we cannot accurately describe the real magnitude of eye disorders.
This alarming situation is aggravated by increased life expectancy and increased population, a scarcity of specialized services, difficult access to eye care for the population, economic problems and/or a lack of educational programs promoting the adoption of preventive measures (3) .
Given the evidence that people respond better and are more prone to make decisions when educational strategies using diversified material that favor inclusion, interactivity and accessibility are used (4)(5) , we sought to adapt eye self-exams to virtual media and validate them with experts and test them with nursing students.
The idea is to enable people to identify eye problems in order to seek specialized care at an earlier point in time.
For this strategy to be successful, however, assessment of Interrater reliability is necessary to determine whether the results of self-exams are consistent with examinations performed by trained staff.
Eye self-examination enables people to understand their eyes and identify changes that may occur in visual acuity, external eye structures, visual field and eye movement. Hence, health workers should encourage the practice. The relevance of this study is that it makes technology providing eye assessment available and encourages the search for specialized care when warning signs are detected. The objective was to compare Interrater reliability of two eye assessment methods. The hypothesis of interest is that eye examinations performed by specialized staff, considered to be the gold standard, obtain results that are compatible with self-exams when using a virtual manual.

Method
This quasi-experimental study focuses on the assessment of Interrater reliability concerning an eye examination technique performed by a trained researcher. The participants performed an eye self-examination using a virtual manual on eye self-examination.
Reliability is the extent to which repeated measurements, of a stable phenomenon, performed by different people and/or instruments at different times and places, present similar results. Interrater reliability assesses level of agreement among raters or the consistency of the performance of two or more raters in recording the same answers at the same time (6) . This property was verified in this study using the results of eye assessments performed by the researcher and team.
College students using a virtual manual addressing eye self-examination performed the eye self-examination.
The study was conducted from January to May 2014 in a public university with a population of 2,060 students. This institution was chosen because of the availability of computers, Internet access at the institution and the ease and access the primary author had to conduct data collection.
A convenience sample was used and inclusion criteria were: being in a physical condition to perform the eye self-examination and mastering basic informatics.
Those with diagnosed eye conditions or enrolled in programs from the health field were excluded.
Sampling calculation was carried out considering 95% confidence level, estimated proportion of success of 50%, and 5% level of precision, resulting in a sample size of 324 observations, according to the formula for finite populations:

Considering the risk of communication among
participants regarding the exam to be performed, and aiming to avoid data collection bias, students from different classes, programs and shifts were invited in alternated weeks. Data were collected from seven programs, which were stratified according to the number of students in order to ensure representativeness.

Discussion
Self-examination is a self-care strategy that has proven benefits for the survival of many individuals A presymptomatic eye assessment, which identifies age-related cataracts, for example, can promote preventive measures, as the eye is easily accessible for the topical application of medication. Its potential is translated as pharmacologically practical prevention or even the treatment of cataracts (12) .
Hence, easy and reliable ophthalmic screening conditions may impact quality of life and result in costs in the health field (13) .
The virtual manual presented positive effects regarding the correct way to perform a self-exam and can be considered to have the potential to expand clinical results, as it is in agreement with new conceptions of health care management (14) .

Conclusion
In this study, eye assessment performed by trained people was compared to college students' selfexaminations and Interrater reliability was attested as Kappa index concerning level of agreement ranged from regular to total reliable.
There are some limitations, however, such as the This study shows the use of the Internet and educational media is a reality worldwide and is booming in the fields of health, but investment in the nursing field is needed.