A systematic review of the current availability of mobile applications in eyecare practices

Abstract

Background

There is an increasing shift towards non-communicable eye diseases (NCEDs) because of a demographic transition. Incorporation of telemedicine into everyday clinical practice is becoming increasingly popular. We sought to carry out a systematic review to look at which applications on portable devices are available for use in eyecare practices for patients with NCEDs, specifically refractive error, diabetic retinopathy, age-related macular degeneration (AMD) and glaucoma.

Methods

Pubmed, EMBASE, Medline, PsychInfo databases were systematically searched using keywords and MeSH terms. Eligible articles included peer-reviewed, original full text articles evaluating apps for use on portable devices aimed at patients with NCEDs.

Results

The initial search yielded 100 studies. Nine studies met the inclusion criteria, and an additional eight studies were identified through reference screening. Of the included studies, 29.4%% (n = 5) evaluated applications aimed for use to identify refractive errors, 35.3% (n = 6) aimed at patients with glaucoma, 23.5% (n = 4) for use by patients with AMD, 11.7% (n = 2) for the non-specific monitoring of visual acuity/fields. 76.5% (n = 13) of the studies showed that the application evaluated was an effective and reliable tool compared to clinical standards.

Conclusions

Portable device applications in patients with NCED have been shown to be effective. The use of these apps for patients is limited to either diagnostic or monitoring use. There is scope for apps which encompass other aspects of patient care that have been used in other specialties that may be applied to ophthalmic patients, including those with an educational aim which have a role in increasing compliance.

Introduction

Globally, the leading cause of blindness in those aged 50 years and older in 2020 was cataract followed by glaucoma, under corrected refractive error, age-related macular degeneration (AMD) and diabetic retinopathy (DR).¹ Non-communicable eye diseases (NCEDs) such as DR, glaucoma and AMD represent a growing proportion of the causes of blindness and moderate and severe visual impairment.² The explanation for this lies in the epidemiological transition, that is a change in lifestyle leading to a change in population health, amongst middle- and now low-income countries. These lifestyle changes reflect the increase in cardiovascular diseases and diabetes that we see. The demographic transition occurring, as people are living longer now, has also resulted in an increased number of people with DR, glaucoma and AMD.² Indeed, the aforementioned conditions mainly affect people over the age of 50. Simultaneously shifts in lifestyle trends will mean that myopia and myopia-related complications will also increase as a result of increased near work activity and decreased time spent outdoors.³ These diseases represent a particular challenge to management as they are chronic and require on-going monitoring and input to prevent severe visual impairment that would adversely impact quality of life. Moreover, uncorrected refractive error is a growing burden, with the number of people with myopia expected to rise to 55% by 2050.⁴ The increase in average life expectancy will naturally result in an increase of the burden of presbyopia.

Much like the population, the role of technology in clinical medicine is growing at an exponential rate.⁵ Technology invaluable in the storage and exchange of data. Recent breakthroughs, and the development of artificial intelligence (AI), mean that it has diagnostic and therapeutic potential as well. The number of smartphone users today surpasses 75% of the world's population and is forecast to further grow.⁶ Considering that patients are familiar with mobile solutions in other areas of their lives, we can anticipate the adoption of portable online solutions in the context of healthcare delivery. Telemedicine is a growing field and has proven to be invaluable. For example Abramoff and colleagues found that software labelling fundus photographs was highly sensitive and specific in detecting referable diabetic retinopathy.⁷ Despite this, little is known about the adoption of mobile applications among those affected by eye diseases and their efficacy in achieving their intended outcomes compared to current standards of clinical practice. Mobile applications have a huge potential of helping healthcare professionals meet the demands of the growing patient population which represent new challenges.

This systematic review of the literature aims to evaluate the current availability of applications (apps) on portable devices which are offered for use by patients with non-communicable eye diseases (NCED), specifically, refractive error, DR, AMD, and glaucoma.

Methods

The primary objective of this systematic review was to summarise which applications are available for use by patients with NCED, specifically, refractive error, DR, AMD and glaucoma and their efficacy compared to gold standard monitoring equivalent. An electronic database search was conducted using EMBASE, MedLINE, PubMed,

PsychInfo databases from their inception to March 2021. This systematic review was conducted in line with the Cochrane Handbook for Systematic Reviews and Interventions⁸ and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.⁹

Inclusion criteria for this study were applications for use on either mobile or tablet devices aimed for use by patients with refractive error, DR, AMD or glaucoma. Only full text articles of primary or secondary research were included which were written in English. Publications that did not fulfil our inclusion criteria and duplicates were excluded from the final data extraction.

Search strategies were developed using the following index and free text terms: ((mobile) OR (tablet) OR (portable electronic) OR (cell* phone)) AND ((app) OR (application)) AND ((glaucoma*) OR (retinopathy) OR (macul* degene*) OR (refract* error) OR (short sight*) OR (long sight*) OR (myopi*) OR (hyperopi*) OR (hypermetrop*)).

After pooling and electronic de-duplication, two authors (SF and MS) independently screened all abstracts against the pre-defined inclusion criteria. Disagreements were resolved by review with a third author (DA). Studies which passed the screening criteria were put forward to full-text review for assessment against the pre-defined inclusion criteria to yield the final studies for the data extraction stage. Data from the final studies were extracted independently by two authors (MS, SF) and collated onto a Microsoft Excel (Version 16.52).

Results

We analysed the aggregate data from 17 studies including 1236 participants. A literature search retrieved 100 studies from EMBASE, MedLine and Pubmed databases. After excluding 21 duplicates, 79 records were screened and a total of 17 studies were eligible for data extraction following full-text screening. The study selection process is summarised in the PRISMA flow diagram (Figure 1).

Figure 1.

PRISMA flow diagram outlining the process of study selection.

Of the included studies in this systematic review, 82.3% (n = 14) of the studies were of cross-sectional design with the remainder being of longitudinal design (n = 2)^10,11 and a single study of blinded diagnostic design¹² (Table 1). The majority of the studies were carried out by researchers in the west; five studies were carried out by researchers in Europe,^10,13–16 six were based in Oceania^{11,12,17–20} and two in the United States of America.^21,22 The remaining studies were based in India,^10,23 China,²⁴ South Korea²⁵ and Nepal.²⁶ These studies spanned a nine-year period. The studies included utilised mobile application devices for screening purposes in eye clinics or monitoring purposes in eyecare practices for use by patients.

Table 1.

Summary of studies included in the final text analysis, including demographics, methodology and outcomes in each study.

	Disease target	Study title, year and author	Design, study location	Participants	Control	Intervention	Outcomes Measured	Result, P value	Is App effective?
1	Glaucoma	Glaucoma Home Monitoring Using a Tablet-Based Visual Field Test (Eyecatcher): An Assessment of Accuracy and Adherence Over 6 Months. 2021, Jones et al.13	Prospective longitudinal study. United Kingdom	20 patients (aged 62–78 years) with established diagnosis of glaucoma	Humphrey Visual Field Analyser (HFA) 24-2 SITA-fast	‘Eyecatcher’ tablet perimeter	Concordance between Visual Fields measured at home and in the clinic Adherence to home Visual Field monitoring	The concordance between visual field measured at home and in the clinic was r = 0.94 (P < 0.001) The adherence was 98% Median test time: 4.5 min	✓
2	AMD	Reliability and diagnostic performance of a novel mobile app for hyperacuity self-monitoring in patients with age-related macular degeneration (AMD). 2019, Schmid et al.14	Prospective cross-sectional study. Zurich, Switzerland	89 patients with AMD (63 wet and 26 with dry AMD) (average age 81), 19 age-matched healthy (average age 67) and 34 young healthy individuals (average age 38)	Comprehensive ophthalmic examination in clinic	Alleye mobile application (implements an alignment hyperacuity)	Extent to which Alleye discriminated between healthy eyes and eyes with wet and dry AMD, using the area under the receiver operating characteristic curve (AUC) (95% confidence intervals).	The test was highly accurate to detect wet AMD. The AUC to detect dry AMD was 0.799 (0.675–0.923), and 0.969 (0.940–0.997) to detect wet AMD compared to young healthy subjects,. The discrimination between dry and wet AMD was moderate (AUC 0.660 (0.520–0.799)).	✓
3	Refractive error	Fast measure of visual acuity and contrast sensitivity defocus curves with an iPad application. 2019, Fernández et al.15	Prospective cross-sectional study. Spain	59 patients (average age 59 years) implanted with multifocal intraocular lenses	Early Treatment Diabetic Retinopathy Study (ETDRS) chart	iPad application	Repeatability of the fast measurement of the visual acuity (VADC) and contrast sensitivity (CSDC)	A total of 45.8% of eyes showed no differences between both tests and the difference was less than one line of visual acuity (VA) in 96.6% of the eyes. The intrasubject repeatability was under one line of VA along all the defocus curve except for positive defocus levels. The CSDC was less repeatable than VADC.
4	Glaucoma	Performance of iPad-based threshold perimetry in glaucoma and controls. 2018, Schulz et al.20	Prospective, cross-sectional study. Australia	60 patients with glaucoma (average age 64) 25 healthy controls (average age 53 years)	HFA SITA 24-2	Melbourne Rapid Fields (MRF) iPad perimetry	Linear regression and Bland Altman analyses of global indices sensitivity/specificity using (ROC) curves, intraclass correlations	MRF detected 39/54 abnormal hemifields with a similar threshold-based criteria. Global indices were highly correlated between MRF and HVF: MD r2 = 0.80, PSD r2 = 0.77, VFI r2 = 0.85 (all P < 0.0001). ROC analysis of global indices showed reasonable sensitivity/specificity	✓
5	AMD	A tablet-based retinal function test in neovascular age-related macular degeneration eyes and at-risk fellow eye. 2018, Ho et al.19	Prospective cross-sectional study. Melbourne, Australia	53 subjects (53 nAMD eyes and 21 at risk fellow eyes) in patients: Undergoing anti-VEGF treatment BCVA of 20/80 or better and refractive error of ± 5.00 dioptre or less	Comprehensive ophthalmic examination including spectral-domain optical coherence tomography (SD-OCT)	Examination on the tablet visual field–testing application (created using PsyPad)	Correlation of SD-OCT structure to retinal sensitivity function on the iPad	App able to detect differences in retinal sensitivity related to the underlying pathologic changes associated with AMD, including presence of atrophy, disruption of RPE and absent ellipsoid zone (p < 0.01 for all) like that reported with formal microperimetry	✓
6	Glaucoma, DR	Performance of an iPad Application to Detect Moderate and Advanced Visual Field Loss in Nepal. 2017, Johnson et al.26	Prospective, cross-sectional validation study. Kathmandu, Nepal	206 participants (411 eyes). Of these eyes: 210 were normal, 183 glaucoma, 18 diabetic retinopathy	HFA using 24-2 SITA Standard tests	Visual Fields Easy (VFE) iPad App	Visual fields Time taken to do test using app	The number of missed locations on the VFE correlated with MD (r = 0.79, p < 0.001) and PSD (r = 0.60 iPad suprathreshold perimetry was able to detect most visual field deficits with moderate and advanced loss but had greater difficulty in detecting early loss. Average time to perform the VFE test was 3 min, 18 s	✓
7	AMD	Home monitoring of retinal sensitivity on tablet devices in age-related macular degeneration. 2018, Adams et al.11	Randomised longitudinal prospective study. Melbourne, Australia	38 participants with AMD	Central retinal sensitivity using the macular assessment integrity analyzer (MAIA; CenterVue, Padova, Italy) microperimetry	PsyPad App to measure retinal sensitivity (RS)	RS from PsyPad compared to clinic-based measurements	Mean RS from the tablet device was not significantly different from the clinic-based microperimetry performance (P = 0.58)	✓
8	Refractive Error	Visual Acuity Testing Using a Random Method Visual Acuity Application. 2016, Rhiu et al.25	Prospective cross-sectional study. Cheongju-si, South Korea	43 normally sighted individuals	ETDRS chart	iPad based application	LogMAR Visual acuity (VA)	The logMAR VA showed no significant difference between the ETDRS chart (P = 0.66) and the iPad Snellen chart and iPad Arabic figure chart (P = 0.29). The logMAR VA of the ETDRS chart was significantly better than that of the iPad Tumbling E chart (p < 0.01) or iPad Landolt C chart (p < 0.01)	✓
9	AMD, DR	Handheld shape discrimination hyperacuity test on a mobile device for remote monitoring of visual function in maculopathy. 2013, Wang et al.22	Prospective cross-sectional study. Dallas, Texas, United States	100 patients (27 normal, 37 AMD, 36 diabetic retinopathy)	Desktop computer shape discrimination hyperacuity (dSDH) test	Handheld shape discrimination hyperacuity (hSDH) test iPhone app	Visual acuity User survey to assess the usability	The hSDH test and dSDH test visual acuity measurements were highly correlated (P < 0.0001) 98% of 46 patients (10 with AMD and 36 with DR) who completed the survey reported the hSDH test was easy to use.	✓
10	Refractive Error	The Handy Eye Check: a mobile medical application to test visual acuity in children. 2014, Toner et al.21	Prospective cross-sectional study. Indiana, United States	60 patients (average age 10 years)	Handy Eye Chart	Handy Eye Check iPad application	Visual acuity Reliability and validity of the Handy Eye Check App	There was a strong linear correlation (r = 0.92) between the two tests and the mean difference in acuity was − 0.005 logMAR, or <1letter (95% CI, − 0.03 to 0.02), between the two tests. Test-retest reliability was high, with 81% of retest scores within 0.1 logMAR (5 letters) and 100% within 0.2 logMAR (10 letters).	✓
11	AMD	Measurement of Retinal Sensitivity on Tablet Devices in Age-Related Macular Degeneration. 2015, Wu et al.18	Prospective cross-sectional study. Australia	30 patients with AMD (average age 70 years)	Macular Integrity Assessment microperimeter	PsyPad application for iPad	Comparison of central retinal sensitivity measurement with iPad App vs standard microperimetry	The mean central retinal sensitivity (±SEM) for all eyes was 25.7 ± 0.4 and 26.1 ± 0.4 dB measured using PsyPad and microperimetry, respectively, and was not significantly different using these two methods (P = 0.094).	✓
12	Glaucoma	Comparison of Perimetric Outcomes from Melbourne Rapid Fields Tablet Perimeter Software and Humphrey Field Analyzer in Glaucoma Patients. 2020, Kumar and Thulasidas23	Cross-sectional observational study. India	28 patients with glaucoma (average age 52)	Humphrey Field Analyzer	Melbourne Rapid Fields (MRF) iPad based perimeter	Comparison of visual field outcomes in the following parameters: - Mean deviation (MD) - PSD - VFI/VC - Foveal threshold - Number of points depressed at on the PSD probability plot - Glaucoma hemifield test/color coded indicator	Perimetric outcomes that were significantly different between the two methods used included mean MD, mean PSD, foveal threshold and the number of points depressed at on the PSD probability plot (all p < 0.01). Bland–Altman plots showed that considerable variability existed between the programs.
13	Glaucoma	Six-month Longitudinal Comparison of a Portable Tablet Perimeter With the Humphrey Field Analyzer. 2018, Prea et al.10	Multicenter longitudinal observational clinical study. New Delhi, India (n = 21) Cambridge, United Kingdom (n = 39)	60 patients with stable glaucoma/ocular hypertension/glaucoma suspects	Humphrey Field Analyzer (HFA) 24-2 SITA standard and fast programs	Melbourne Rapid Fields (MRF) iPad based perimeter	Medium-term repeatability of the MRF compared to the HFA	MRF correlated strongly with HFA across 4 visits over a 6-month period and has good test-retest reliability.	✓
14	Glaucoma	A Comparison of Perimetric Results from a Tablet Perimeter and Humphrey Field Analyzer in Glaucoma Patients. 2016, Kong et al.17	Prospective cross-sectional study. Melbourne, Australia	90 patients (12 had normal optic nerves and 78 had glaucoma with various degrees of visual field loss (41 mild and 37 moderate-severe)	HFA 24-2 SITA standard	Melbourne Rapid Fields (MRF) iPad based perimeter	correlation between the perimetric outcomes from perimetry using MRF versus HFA	The test durations were shorter on MRF than HFA (P < 0.001). MRF showed a high level of concordance in its outcomes with HFA MRF also showed levels of test–retest reliability comparable to HFA	✓
15	Glaucoma	Portable Perimetry Using Eye-Tracking on a Tablet Computer – A Feasibility Assessment. 2019, Jones et al.16	Prospective cross-sectional study. London, United Kingdom	12 glaucoma patients and 6 age-similar controls	HFA SITA standard 24-2	Eyecatcher App	Paracentral visual field assessment	Eyecatcher scores were strongly correlated with MD scores (r2 = 0.64, P < 0.001), and there was good concordance between corresponding VF locations (∼84%). Participants reported that Eyecatcher was more enjoyable, easier to perform, and less tiring than SAP (all P < 0.001).	✓
16	Refractive Error	An assessment of the iPad as a testing platform for distance visual acuity in adults. 2013, Black et al.12	Single centre, blinded, diagnostic test study. Auckland, New Zealand	85 university and staff students with VA better than 6/60	ETDRS chart and the Medmont system	Visual Axuity XL iPad application	Visual Acuity	Acuity measures with iPad were significantly poorer (approximately 2 LogMAR lines) than those made using an ETDRS chart and Medmont system Antiglare screen and veiling the iPad resulted in acuity measurements that were equivalent those made using gold standard charts (n = 29).
17	Refractive Error	A pilot trial of the iPad tablet computer as a portable device for visual acuity testing. 2013, Zhang et al.24	Single centre cross-sectional study. Guangzhou, China.	120 patients (mean age 47 years) with a mix of ophthalmic diagnosis (10 normal eyes)	Light-box chart	Eye Chart Pro app on iPad	Visual Acuity	In patients with Snellen VA >20/200 there was no significant difference between the two methods. In patients with Snellen VA ≤ 20/200, the iPad results were significantly lower (p < 0.001) than the light-box chart

nAMD, neovascular age-related macular degeneration; BCVA, best corrected visual acuity; VEGF, vascular endothelial growth factor; RPE, retinal pigment epithelium; MD, mean deviation; PSD, pattern standard deviation; VFI, visual field index; VC, visual capacity; SITA, Swedish interactive threshold algorithm; RS, retinal sensitivity; DR, Diabetic Retinopathy.

Shaded rows indicate studies evaluating applications aimed at disease monitoring, unshaded indicates studies evaluating applications aimed at disease diagnosis.

Most of the studies included evaluated mobile applications aimed for use by patients with glaucoma (35.3% of included studies, n = 6)^{10,13,16,17,20,23} and at patients with refractive errors (29.4% of included studies, n = 5).^{12,15,21,24,25} The remaining applications were targeted at patients with AMD (23.5% of included studies, n = 4)^11,14,18,19 and finally, two studies (11.7%) evaluated applications assessing visual acuity/sensitivity and visual fields assessment for use in patients with relevant ophthalmic diagnosis where this may be a suitable form of disease monitoring.^22,26 This is summarised in Figure 2.

Figure 2.

Pie chart summarising the focus of the mobile applications included in the final study selection (n = 17).

We categorised the applications evaluated in the studies included into which aspect of the patient's healthcare journey was aimed. We identified two categories; 12 of the included studies evaluated applications targeted at disease monitoring^{10,11,13,14,16–20,22,23,26} and five evaluated applications targeted at disease diagnosis^{12,15,21,24,25} (namely diagnosis of refractive error) (Table 1). 76.5% (n = 13) of the included studies yielded significant results which support the efficacy of the application evaluated versus the gold standard control used.^{10,11,13,14,16–22,25,26} Further comprehensive details on each study with their findings are outlined in Table 1.

Discussion

In parallel with technological advancements, tablet and mobile based applications have great potential in aiding healthcare professionals in the care of the increasing ophthalmic population. We analysed 17 studies in this systematic review which evaluated the use of applications on portable electronic devices for patients with NCEDs, specifically, DR, glaucoma, AMD and refractive errors.

Six studies looked at applications aimed for use by glaucoma patients. Four of them evaluated the Melbourne Rapid Fields (MRF) application for use in disease monitoring for glaucoma patients compared to the current gold standard, Humphrey Field Analyzer (HFA).^10,17,20,23 Prea et al. concluded that MRF correlated strongly with HFA across 4 visits over a 6-month period and had good test-retest reliability.¹⁰ Schulz and colleagues concluded that MRF perimetry, despite using a completely different test paradigm, shows good performance characteristics compared to HFA for detection of defects.²⁰ Similarly, Kong et al. found that the perimetry results from the MRF have a strong correlation to the HFA outcomes and had test–retest reliability comparable to HFA.¹⁷ Kumar and Thulasidas found that perimetric outcomes were significantly different between the two methods used, including mean deviation (MD), mean pattern standard deviation (PSD), foveal threshold and the number of points depressed at on the PSD probability plot (all p < 0.01).²³ Therefore, although a good cost-effective and time-saving tool for monitoring visual fields there are limitations and scope for improvement in the widespread adoptability of MRF for glaucoma monitoring. The main limitations include the apps inability to detect early cases²³ and heterogeneity reproducibility for individual points.²⁰ Therefore, it is more suitable for use in settings where conventional perimetry is not readily accessible or suitable.¹⁰ The remaining two studies evaluated the Eyecatcher application, an alternative visual field measuring application, compared to the HFA.^13,16 Both studies were carried about by Jones and colleagues and both showed significant concordance between visual field measures using the application and in the clinic. The application was favoured by participants and had high adherence rates.^13,16

Four of the studies evaluated applications for use in patients with AMD.^11,14,18,19 Schmid et al. tested the Alleye mobile application, which applies a similar principle of hyperacuity testing as the Amsler grid test, in 89 patients with AMD (63 with wet AMD and 26 with dry AMD) versus 19 age-matched controls and 34 young healthy individuals.¹⁴ The test was highly accurate to detect wet AMD and reasonably accurate to classify dry vs. wet AMD. Similar to the Amsler grid test, this application was intuitive and needed very little explanation for first time users. The remaining three studies evaluated the PsyPad application which measures retinal sensitivity within the central 5° of vision. Ho et al. compared the retinal sensitivity detected by the application with pathological changes in patients with neovascular AMD seen using spectral-domain optical coherence tomography (SD-OCT).¹⁹ PsypAD was able to detect differences in retinal sensitivity that are related to the underlying pathologic changes associated with AMD, including presence of atrophy, disruption of RPE and absent ellipsoid zone (p < 0.01 for all), similar to that reported with formal microperimetry.¹⁹ Importantly, they noted that this app had a short test time and was able to be used by their cohort. Adams et al. tested the same app in 38 patients with intermediate AMD in detecting retinal sensitivity versus clinic based microperimetry using the macular assessment integrity analyzer and found no significant difference in the performance of the two.¹¹ Wu et al. also evaluated the PsyPad application in 30 patients with AMD with no statistical difference compared to Macular Integrity Assessment microperimeter.¹⁸

Five studies evaluated the use of mobile applications in the detection of refractive error.^{12,15,21,24,25} Of these, three studies that the visual acuity applications yielded largely different results to current gold standards.^12,15,24 Rhiu et al. found that there was no significant difference in visual acuity measurements between the ETDRS chart (p = 0.66) and the iPad Snellen chart and iPad Arabic figure chart (P = 0.29).²⁵ However, there were differences when comparing to the iPad Tumbling E and Landolt C chart (p < 0.01).²⁵ Toner and colleagues found a strong linear correlation between visual acuity attained through the Handy Eye check iPad application and the Handy Eye chart.²¹ This heterogeneity may, in part, be explained by the different interfaces used with some studies using a mobile device to deliver the application and others using tablet device. Indeed, Black et al. demonstrated that measures to reduce glare resulted in improved accuracy gained from visual acuity testing using the iPad application.¹²

Finally, two studies evaluated applications for use by non-specific pathologies.^22,26 Wang et al. found that a handheld shape discrimination hyperacuity (hSDH) iPhone application test designed for visual function self-monitoring in patients with AMD and DR was highly correlated (p < 0.0001) to the desktop test (dSDH).²² Moreover, 98% of the patients who completed the usability survey found that it was easy to use.²² Johnson et al. found that the number of missed locations on the visual fields easy iPad application was significantly correlated with the MD from the HFA in 206 patients (210 normal, 183 with glaucoma and 18 with DR).²⁶

To our knowledge, this is the first systematic review evaluating the use of mobile applications for patients with NCEDs. Skrzypecki and colleagues performed a search of the available patient-oriented mobile applications in ophthalmology on the App store and Google Play mobile stores (yielding results in English language only).²⁷ They found 56 apps in total, with the large majority focusing on dry eye diseases (13 apps), AMD (14 apps) and strabismus (14 apps). Most apps came into the category of educational applications (23 apps), self-diagnosis (20 apps) and those focussing on compliance (11 apps). Interestingly, they found no correlation between prevalence of a particular disease and the number of available applications, or their downloads. They hypothesise that the demand for applications increases ‘when the onset of the disease matches the age of the average mobile application user’.²⁷ Bearing in mind that the prevalence of NCEDs increases with age, the studies included in our final analysis have reported, where collected, positive feedback from patients regarding usability. Whereas Skrzypecki et al. highlight the abundance of these ophthalmic applications to patients, we highlight the paucity of literature evaluating and validating these applications for use by patients.

Limitations of the studies

This study has limitations. First, there was only one multi-centre trial included,¹⁰ the rest were small single centre trials. Geographical bias also exists, as 76.5% were carried out in Western populations.^{10,11,12–19,20–22} For a more representative and externally valid data set, future work will need to focus on expanding to larger multi-centre randomised control trials with a wide demographic base. This will also help to identify specific patient groups which can be targeted for more specialised educational intervention on using the apps correctly or addressing specific concerns.

Secondly, there was heterogeneity of study design (for example participant recruitment, device choice, eye examined and explanation of how to use the application) and measures of reported outcomes of the studies included. Therefore, it is challenging to reach any firm conclusions on the validity of the use of mobile applications in patients with NCEDs. Furthermore, device settings such as illumination and the patient level of proficiency with technology were not equal across all studies and difficult to control for.

Additionally, many of the studies tested the mobile application in clinics under a supervised setting. The real potential advantage of these mobile application testing in the home or field is not well addressed and will require carefully designed studies to evaluate the potential for increased variability in unsupervised conditions. Moreover, a requirement for these applications is patient access to internet. Currently, there is a paucity of literature available in the sources we searched on the reliability and efficacy of these applications in patients in rural populations, when in fact these patients could potentially benefit the most from such applications.

Mobile applications are becoming increasingly popular in clinical medicine and are versatile in what they can offer patients, including screening diseases, improving patient understanding and guiding management. Interestingly, we found that most of the literature on ophthalmic applications for patients with NCEDs are either aimed at diagnosing pathology or aiding disease monitoring. In other areas of medicine, there are applications which educate patients on possible lifestyle modifications which have proved popular amongst patients.^28–30 Likewise, applications targeting other aspects of ophthalmic patient care would be worthwhile.

Conclusion

This systematic review has found promising evidence on the use of mobile applications for ophthalmic patients. Applications offer an exciting and innovative way in which patients can feel more empowered and involved in their care in the digital age. Randomised clinical trials with larger sample sizes should be performed to evaluate the efficacy of these applications. It needs to be established whether home monitoring and screening will be sustainable over longer periods of time and whether the patients will be compliant. For screening devices, the ability to adequately spot a rapid progression will be essential. Cost-effectiveness of use of such applications in clinical practice also needs evaluation. Ultimately, the ability to develop and perform sensitive measures of visual function on a portable and low-cost tablet device has enormous potential for managing ophthalmic diseases.