Three-year Outcomes of an Expanded Asynchronous Virtual Glaucoma Clinic in Singapore

Abstract

Purpose: 

Virtual glaucoma clinics can help increase health care capacity, easing the disease burden of glaucoma. This study assesses the safety, rate of glaucoma progression, time efficiency, and cost-savings of our expanded Glaucoma Observation Clinic (GLOC) at the Singapore National Eye Centre over 3 years.

Methods: 

All patients seen at GLOC between July 2018 and June 2021 were included. Visual acuity, intraocular pressure, and visual fields or optic nerve head imaging were recorded, followed by a virtual review of the data by an ophthalmologist. An objective review of the management of 100 patients was conducted by 2 senior consultants independently as a safety audit. Patient outcomes including the rate of instability (due to worsening of clinical parameters necessitating a conventional clinic visit), glaucoma progression, the consultation review time efficiency, and cost-savings of GLOC were measured.

Results: 

Of 3458 patients, 16% had glaucoma, and the others had risk factors for glaucoma. The safety audit demonstrated a 95% interobserver agreement. The rate of instability was 14.6%, of which true progression was observed in 3.12%. The time taken for a glaucoma specialist to review a GLOC patient was 5.75±0.75 minutes compared with 13.7±2.3 minutes in a conventional clinic. The per capita manpower cost per GLOC patient per visit was $36.77 compared with $65.62 in the conventional clinic. This translates to a cost-saving of $280.65 over the lifetime of a hypothetical patient.

Conclusions: 

Our expanded virtual glaucoma clinic is a safe, time-efficient, and cost-effective model with low rates of glaucoma progression, which could allow for significant health care capacity expansion.

INTRODUCTION

The number of glaucoma cases worldwide is expected to increase to 111.8 million in 2040, particularly affecting patients who reside in Asia and Africa.¹ The disease burden of glaucoma in Singapore is rising, consistent with our aging population.² As glaucoma tends to affect older persons, its magnitude in Singapore is estimated to increase from 59,030 in 2010 to 150,965 in 2040.³ With the increasing disease burden of glaucoma, there will be a growing need for the better provision of services to screen, diagnose, and manage patients.

The diagnosis and management of glaucoma require the expertise of an ophthalmologist and often lifelong follow-up. Conventional models of care require patients to perform glaucoma-related investigations and attend doctor’s consultations, which are labor-intensive and resource-intensive. One approach to reduce manpower cost is through the adoption of telehealth services to screen for glaucoma, evaluate the severity of glaucoma, determine a management plan, and monitor glaucoma progression.⁴

The Royal College of Ophthalmologists in the United Kingdom has issued guidelines to conduct virtual glaucoma clinics (VGCs) while increasing capacity.⁵ The recent coronavirus disease 2019 (COVID-19) pandemic has been a powerful catalyst for the implementation and scaling up of telemedicine initiatives.⁶ At the Singapore National Eye Centre (SNEC), the Glaucoma Observation Clinic (GLOC), an asynchronous virtual clinic, was set up in 2018 as a follow-up scheme for stable glaucoma patients and to monitor those at increased risk of developing glaucoma. In a pilot study involving 377 patients from July 2018 to June 2019, we showed that GLOC reduced waiting and consultation times and manpower costs and was a sustainable model of care for managing patients with low risk of development or progression of glaucoma.⁷ There was also a high level of patient acceptance and satisfaction.⁷ After this successful pilot, we decided to increase the decanting of patients into the GLOC.

Our initial GLOC recruitment focused on patients with a low risk of developing significant visual loss, such as those with ocular hypertension, disc suspects, pseudophakic primary angle-closure, normal-tension glaucoma, or patients being monitored due to risk factors for glaucoma.⁷ A study conducted in Moorfields Eye Hospital revealed the capability of delivering high-quality glaucoma care through a VGC while expanding patient eligibility.⁸ This expansion allowed the inclusion of more varied glaucoma phenotypes and disease stages while delivering safe and effective care and maintaining high levels of patient satisfaction.⁸ Singapore has adopted a health care system, which is different from the National Health Service in the United Kingdom. We also have a different spectrum of diseases and ethnic mix. Thus, it was important to evaluate whether a similar expansion in the patient eligibility for our VGC would yield outcomes comparable to the European population. From 2019, we increased the patient eligibility by including patients: (1) on glaucoma medications, (2) diagnosed with mild-to-moderate primary open-angle glaucoma or primary angle-closure glaucoma, (3) diagnosed with mild secondary glaucoma, and (4) selected only-eyed patients with mild glaucoma. Patients who needed concurrent retinal screening for diabetes or age-related macular degeneration could also be reviewed in the VGC, which runs in conjunction with a virtual or physical retina clinic.

In this study, we report the 3-year outcomes of the expanded GLOC. We aim to assess the rate of glaucoma instability (ie, those requiring referral back to the conventional clinic for further assessment), safety, rate of glaucoma progression, time efficiency, and cost-savings of this model.

METHODS

Since its introduction, GLOC has been running with the approval of the institution’s management team and in accordance with the consensus guidelines that ensure patient safety in a virtual clinic setting.⁹ The GLOC model can be divided into 3 components:

• Patients undergo relevant tests and investigations;
• The ophthalmologist analyses the data and formulates a management plan;
• The nurses communicate the management plan to the patients.

The data collected in this study were part of a quality assessment and did not require ethics approval.

GLOC Setup

Initially, the investigation area for GLOC was located in the glaucoma clinic at the SNEC. No new equipment or hiring of staff was required. With the expansion of GLOC’s patient criteria, we shifted the investigation room to 2 separate locations outside SNEC main center (named A and B). Each GLOC investigation site was staffed with nurses and ophthalmic technicians who would perform the relevant tests and patient service associates who would provide registration and billing services at the counter. These investigation sites also provided clinical services to other ophthalmology subspecialties. Location A had the following equipment: an iCare tonometer (iCare IC200 Finland Oy, Helsinki, Finland), a Humphrey Visual Field (HVF; Analyzer II: Carl Zeiss Meditec, Dublin, CA), an optical coherence tomography (OCT; Zeiss Cirrus 5000, Oberkochen, Germany), a fundus camera (Topcon TRC-NW8 Oakland, NJ), and an autorefractor (Topcon KR-1). Location B had the following equipment: an iCare tonometer (iCare IC200 Finland Oy, Helsinki, Finland), a HVF (Analyzer II: Carl Zeiss Meditec, Dublin, CA), an OCT (Zeiss Cirrus 5000, Oberkochen, Germany), a fundus camera (Zeiss Cirrus 600), and an autorefractor (Topcon KR 800).

GLOC Workflow

GLOC is an asynchronous telemedicine service, where glaucoma patients visit the GLOC investigation site for the purpose of glaucoma testing, without seeing an ophthalmologist. Patients were referred to GLOC after a face-to-face visit to the glaucoma clinic, where they were vetted for suitability by a glaucoma consultant. The GLOC patient criteria included patients above the age of 21 years who were deemed to be at low risk for glaucoma progression, as well as patients with selected phenotypes and severity of glaucoma (Table 1).

TABLE 1 - Initial and Expanded GLOC Referral Criteria

Initial GLOC Referral Criteria Ocular hypertension (IOP 24–27 mm Hg, CCT ≥500 microns, not on glaucoma medications) Disc suspects Pseudophakic primary angle-closure  (at least 4 mo post–cataract surgery, with IOP<24 mm Hg without medications and no peripheral anterior synechiae on gonioscopy) Normal-tension glaucoma (stable for at least 3 y, not on glaucoma medications) Risk factors for glaucoma including:  a) family history of glaucoma in first-degree relatives  b) pseudoexfoliative syndrome  c) pigment dispersion syndrome  d) angle recession  e) nevus of Ota  f) Sturge Weber

Expanded GLOC Referral Criteria Patients on any number of topical glaucoma medications Mild-to-moderate POAG and pseudophakic PACG (stable for at least 3 y, without filtration surgery; if the patient has previous Schlemm canal minimally invasive glaucoma surgery, he/she should be stable for 2 y postsurgery) Mild secondary glaucoma (stable for at least 3 y, without filtration surgery) Only-eyed patients: without secondary glaucoma in better eye, with mild primary glaucoma on up to 2 medications Patients who require IOP check after starting or changing medications

CCT indicates central corneal thickness; GLOC, Glaucoma Observation Clinic; IOP, intraocular pressure; PACG, primary angle-closure glaucoma; POAG, primary open-angle glaucoma.

The glaucoma severity of each patient was defined by the mean deviation on the Humphrey visual fields. Mild glaucoma was defined by a mean deviation between 0 to −6 dB, while moderate glaucoma was defined by a mean deviation between −6 and −12 dB, and severe glaucoma was defined by a mean deviation worse than −12 dB.¹⁰

Before referral to the GLOC, patients are required to undergo 4 baseline examinations. These include (1) Snellen visual acuity, (2) intraocular pressure (IOP) measurement by Goldmann applanation tonometry, (3) slitlamp examination, and (4) gonioscopy. Visual acuity was tested by nurses, while the other 3 baseline examinations were performed by an ophthalmologist. Baseline ophthalmic investigations were also performed. These include corneal pachymetry (Nidek Echoscan US-500, San Jose, CA) to assess the corneal thickness, HVF (Analyzer II: Carl Zeiss Meditec, Dublin, CA) using the Swedish Interactive Threshold Algorithm (SITA) Fast algorithm with a 24-2 test pattern, or OCT of the retinal nerve fiber layer (RNFL) (Zeiss Cirrus 5000, Oberkochen, Germany). Patients were counseled and given an information sheet about the virtual nature of follow-up at the GLOC.

During the GLOC appointment, nurses would reassess the patient’s Snellen visual acuity and measure the IOP using iCare tonometry. Technicians would perform either an HVF or an OCT-RNFL for the patient, which was preordered by the clinician at the previous review. The test performed would depend on the clinical need and reproducibility of the previous investigation results. Disc or fundus photographs may also be taken at the visit. Thereafter, the patient would leave the clinic. The results were entered into the patient’s electronic medical record (EMR) (Fig. 1).

FIGURE 1: Workflow when patient attends GLOC appointment. EMR indicates electronic medical record; GLOC, Glaucoma Observation Clinic; HVF, Humphrey Visual Field; IOP, intraocular pressure; OCT-RNFL, optical coherence tomography of the retinal nerve fiber layer.

An ophthalmologist from the glaucoma department would review those results remotely within 1 month and enter the management plan into the patient’s EMR. Only 3 outcomes were permitted. The patient might be discharged with no further follow-up, continue follow-up in the GLOC, or return to the conventional clinic for a physical consultation with the doctor. Nurses would then book the subsequent appointment and inform the patient about the outcome of the visit via a Short Message Service. Pharmacy staff would arrange for medications to be delivered to the patient’s home if required.

In the first 2 years of establishing the GLOC, all GLOC patients were reviewed by associate consultants who were trained ophthalmologists undergoing their glaucoma fellowship training. Glaucoma consultants and senior glaucoma consultants, who had completed their glaucoma subspecialty training, were subsequently included in the pool of doctors reviewing GLOC patients to cope with the increased number of patients in the expanded GLOC.

We also created more flexibility in referring patients between GLOC and conventional glaucoma clinics. For example, a patient with moderate disease who was deemed stable may alternate between visits to the GLOC and the conventional clinic during the course of follow-up. This would allow for increased safety, where GLOC is used to supplement rather than totally replace conventional clinics.

Data Collection

The following information about the GLOC patients was obtained from the EMR: age, sex, ethnicity, glaucoma diagnoses, and management plans. Further information was collected on the following outcome measures:

Safety

One hundred consecutive GLOC patients reviewed in February 2022 were selected for the safety audit. This audit was performed by 2 independent glaucoma senior consultants, where the diagnoses and proposed management plan by the reviewing doctor were compared against that of independent glaucoma senior consultants. If the diagnosis and management plan of the reviewing doctor were in agreement with the 2 senior glaucoma consultants, then the case management would be considered safe.

Rate of Instability and Glaucoma Progression

The number of patients whose glaucoma was deemed unstable was recorded. These patients would require referral back to the conventional clinic. The reasons for instability include either one or more of the following: (1) deterioration of visual acuity by more than 2 Snellen lines; (2) IOP above a predetermined target; (3) HVF progression; or (4) OCT-RNFL progression. The patients’ outcomes were monitored over time and the true progression rate was calculated.

Time Efficiency

Doctors of varying seniorities provided an estimated time required to review a single GLOC patient, as an average of over 10 patients. The time required to review patients in a physical consultation was also calculated based on at least 6–10 clinic sessions per grade of doctor and averaged over the number of patients reviewed per clinic session. The time taken for visual acuity, intraocular pressure measurement, and ophthalmic investigations were excluded as these were deemed to be comparable in the GLOC and conventional clinics.

Net-cost Analysis

Annual manpower costs for staff involved in the operation of the face-to-face and GLOC clinics in 2021 were obtained from the Human Resources Department at the SNEC. The expected manpower costs of a given visit were calculated by summing up the costs of performing the eye tests, reviewing and communicating results, and the potential referral back to conventional clinics for unstable GLOC cases. Differences in cost were driven by differences in time to perform each task at each seniority level, the cost of telecommunication for GLOC patients, and the cost of a follow-up conventional consultation for unstable GLOC patients. Although more manpower was required to staff each new investigation site, this manpower was also utilized to provide clinical services to other ophthalmology subspecialties. This cost was hence excluded from our calculations.

Taking into account the risk of glaucoma progression and all-cause mortality, we estimated the lifetime cost-savings for a hypothetical cohort of stable glaucoma patients aged 65 years, over a 25-year horizon using a Markov model with annual cycles and 3 health states: (1) nonprogressive glaucoma, (2) progressive glaucoma, and (3) dead. Patients begin in the nonprogressive state and receive annual follow-up visits. At any given visit, if the investigations showed any signs of change, suggesting the condition to be unstable, the GLOC patient will be recalled for a conventional consultation at the next visit. The patient would then remain in the conventional clinic for subsequent annual visits. Unstable but nonprogressive GLOC patients will return to the GLOC clinic for their next annual visit. In the conventional model, all glaucoma patients are seen in conventional clinics. A 3% discount rate was applied to cost incurred beyond the first year.

RESULTS

Between July 2018 and June 2021, 3458 patients were seen in the GLOC. The mean age was 64.82 (±12.43) years old, 39.4% were male, and most were Chinese. Patients with varying diagnoses were seen, with the majority being disc suspects (57.7%) (Table 2).

TABLE 2 - GLOC’s Patient Characteristics

Patient characteristics	N=3458, n (%)
Sex
Male	1361 (39.4)
Female	2097 (60.6)
Ethnicity
Chinese	3152 (91.2)
Malay	138 (4)
Indian	150 (4.3)
Others	18 (0.5)
Diagnosis
Disc suspect	1994 (57.7)
Ocular hypertension	258 (7.5)
Family history of glaucoma	241 (7)
Pseudoexfoliation/pigment dispersion/angle recession	38 (1.1)
Primary angle-closure suspect	248 (7.2)
Primary angle closure	93 (2.7)
Previous APAC	14 (0.4)
NTG	246 (7.1)
POAG	187 (5.4)
PACG	95 (2.7)
Secondary glaucoma	43 (1.2)

APAC indicates acute primary angle closure; GLOC, Glaucoma Observation Clinic; NTG, normal-tension glaucoma; PACG, primary angle-closure glaucoma; POAG, primary open-angle glaucoma.

GLOC Safety

An audit was performed to assess the safety of the extended GLOC. Two senior consultants performed an independent review of 100 consecutive GLOC patients who had been reviewed by other glaucoma specialists in February 2022. Of the 100 patients, the senior consultants concurred with the management of 95 patients. For the 5 patients with no concurrence, 3 were noted to have diabetes and had not been referred for diabetic eye screening, 1 patient was a disc suspect who should have been discharged, and 1 patient had borderline IOP who should have been trialed off medications and considered for discharge.

Rate of Instability and Glaucoma Progression

Among all the GLOC patients, 6.6% were discharged with no further follow-up required, 78% were stable and given a follow-up appointment, and 14.6% were deemed to be unstable. The criteria for instability included a reduction in visual acuity of 2 or more lines, IOP above a predetermined target, or progression on HVF or OCT-RNFL.

Of the 505 (14.6%) unstable patients, 180 (35.6%) had changes in their HVF, 143 (28.3%) in the OCT-RNFL, 93 (18.4%) patients had IOP above a predetermined target, and 81 (16%) patients had decreased visual acuity (Table 3). The causes of visual deterioration were all due to nonglaucomatous issues, such as cataracts, refractive error, posterior capsular opacification, and age-related macular degeneration. A hundred eight (21.4%) unstable patients were diagnosed with true glaucoma progression and had medication started or added. Hence the rate of glaucoma progression was found to be 3.12%.

TABLE 3 - Reasons for Instability

HVF changes	OCT-RNFL changes	IOP above target	VA deterioration	N=505 (%)
√	—	—	—	180 (35.6)
—	√	—	—	143 (28.3)
—	—	√	—	93 (18.4)
—	—	—	√	81 (16)
√	√	—	—	4 (0.8)
√	—	√	—	1 (0.2)
√	—	—	√	1 (0.2)

HVF indicates Humphrey Visual Field; IOP, intraocular pressure; OCT-RNFL, optical coherence tomography of the retinal nerve fiber layer; VA, visual acuity.

Time Efficiency

The average time taken for a glaucoma specialist to perform a GLOC review was 5.75±0.75 minutes. This period contrasts with the average time taken to perform a conventional clinic consultation, estimated at 13.7±2.3 minutes. This shortened consultation time translates to a 58% of time-saving per GLOC consultation compared with a physical consultation. Hence the glaucoma specialists were able to review a larger volume of GLOC cases compared with the conventional glaucoma clinic in the same amount of time.

Patients who attend the conventional clinic would need to have their visual field tests done on a separate visit 1–2 weeks before, resulting in 2 separate clinic visits for each review. In contrast, patients who attend the GLOC clinic have to attend only 1 session to perform the tests as the doctor’s review can be completed virtually. This reduces travel time for GLOC patients and eliminates the wait time for a physical doctor’s consultation. Unlike the conventional clinic, the GLOC patients did not have to wait at the appointment desk and pharmacy, as they would be informed of the clinical outcome and appointments via Short Message Service and medications would be delivered to their homes.

Cost-analysis Results

Per Visit Savings

With the expanded eligibility and increased number of doctors reviewing patients in the GLOC, the per capita manpower cost per GLOC patient per visit was $36.77 as opposed to $65.62 per patient per visit in the conventional clinic. This translates to a $28.85 of manpower cost-saving per visit to the institution. Cost-savings were mainly driven by less time spent on reviewing cases under the GLOC model.

Lifetime Savings

As of 2021, the life expectancy for Singaporeans is estimated to be 84.3 years for males and 88 years for females. The per capita manpower cost of managing a 65-year-old glaucoma patient under the GLOC model would be $525.75 over his/her expected lifetime, that is, 25 years. This is compared with the cost of a face-to-face model, which is $806.40. This translates to an expected per capita manpower cost-savings of $280.65.

DISCUSSION

To address the issues of increasing health care manpower costs and limited physical space in conventional clinics, the SNEC has adopted and expanded the GLOC to address the increasing burden of glaucoma effectively. At the height of the COVID-19 pandemic, the capacity of conventional clinics was significantly reduced to decongest physical clinics and reduce the spread of infection. In contrast, the rapid turnover of patients in the GLOC model helped avoid stagnation and facilitate social distancing measures, allowing glaucoma patients to have continued access to care, thus reducing the risk of adverse outcomes due to poor follow-up. This model also catered to patients who were uncomfortable with physical consultations during this time.^6,11 We found the expanded model to have a high rate of safety, low rate of glaucoma progression, and to be time-efficient and cost-saving.

Our findings concur with those of Nikita et al⁸ who used similar expanded patient eligibility criteria while providing safe and effective high-quality glaucoma care. Telemedicine for glaucoma has also been shown to cost significantly less than conventional clinics.¹² In our Asian population, the GLOC has been successful at increasing the capacity of our conventional glaucoma clinics while maintaining high standards of safety. From July 2020 to June 2021, 45,416 glaucoma patients were seen in our conventional clinics and 5968 (13%) patients were referred to the GLOC. This translates to a capacity expansion of 13% for conventional visits, allowing increased access to care for patients with more complex needs.

Although the patient eligibility for the GLOC had increased to include patients with mild-to-moderate glaucoma, the rate of instability was low (14.6%), and the true rate of glaucoma progression was even lower (3.12%). This is compared against glaucoma progression in a conventional clinic, which is estimated to be between 3% and 17%, suggesting that the expanded patient eligibility and the referrals made were appropriate for the GLOC model.¹³

The safety audit performed in February 2022 showed that there was a 95% agreement between 2 senior consultants and other reviewing doctors for the management decisions of the GLOC patients. The nature of the disagreement between consultants did not result in harm to the patients involved.

The cost calculations show that our asynchronous VGC is a cost-effective model as it allows more patients to be reviewed within a given amount of time compared with the conventional clinic. There is a 44% reduction in cost per patient per visit to the institution when a patient is reviewed in the GLOC clinic as opposed to the conventional clinic. Over the patient’s lifetime, there is a 34.8% reduction in cost to the institution if the patient is reviewed in the GLOC as compared with the conventional clinic. From the patient’s perspective, there are significant time-savings because there is a reduced waiting time for a doctor’s consultation, reduced need to attend a separate visit for visual field tests unlike in a conventional model, reduced waiting time at the pharmacy, and reduced waiting time for payment and appointments. The cost of the GLOC consultation is also 16.7% lower than that of the conventional consultation.

Our cost calculations did not include drug delivery or the cost of ophthalmic investigations (such as HVF and OCT-RNFL) as these costs were applicable to patients in both the conventional and GLOC clinics and therefore incurred the same cost to the institution. The cost calculations performed took into account the difference in the costs incurred by each clinic due to the different work processes.

However, cost-savings are reduced when GLOC patients are deemed to be unstable, requiring additional assessment. This could be due to reduced visual acuity, inadequate IOP control, or new changes noted on investigations. For safety purposes, this reassessment is normally conducted in a conventional clinic. Of the 14.6% of patients who were deemed to be unstable and referred back to the conventional clinic, less than a quarter were found to have true glaucoma progression requiring intervention. Appropriate patient selection and choice of investigations would reduce the number of patients referred back to the conventional clinic for additional assessment, thereby increasing the cost-savings of the GLOC model.

While there are a number of telehealth programs that have been setup and implemented globally, we have chosen an in-house nurse and a technician-led, asynchronous virtual clinic model. We use the existing infrastructure and some equipment that is already available in the SNEC and satellite centers across the country. This allows for the implementation of telehealth without the typical high start-up cost.¹⁴ This model aims to optimize the efficiency of reviews by the glaucoma specialist by 241% while maintaining safety and a high standard of care. This model may allow for better utilization of manpower resource, the largest health care expenditure, while not compromising outcomes.^3,15

To the best of our knowledge, our study is the first in Asia to describe the long-term outcomes of a hospital-based VGC with expanded patient eligibility. This study has also shown that our GLOC model is safe and provides the benefits of time-savings and cost-savings to both the patient and health care institution over the patient’s lifetime.

Limitations

This study reviews the 3-year data of our GLOC model. Because of the COVID-19 restrictions, we were unable to time the patients’ journeys in the GLOC and conventional models. We could not compare the interobserver agreement between the GLOC and conventional clinics. However, Clark et al¹⁶ have shown a low rate of adverse misclassification in the evaluation of 204 patients. Patient satisfaction was not collected as it was completed in our previous study.⁴

The time taken for the review of GLOC patients was calculated as an average of over 10 patients. The time taken for each face-to-face consultation was estimated by taking the total amount of time required by the doctor to review all patients in 6–10 sessions, divided by the total number of patients in those 6–10 sessions. This has a potential for bias as opposed to measuring the length of consultation in true time.

Although iCare tonometry was used to measure IOP during the GLOC visits, no baseline IOP measurement by iCare was obtained to ensure correlation to baseline Goldmann applanation tonometry. iCare tonometry has been shown to have up to 95% agreement with Goldmann applanation tonometry but may have less reliability at extremes of IOP.¹⁷

CONCLUSIONS

The expansion of inclusion criteria in the asynchronous VGC was shown to be a safe, time-efficient, and cost-effective model of care with a low rate of glaucoma progression. The VGC has been resilient throughout the COVID-19 pandemic, a valuable method of managing glaucoma patients effectively in a remote manner, and has allowed for significant health care capacity expansion despite ongoing global challenges to address the increasing disease burden of glaucoma in our population.