Dr. Shroff’s Charity Eye Hospital (SCEH), based in New Delhi, is a tertiary referral center providing general and subspeciality services and training. SCEH conducts one-day cataract screening clinics in villages closer to the city of New Delhi. This is made possible through the financial support of private donors. These programs target patients who have not sought out care due to limited or non-existent local eye care facilities, financial constraints, or in most case a lack of awareness regarding eye-care. In anticipation of the one-day screening clinics, local organizers are responsible for advertising the clinic (e.g. through posters, flyers, or door-to-door visits), managing volunteers and procuring an appropriate facility. The cost of transportation of patients to SCEH is borne by hospital These camps are managed by Program Manager (Outreach) based at SCEH with regards to planning, monitoring, and strategy.

In our data of the camps conducted by the outreach team in Delhi, between June 2016 and May 2017, out of 8283 population of ≥40 years age screened, only 0.25% (n=21) were identified as glaucoma suspects. This is well below the prevalence (1.6-3.5%) found in the rural and urban studies in India ¹⁹ . Model based studies of community screening for glaucoma in rural and urban India has been shown to be cost-effective [20, 21]. The rationale of this study is to analyse the cost and detection rate of detecting glaucoma in cataract camps. If found to be cost effective, this model can be scaled up for detecting glaucoma suspects from cataract screening camps and referring them for further investigations and management.

In addition to the original team for the outreach camp, one of the trained optometrists, a vision technician for handling FDT and a counsellor would travel for the interventional camps.

All people of age ≥40 years, not detected as having a cataract would undergo screening for glaucoma by trained optometrist using additional equipment. Those detected with cataracts are not included in the study as they would have undergone glaucoma screening as a routine before being advised surgery. Those included in the study would have already undergone vision testing and refraction if indicated. The registration number given at the start of the camp would be recorded by the counsellor. There would be two stations for glaucoma screening. The trained optometrist would carry out disc examination, ACD examination and IOP recording. At the second station FDT would be carried out. Any result beyond cut-off and requiring further intervention would be flagged off. After that the person screened would be received by the counsellor. The test used for screening would be disc examination, FDT and IOP using Icare rebound perimetry.

Training. There will be two optometrists trained for the study. The optometrists selected would already be carrying out refractions, basic history taking and slit-lamp examination. The curriculum would pertain to glaucoma-related history-taking and slit-lamp examination. Specifically, for the study the following training would be included for various test to be conducted in the study

The optometrists would have explained to them the need to calibrate before starting measurements and explained the process of calibration. The principles of avoiding injury while recording IOP would be discussed. The cut-offs being used in the study would be stressed upon

The optometrists would be trained to use direct ophthalmoscope and focus on the disc. They would be exposed to patients and photographs with a variety of disc findings.

The concept of the Van Hericks test would be explained. Although they would be already using a slit-lamp, they would be trained to use a hand-held slit lamp. The cut-offs to be used would be explained. The optometrists would be maintaining a log-book to record their findings, and these would be checked and signed by a glaucoma consultant or a senior fellow.

To carry out a field test with FDT, the optometrists would be trained in handling the equipment, giving instructions to the patients and commenting on the printouts keeping the cut-offs of the study in mind.

The training would be conducted in a classroom with visual presentations for theory. A senior glaucoma fellow ophthalmologist would be involved to take classes with a problem-based approach. Theory class would be scheduled for 1 hour in every working day for 2 weeks. A total of 12 hours would be spent in theory classes. The topics to be covered are mentioned in Table 1.

Junior glaucoma consultant would be involved in imparting practical training. In total, 30 h would be taken to develop their practical skills for each trainee. A log book record would be maintained for the patients examined. The criteria used for assessing the level of training are highlighted in Table 2.

A prospective cohort study of glaucoma detection and cost assessment of the glaucoma screening conducted in the community. The total and incremental costs of the intervention prospectively from a societal perspective, measuring programme, provider and household costs.

The camps will be held in urban slums in and around the national capital region of Delhi in North India. Figure 1 and Figure 2 depict the flow of patients, that will be maintained in the routine camps and intervention camps, respectively.

Figure 1 shows the patient flow in a routine cataract camp, whereas, in Figure 2, the patient flow in camps which constituted interventional arm is depicted.

Included patients would undergo ophthalmoscopy, AC depth examination, FDT and IOP measurement with rebound tonometry.

In hospital, these referred cases will undergo the following examinations:

Those ≥40 year old identified from routine cataract camps without any extra equipment or trained optometrist for glaucoma diagnosis:

Glaucoma suspects ( Table 4) in this arm would come from routine cataract camps if

Undergo tests similar as intervention group in hospital:

Patients suspected glaucoma for the first time will undergo the following diagnostic tests:

If diagnosed as glaucoma ( Table 4), appropriate treatment is advised.

The detection rate or yield will be measured by dividing the number of new cases identified during the study by the total number of cases examined. However, various other indicators will also be estimated through the study, which are expected to throw light on the efficacies of the FDT setting at the intervention camps in comparison with normal cataract camps and conventional detection setting at the hospital. Therefore, the sample sizes are estimated for different components of the study separately, based on the desired accuracy of the indicators at different level. For an accuracy of ±0.1 with a confidence interval of 95% and assuming a rate of around 50% positive predictive value (PPV) and negative predictive value (NPV), the sample size was estimated to be 97 suspects as diagnosed by the additional setting.

The sample size mentioned above is applicable for both FDT (used in camps and hospital set-up B) and conventional detection setting (triage) at the hospital set-up A. This sample sizes are estimated using the following formula ²⁴ :

n = ( z α /2 ) 2 d 2 × p o p u l a t i o n v a r i a n c e

Where

α is such that (1- α) is the probability of confidence interval

Z _1-α/2 is defined as probability {|X|< Z _1-α/2 | X~ N(0,1)} = (1- α)

d = desired accuracy of the estimate. In other words, length of the confidence interval is 2d.

Since the individuals not suspected of having glaucoma in the intervention camps will not be referred to the hospital for confirmatory diagnosis as per the study protocol, the sample for normal population will be captured by FDT setting at the hospital set-up B. We assume that only 75% post-screening follow-up compliance rates, i.e. people suspected of having glaucoma, referred from the intervention camps turning up for confirmatory diagnosis. However, the confirmatory diagnosis test can be conducted on all those diagnosed as suspected of having glaucoma in the hospital arm as they will be present in the hospital itself. It has been assumed, for the purpose of estimation, that the rate of detection of people suspected of having glaucoma would not exceed 12% in the community. We further assume that it can be, in reality, as low as 10%. The number of screenings in the intervention, non-intervention and hospital arm should be adjusted accordingly in order to capture the required minimum sample sizes for people suspected of having glaucoma (required for estimating PPV) and normal (required for estimating NPV) populations. Although the FDT setting in the non-intervention arm will also capture a few suspected glaucoma cases, its primary objective is to identify normal individuals to be referred for confirmatory diagnosis, the target for capturing suspects at the intervention camps is reduced. Conventional detection setting in the the Hospital arm will identify both suspects and normal for confirmatory diagnosis.

Given the above assumptions on (i) post-screening follow-up compliance rates, (ii) prevalent suspect rates for FDT setting and conventional detection setting and (iii) required minimum sample sizes for estimating PPV and NPV, the number of screenings required for different components are estimated as per Table 5.

All people of age ≥40 years, not detected as having cataract who would consent for screening for glaucoma would be included. A trained optometrist and vision technician would conduct various screening tests, such as disc examination, ACD depth, IOP reading, and FDT. Any result beyond cut-off points as mentioned in Table 3 would be flagged as suspected glaucoma cases and would be received by counselor. For those travelling to the hospital, transport and investigations in the hospitals would be offered free of cost. Overnight stay if required would also be provided free.

In the hospital, a detailed workup would be done for the glaucoma suspects. For glaucoma, they would undergo applanation tonometry, gonioscopy, automated visual field analysis and nerve fibre analysis. After the investigations, they would be examined by a glaucoma consultant, who would be masked to the study arm to which patient belongs. A diagnosis of glaucoma, suspected glaucoma, angle closure disease or impending angle closure requiring intervention will be made. The patient requiring intervention would be counselled and others would be discharged. Those requiring surgery would be offered surgery at fixed rates meant for camp patients needing non-cataract surgery. The patients staying overnight would be offered free transport both to the hospital from the camp and back to the camp site after discharge. Records would be maintained of all patients offered free transport.

A prospective observational study of 1000 adults ≥40 years age attending for glaucoma screening in cataract screening camps will be conducted. These adults would be recruited over an 18-month period across camps being conducted in villages near Delhi. The study is ongoing and recruitment is expected to be completed by December 2019.

The comparator group would be an equal number of adults of similar age who are screened in cataract camps in other nearby similar areas by SCEH and are identified as glaucoma suspects and as glaucoma-positive during their visit to the hospital. Additional comparator would also be patients >40 years age of similar socio-economic backgrounds who visit the hospital for the first time as routine or referral patients for eye check-up and are identified as glaucoma suspects and glaucoma positive.

The primary outcome of the screening algorithm would be the detection rate of the screening tests. The detection rate will be calculated by dividing the number of new cases identified during the study by the total number of cases examined [28].

We will use a step-down costing methodology for capturing costs of screening camps, whereby program costs will be entered into a customized tool created in MS Excel (Annexure 4, extended data) ²⁰ . Cost data will be regularly entered into the tool for a period of 3 months to reflect any changes in the cost structure of the screening program. Using a step-down method, the main worksheets for entering data allocate costs to one of the following categories: training, pre-screening and screening.

The costs will be calculated by adding the unit costs of all resources used in the different activities occurring the screening process. This process includes all activities and procedures performed to detect cases with glaucoma from the invitation to participate in the screening campaign to the moment when the ophthalmologist establishes the diagnosis in the hospital. We will use the concept of activity-based costing for the screening program, but for the hospital we will use a standardized cost-accounting system. For the activity-based costing, the costs inputs will include screening invitation, screening costs (health professionals, devices, screening venue etc.). Personnel costs will be calculated from the total cost of each professional participating in a certain activity according to the time specifically dedicated to that particular activity. For example, optometrists dedicated 5–6 hours of their daily work time to the screening and only that part of their full salary will be considered as a cost. For ophthalmologists conducting different tasks (training, examination or test interpretation), only the fraction of their work time dedicated to the specific task involved in the glaucoma diagnostic process will be added.

All diagnostic equipment with a usage life of more than a year will be considered as capital equipment. The annual and monthly rental value of the capital equipment, and screening clinic costs per year/month will be calculated as per methods for assessing costs of glaucoma screening in India in other studies [20, 21]. All other diagnostic equipment will be considered as non-capital and will be considered with a depreciation rate of 100% in the first year itself.

Key information interviews with Program Manager (Outreach) will be conducted in identifying any donated goods requiring reevaluation and in allocating joint costs between program components. The allocation of joint costs to program components and activities, will also be informed by monthly sheets of screening camps.

The fixed costs at the hospital (electricity, water, gas, maintenance, and security) will be estimated from the annual structural costs of the facility divided by the number of patients seen and multiplied by the number of patients evaluated in the hospital in both screening and conventional arms. The structural costs at the hospital (glaucoma consultation) will be calculated from the hospital cost-accounting system.

All costs captured from accounting statements, such as fixed costs, salaries, etc., will be converted to economic costs. This means that capital costs will be annualized over their expected useful life, and any donated goods or volunteer time appearing as zero costs in the accounting data, or not appearing in the accounting data in any form, will be added to the cost sheets and assigned their current market value. Costs to the community would be captured using a structured questionnaire and will include direct medical (consultation, medicines etc.), direct non-medical (travel), and indirect costs (wage loss of patient) (See Annexure 2, extended data) ²⁰ .

For the cost-effectiveness analysis, the cost per case detected would be the primary effectiveness outcome measure, in comparison with conventional detection, i.e. opportunistic case finding in hospitals.

The costs for the screening program will be based on the study sample and may vary significantly depending on the size of the target population and other factors. Hence, a sensitivity analysis will be performed using mean of detection rates from other studies in similar developing country settings.

The costs will also be recalculated for the community of 1 million subjects in order to generalize the findings to India as a whole. Costs will be presented in 2019 prices in Indian rupees and United States dollars (USD). Since the study period is less than 12 months, costs will not be adjusted for inflation, nor will discounting be considered for costs and outcomes.

Data entry will be conducted by a single data entry operator at the research unit of SCEH. One of the co-authors (AM) will be reviewing the data entry to check for any discrepancies including any data entry errors from the data entry form. The data will be stored in a desktop computer with access to the data entry operator, and co-auth or (AM). Once the data entry is completed, and cleaned, the data sheet will be transferred to laptops of co-author (AM & DJ) for further analysis. After the analysis these data sheets will be destroyed in these laptops and the data sheet would be available only with the desktop present at research unit of SCEH.

The patients being screened would be undergoing standard investigative procedures and no experimental investigation or procedure would be carried out. Informed consent would be taken before administering the questionnaire regarding cost incurred by patient/relative for screening and hospital examination (Extended data, Annexure 3) ²⁰ . The identity of the person would not be disclosed, and no identifiable data would be shared or published. The identifiable data stored in database would be password protected.

The study received ethical approval from Dr Shroff’s Charity Eye Hospital-Ethics Committee (study reference number SCEHEC/2018/02) on 13.02.18.

The study results will be submitted to a suitable peer-review publication within 6 months of study period (i.e. patients entering camps and then arriving at hospital) with an acceptable sample size as described in relevant section above. Additionally, the results will also be presented in suitable national/international conferences based on resources available for participation.