The FLuorometholone as Adjunctive MEdical Therapy for Trachomatous Trichiasis Surgery (FLAME) Trial: Study Design

Abstract

Purpose:

To report the design of FLuorometholone as Adjunctive MEdical therapy for TT surgery (FLAME) trial.

Design:

Parallel design, double-masked, placebo-controlled clinical trial with 1:1 randomization to fluorometholone 0.1% eye drops twice daily or placebo twice daily for 4 weeks in eyes undergoing trachomatous trichiasis (TT) surgery for assessing the efficacy, safety, and cost-effectiveness of fluorometholone 0.1% in preventing recurrent postoperative trichiasis (PTT).

Methods:

Up to 2500 eligible persons with trachomatous trichiasis undergoing lid rotation surgery were enrolled in the Jimma zone, Ethiopia. Participants, surgeons, study field staff, and study supervisors leading operational aspects of the trial are masked to treatment assignment. Randomization is stratified by the surgeon and is simultaneously stratified by the district. Study visits (in addition to programmatic follow-ups) are at the baseline/enrollment, at four-week post-enrollment, and after 6 months and 1 year. The primary outcome is cumulative one-year PTT incidence, defined as: ≥1 lashes touching the globe, evidence of epilation, and/or repeat TT surgery. Secondary postoperative outcomes include the number of trichiatic lashes, location (touching the cornea or not), evidence of post-operative epilation, entropion, changes in corneal opacity, IOP elevation, need for cataract surgery, visual acuity change from the baseline, eyelid contour abnormality, granuloma, eyelid closure defect, and the occurrence of adverse events. Health economic analyses center on calculating the incremental cost per case of PTT avoided by fluorometholone treatment.

Conclusion:

The FLAME Trial is designed to provide evidence of the efficacy, safety, and cost-effectiveness of adjunctive topical peri-/postoperative fluorometholone 0.1% therapy with trichiasis surgery, which is hypothesized to reduce the risk of recurrent trichiasis while being acceptably safe.

Clinical Trial Registration: https://www.clinicaltrials.gov/study/NCT04149210

Introduction

Trachoma is the leading infectious cause of blindness worldwide.^1,2 Trachomatous trichiasis (TT), a sequela of conjunctival scarring resulting from repeated infection/ chronic inflammation, is a key mechanism leading to blindness, and also causes severe, chronic eye pain. Approximately 1.9 million people are visual impaired from trachoma, including 0.45 million who are irreversibly blind; as of April 2023, it was estimated that approximately 1.5 million people have untreated TT and are at risk of blindness in addition to having eye pain.³ TT surgeries are performed by trained eye care workers as part of the permanent public health system to speed up the TT backlog clearance.⁴

Surgery to relieve trichiasis is one of the four World Health Organization (WHO)-endorsed “SAFE” Strategy priority interventions for programs aiming to prevent trachoma blindness.¹ The elements of the strategy, under the acronym “SAFE,” include: Surgery for trichiasis, Antibiotics to clear C. trachomatis infection, Facial cleanliness to reduce C. trachomatis transmission, and Environmental improvement to remove conditions facilitating endemic trachoma. Unfortunately, a high recurrence rate following trichiasis surgery limits the benefits of surgery and also reduces community confidence in the surgery, resulting in less utilization by people who could benefit from it. Reports of the incidence of recurrence of TT following TT surgery (“post-operative TT” [PTT]) have been highly variable and are hard to combine due to heterogeneous methods used in reporting. In a comprehensive review summarizing reports through 2012, the median for the currently WHO-recommended procedures (bilamellar tarsal rotation and posterior lamellar tarsal rotation) appears to be in the 20–30% range, with programmatic results tending to be less favorable than clinical trial results, which often involve merit-based selection and intensive training of TT surgeons.⁵ Reports based on selection of the best TT surgeons followed by intensive training in a clinical study, such as a trial, would not be expected to be representative of typical programmatic conditions.

Reported risk factors for PTT include surgeon skill, severity of preoperative disease, older patient age, and ongoing inflammation during the perioperative period.^6–13 Optimization of surgical quality is acknowledged as critical for obtaining good results of surgery, and a subject of considerable programmatic effort, but such efforts have not eliminated recurrence.¹⁴ PTT is a difficult problem, with worse outcomes of surgical repair than primary TT.⁶ According to the WHO, the most experienced surgeon available should manage PTT; epilation is encouraged between diagnosis and review by that professional.¹⁵ Because of the dire situation of patients who have PTT – with high risk of blindness, ongoing pain, and limited management options – and the impact on community observations of such cases in reducing interest in sight-saving TT surgery, avoidance of PTT is of paramount importance.

Ongoing inflammation in the setting of trachoma is associated with progressive conjunctival scarring¹⁶ and often is seen in persons with TT.^9,17,18 Such inflammation is only rarely associated with active C. trachomatis replication,¹⁹ and the specific causes thereof are incompletely understood. In the STAR Trial, where azithromycin therapy was associated with the reduced risk of postoperative TT even though detectable C. trachomatis infection was rare prior to treatment, anti-inflammatory effects of azithromycin were cited as one potential mechanism by which benefit may have occurred.⁸ Considering the potential importance of ongoing perioperative inflammation in relation to PTT incidence and the rarity of active C. trachomatis replication among TT cases, members of our group previously undertook a randomized phase 2-like clinical trial evaluating whether the perioperative use of a low potency, low-cost corticosteroid eye drops (fluorometholone 0.1%) might improve outcomes. Upon evaluating three potential doses, we found that all three were associated with an approximate one-third reduction in PTT compared with placebo-treated eyes and with contralateral untreated eyes; the safety profile appeared favorable.²⁰

Because the phase 2-like trial was not sufficiently powered to definitively prove that topical fluorometholone 0.1% is effective in reducing the incidence of PTT, we have undertaken a full-scale clinical trial adequately powered to settle the issue, described here: the FLuorometholone as Adjunctive MEdical Therapy for TT Surgery (FLAME) Trial. The specific aims of the FLAME Trial are as follows: 1) to assess the efficacy of fluorometholone 0.1% one drop twice daily for 4 weeks in reducing the incidence of post-operative TT when given as adjunctive therapy with TT surgery in the programmatic setting; 2) to assess whether such treatment is sufficiently safe for wide-scale implementation in TT programs; and 3) to estimate the costs of adding fluorometholone 0.1% treatment to TT surgery per case of postoperative TT averted and to characterize the value of such treatment under a range of plausible health economic circumstances.

Methods

Overview of trial design

The FLAME study is a National Eye Institute-funded randomized, double-masked, parallel-group, placebo-controlled clinical trial (Clinicaltrials.gov # NCT04149210) in which a total of up to 2,500 eligible study participants with TT undergoing lid rotation surgery (Table 1) are enrolled to assess the efficacy and safety of fluorometholone 0.1% as ancillary therapy for trachomatous trichiasis (TT) surgery. The Trial’s location is in the Jimma zone of the Oromia Region of Ethiopia, wherein the SAFE strategy for trachoma blindness mitigation is being implemented (including the conduct of TT lid rotation surgery for a large volume of patients). Study participants are randomized in a 1:1 ratio to fluorometholone 0.1% versus placebo (artificial tears). Participants may have one or both eyes enrolled in the trial; when both eyes are enrolled, the participant uses the same treatment for both eyes. Patients are examined at the baseline, 4 weeks post-surgery, 6 months, and 1 year at the trial exit (see Table 2). Before recruiting any study participants into the study, written approvals of the protocol and of the informed consent form were obtained from all governing Institutional Review Boards (IRBs), and approvals were maintained thereafter. The study participants’ confidentiality is protected by providing each eligible participant with an enrollment ID and locking away all documents containing identifiers, such participants’ consent forms, contact information forms, and enrollment logs. The Ethiopian IRBs include the National Research Ethics Review Committee of the Ministry of Education (formerly under the Ministry of Science and Technology) and the Ethiopia Food and Drug Administration. The Oromia Health Bureau also reviewed and supported the study. In addition, the IRB approval was obtained from the Study Chair’s institution (Massachusetts Eye and Ear under Mass General Brigham, Boston, Massachusetts, USA) and the Vice-Chair’s institution (London School of Hygiene and Tropical Medicine, London, United Kingdom). As per the National Institutes of Health “single IRB” requirements, the Data Center Director’s institution (University of Pennsylvania) formally ceded responsibility for the IRB approval to Mass General Brigham’s IRB.

Table 1.

Inclusion and exclusion criteria for the FLAME trial.

Study participants are eligible for the study if they have been selected to have TT surgery by the Surgical Team, all of the Inclusion Criteria are met and none of the Exclusion Criteria are met.

Inclusion Criteria:

(1) Age 15 years or more, corresponding to the age of patients treated in the Fred Hollows Foundation/Federal Ministry of Health Program at field sites without general anesthesia (i.e., under local anesthesia).

(2) One or both eyes with upper lid trachomatous trichiasis (TT) – with one or more eyelashes touching the eye or evidence of epilation – for whom a decision already has been made to undergo TT surgery on at least one upper eyelid.

(3) Collection of all baseline data prior to randomization

(4) Signed, informed consent (and assent, when applicable)

Exclusion Criteria:

(1) Contraindication(s) to the use of the test articles, including a known allergy or sensitivity to the study medication (fluorometholone) or its components, and contraindication(s) to use of azithromycin

(2) Patients known to be pregnant

(3) IOP ≥ 22 mmHg and/or currently taking ocular anti-hypertensive medications in the study eye (prior IOP-lowering surgery is acceptable)

(4) A known severe/serious ocular pathology or medical condition which may preclude

(5) study completion or increase the risk of harm in the study (e.g., suspicion of non-trachomatous active ocular infection or suspicion of glaucoma of a degree to which where an intraocular pressure spike would be vision-threatening).

(6) Any condition known to be present at baseline for which it is anticipated that ocular or systemic corticosteroid therapy will be required.

(7) Any significant illness or condition (e.g., hypertension with systolic blood pressure ≥ 170 mmHg and/or diastolic blood pressure ≥ 110 mmHg) that could, in the study team’s opinion, be expected to interfere with the study parameters or study conduct; or put the patient at significant risk.

(8) Previous upper lid TT surgery on all eyes with upper lid TT. (If one eye has previously undergone upper lid TT surgery but another eye with upper lid TT has not, the patient may be enrolled, and only the latter eye will be counted for the primary analyses).

Safety considerations regarding fitness for surgery are addressed by the Surgical Team in its selection of persons for surgery, prior to a person entering the Screening Population pool of persons.

Table 2.

Visit and data collection schedule.

Assessments	Visit 1 Baseline	Visit 2 Surgery	Visit 3 4 weeks	Visit 4 6 months	Visit 5 12 months
Timing/Interval (Days[D]) pre/post-surgery (= D 0)	D–7 to 0	D 0	D 26–35	D 180 ± 60	D 365 ± 90
Informed Consent, Demographics, Randomization	X
Medical and Ophthalmic History	X		X	X	X
External Examination using + 2.5 magnifying loupe	X			X	X
Visual Acuity, Trachoma & Trichiasis Grading; IOP;	X		X	X	X
Patient reported outcomes (pain, EQ5D, OSDI, poverty**)
Medication Review	X	X	X	X	X
Adverse Event Review; medical care utilization		X	X	X	X
Surgical Details		X
Assessment of Treatment Adherence			X
Study Participant Exits Study					X

^*D = days post TT surgery; IOP = intraocular pressure; EQ5D = EuroQol health utility (Ethiopian form), OSDI = Ocular Surface Disease Index,

^**Poverty assessment at visits 1 and 5 only.

Study organization

The study group consists of four resource centers: the Chairman’s Office, the Data Center, the Field Coordinating Center, and the Surgical Team. The Chairman’s Office provides overall leadership to the FLAME Trial Research Group mentioned in the credit roster. The Field Coordinating Center (FCC) oversees the operational aspects of the study in a manner typical for a clinical trial coordinating center, except for handling the data system, data management, and statistical aspects of the study which are managed by the Data Center. The FCC oversees field implementation of the trial in Ethiopia, employing, training, and supervising all field team members, except those involved in TT programmatic services under the Surgical Team. The Surgical Team oversees the implementation of TT surgery, including conducting TT patients’ mobilization (through mass mobilization and house-to-house visits) and presenting TT patients to the FCC study field team for the consideration of enrollment into the study. The Data Center oversees and manages the data system and conducts data management and statistical analysis for the study.

Study treatment

The study treatments are fluorometholone 0.1% or placebo, with each one drop two times daily for 4 weeks. The study test articles were prepared by the Penn Investigational Drug Service, under the Data Center in Philadelphia, repackaging active treatment or placebo into identical bottles and shipping them to Ethiopia. Fluorometholone 0.1% suspension was donated by Allergan, now operated by Abbvie (North Chicago, Illinois, USA) (FML® Ophthalmic Suspension, USP 0.1%). Artificial tears (Soothe XP® artificial tears produced by Bausch and Lomb) were purchased; both have a similar milky appearance. The shelf-life of the repack-aged drug was assessed by testing a sample of bottles at 6, 12, and 18 months after preparation, and it was determined to be as potent as expected for at least 12 months, after which the study drug is considered expired. The packaging contains a medication box number corresponding to the masked contents of the bottle, on a peel-off label. Each box number maps to the masked treatment group; all are unique so as to prevent unmasking of a large segment of the study population should unmasking of one participant’s treatment be required.

Study participants are asked to complete a diary provided by the study field team regarding eye drop use and to bring the diary and their medication bottle at the week 4 visit. Adherence is assessed by questioning patients at the week 4 visit, by tabulating the number of eyedrop administrations recorded in the diaries and by evaluating the difference in weight of the eye drop bottles from the baseline to 4 weeks using scales (OHAUS™ Navigator™ Portable Balance, Model: NV222) able to detect a difference in weight of 0.05 g (the approximate weight of one eye drops).

Study participant selection and eligibility criteria

Patients who have TT confirmed by the study’s Surgical Team, who plan to undergo upper lid rotation surgery conducted by a participating surgeon, and who potentially are interested in participating in the study formally provide informed consent to undergo screening procedures to determine eligibility (Table 1). Those who are eligible are invited to participate in the study and go through the enrollment process get examined and are randomized into the Trial. Both participants and those who elect not to participate undergo TT surgery as per standard of care.

Surgical treatment

Since the study treatment with fluorometholone 0.1% would be added to existing programs if cost-effective, the goal of the study is to assess the impact of adding fluorometholone treatment to an unaltered existing programmatic system. Therefore, the study does not seek to change the practices of the participating surgeons or the standard operating procedures of the extant TT surgery outreach program. In this region, new surgeons are taught to perform Posterior Lamellar Tarsal Rotation (PLTR or “Trabut”), which was found superior to Bilamellar Tarsal Rotation (BLTR) surgery in a randomized trial reported during the conduct of our preliminary trial,¹⁹ consistent with the WHO guidance for TT surgery as part of the SAFE program. However, the WHO still recommends BLTR to surgeons who are more familiar with BLTR. Participating surgeons can use whichever of these two procedures they prefer. Thirty surgeons, previously trained and assigned to work in various districts of the Jimma zone but with a varying number of years of experience ranging from less than a year to over 6 years, have participated in the FLAME Trial. Among these surgeons, 23 perform PLTR surgeries and 7 surgeons use the BLTR surgical method. Dissolvable sutures are used exclusively.

Visit procedure and schedule

Study participants are scheduled to be seen four or five times in total. After a baseline visit at the time of enrollment (Visit 1) and a surgical visit (Visit 2), which typically occur on the same day, the study participant is scheduled to return for follow-up at 4 weeks (day 26–35) (Visit 3), 6 months (±60 days) (Visit 4), and 1 year (±90 days) (Visit 5) after enrollment/trichiasis surgery (Table 2). After completing the one-year visit, all study participants exit the study. The study participants’ contact information and residence location are collected at the baseline, and each participant is given a patient ID card with visit dates to facilitate follow-up visit procedures.

Per the extant TT program, separate from the trial, surgeons are expected to examine TT patients after their surgery at various times such as the first postoperative day, after 8–14 days, and at 3 months. Limited data about complications are collected at those clinical care visits.

Randomization procedures

A surgeon-stratified randomization schedule was prepared in advance by the study Data Center. Randomized patients either immediately undergo the Visit 2 procedures (which includes trichiasis surgery) or else can be scheduled to return to the center for Visit 2 within 7 days.

Package boxes of 20 medication kits are prepared by the Investigational Pharmacy in using random block sizes of two or four in such a way that every package box is comprised of 10 fluorometholone and 10 placebo bottles. A specific package box is used by each surgeon, who works through one box at a time, thus achieving stratified randomization by surgeon. Surgeons operate in their own geographical areas. Thus, this method of stratification also ensured that there would be an approximately balanced number of patients in each treatment group within each geographical area. When package boxes are incompletely used by a given surgeon, they can be repackaged per instructions from the unmasked biostatistician at the Data Center to create a new package box with 10 active and 10 placebo treatment bottles so as to avoid wasting treatment bottles.

At the surgical location, after obtaining consent and confirming eligibility, the study team enters the patient’s name and date of enrollment on the Enrollment Log for study patients of each participating surgeon. A preselected enrollment ID is assigned to each sequential patient. A medication box number is assigned to each enrolled patient following a randomization sequence, and the medication box kit number label is put on the enrollment form, thus accomplishing randomization at enrollment. Therefore, each patient has a screening ID number (one of each of these is given to each patient screened, including those not ultimately enrolled, for purposes of constructing a CONSORT Diagram). In addition, each enrolled patient receives a patient ID number and a uniquely identified medication box kit. Study participants are counted as enrolled in the study and part of the safety and efficacy analyses, from the moment the patient’s name is entered in the Enrollment Log. Medication box number(s) are used to link to the simultaneously issued actual treatment assignment for the statistical analysis of efficacy and safety data.

After treatment assignment, the study Field Team proceeds to implement the treatment protocol for Day 0, treating the study eye(s) with an eye drop from the assigned treatment kit prior to surgery. The treatment kit contains a bottle with sufficient fluorometholone 0.1% or placebo for all the study eyes to receive 4 weeks treatment. The study participants are carefully instructed on how to apply eye drops, how often to apply them, and for how many weeks. Participants practice using an eyedrop from single-use preservative-free artificial tear containers and a drop from the assigned medication bottle. After receiving their first assigned eye drop treatment just prior to surgery, study participants undergo TT surgery and then resume their study treatment upon removal of any bandages or eye patches the following day at the Day 1 postoperative follow-up performed by the TT surgeons.

Outcome assessments

Assessments of outcomes and other study variables are measured primarily during eye examinations. Efficacy assessments include trichiasis and entropion grading. Postoperative TT is defined as at containing least one of the following: (1) one or more lashes touching the globe in an eye; (2) evidence of epilation on clinical examination; and/or (3) history of repeat TT surgery. Additional assessments include clinical signs of trachoma status, visual acuity, patient-reported outcomes, and elements of cost calculations as mentioned in the section regarding health economic analysis. Adverse outcome assessments include examining corneal scarring, assessing, and documenting the presence of lagophthalmos, and other conditions such as overcorrection, lid margin notching, granuloma, blepharitis, necrosis of eyelid margin, IOP elevation, occurrence of cataract surgery (within the study period), infectious keratitis, and other observed adverse events attributed to study treatment.

Eye examinations are conducted by masked, trained, and study-certified field team nurses, using 2.5× magnifying loupes and a torch. The outcome assessment and examination details along with the visits are given in Supplementary Appendix 1.

The primary outcome of the FLAME Trial is the cumulative incidence of PTT as determined by the trained study team members at 4 weeks, 6 months, and 1 year.

In addition, safety outcomes, patient-reported outcomes, and health economic issues are evaluated. Safety outcome measures include intraocular pressure (IOP) elevation, cataract, and adverse events attributed to study treatment. Patient-reported outcomes include overall patient satisfaction, cosmetic outcome, impact on quality of life, Ocular Surface Disease Index,¹⁶ and the five-level EQ5D and thermometer EuroQol Questionnaire.¹⁷ The latter primarily uses a new official translation into the Oromo language, although pre-existing Amharic and English versions also are available.

Quality assurance

This study is monitored by various study team members at the FCC, Chair’s Office and Data Center using different procedures to assure protocol adherence and high-quality data. The specific quality assurance features related to data integrity of the FLAME Trial are:

• Standard data collection forms and procedures;

• Standard electronic case reporting data system (REDCap) with the built in functions for checking of data ranges, consistency, and completeness;

• Masking of study participants, surgeons, and all study staff in Ethiopia to the assigned treatment group of the study participant;

• Training and certification of study coordinators in data collection and data entry;

• Explicit instructions with the release of new/ revised data collection forms about new/revised questions and instructions;

• Frequent checking of data not only mainly by the Data Center biostatistician but also by the field teams, supervisors, and internal monitors to detect any data problems early and provide feedback to the field teams for any correction;

• Timely data detection (e.g., range checks) and edits for missing, invalid, and unexpected (potentially erroneous) responses;

• Regular reporting on performance (timeliness, completeness, and accuracy of data) of field teams for the data collection;

• Regular field observation and review of data collection process by study officers and team supervisor, including reexamining certain study participants

• Checking by the study supervisors, field study teams, and internal monitors of a random sample of entered data by the field teams against the original source data after data editing has been completed.

• Supervision of the Trial by a Data and Safety Monitoring Committee (DSMC) which is an independent board appointed by the National Eye Institute (NEI) as advisory to the FLAME Trial Research Group and the NEI. The DSMC meets approximately biannually and is responsible for ongoing review of the efficacy and safety data, policy and ethical issues, and study performance.

Sample size

Prior to the enrollment, the sample size was estimated for comparing the fluorometholone 0.1% treatment group and placebo control group for the cumulative TT recurrence rate by 12 months at a type 1 error rate of 0.05. Assuming TT recurrence by 12 months of 20% in the placebo group, and 15% in the fluorometholone group (a 25% reduction in TT rate), 75% participants with bilateral TT surgery, inter-eye correlation coefficient of 0.48 in TT recurrence^19,21 and 10% lost to follow-up, a total sample size of 2254 participants (3944 study eyes) provided 90% power for comparing differences using repeated measures logistic regression for per-eye analysis with the inter-eye correlation adjusted by generalized estimating equation (GEE).²² However, during enrollment, the bilaterality proportion was found to be considerably lower than 75%, so sample size re-calculation was performed using the observed bilaterality rate of 35% as well as a slightly lower observed one-year cumulative incidence of postoperative TT in the placebo group (slightly less than 19%), and the observed lower rate of loss to follow-up (2%), which provided a total sample size target of 2400 participants for enrollment. A revised sample size of approximately 2400 was approved by the DSMC.

The number of eligible participants for enrollment and the number of eligible eyes required are in the following Table 3 for the assumed PTT rate ranging from 17% to 20% in the placebo group. Because in a field trial simultaneously enrolling large numbers of participants daily at sites often off the cellular grid, it is not possible to reach an exact number and stop, a target enrollment of up to 2,500 participants (1,250 participants per group) was selected, the revised total approved by the study IRBs.

Table 3.

Numbers of participants and eyes required to enroll for PTT specific incidence rates in 1 year.

Postoperative trachomatous trichiasis (PTT) incidence rate in one year
Placebo group	FML group	Rate Difference	% Reduction	Number of Eligible Participants Enrolled	Number of Eligible Study Eyes
20.0%	15.0%	5.0%	25%	2190	2956
19.0%	14.3%	4.7%	25%	2383	3217
18.5%	13.9%	4.6%	25%	2434	3286
18.0%	13.5%	4.5%	25%	2484	3354
17.5%	13.2%	4.3%	25%	2665	3598
17.0%	12.8%	4.2%	25%	2727	3681

Statistical analysis plan

Since some participants undergo concurrent TT surgery in both eyes, and their treatment outcome (i.e., recurrence of TT postoperatively) potentially is correlated, the primary assessment of efficacy is the comparison of cumulative proportion of incident post-operative TT by 12 months between the two treatment groups using a repeated-measures logistic regression model, where the inter-eye correlation is accounted for through generalized estimating equations. The difference for cumulative incidence rate of postoperative TT by 12 months, the odds ratio and their 95% confidence intervals for the comparison between the two treatment groups are calculated from the repeated measures logistic regression model. Analyses are performed with their treatment group assignment based on randomization regardless of their treatment compliance (“intent-to-treat”).

To check the consistency of results over subgroups, the effect of modification of the treatment on the primary outcome is assessed with the following factors by including the treatment group indicator, subgroup indicator, and their interaction term in the repeated measures logistic regression model as described above. If we find any important interactions, stratum-specific recurrence rates by the treatment group and their odds ratios (ORs) for the treatment effect will be reported for each effect modifying variable, which are considered for interactions.

1. Baseline upper eyelid trichiasis severity with two severity categories (severe: 6 or more in the total number of upper eyelid lashes or epilation >1/3; not severe: 5 or less in the total number of upper eyelid lashes or epilation < 1/3).

2. Baseline conjunctival (papillary) inflammation (presence or absence)

A secondary analysis also will be performed based on the level of adherence to randomized treatment with fluorometholone or placebo. For this analysis, good treatment adherence will be defined as one of the following criteria meet:

1. Bottle weight change indicates >75% of the expected doses. and the medication diary indicates >75% of the expected doses;

2. Medication diary indicates >75% of the expected doses. and the bottle weight change is unknown;

3. Bottle weight change indicates >75% of the expected. doses and the medication diary is unknown;

4. Self-reported adherence is very good, when both the medication diary and the bottle weight change are unknown.

Because only a small percent (<5%) of participants may be lost to follow-up or will not comply with the trial protocol during the study, the primary statistical analysis for primary outcome and secondary outcomes will be based on participants who complete the 1-year follow-up (completed cases) with their treatment group assignment classified as assigned at randomization (“intent-to-treat”). Sensitivity analyses including “per protocol” analysis and missing data imputation will be performed to assess the robustness of the results with respect to loss to follow-up and non-compliance with the eligibility criteria and the treatment protocol.

Health economic analysis

The cost of the intervention will be calculated as the cost of the drops of fluorometholone 0.1% that are dispensed to the patients. The price of the generic drug will be based on what importers charge for the fluorometholone 0.1% suspension in Ethiopia. The cost of post-index surgery medications related to the eye will be derived from the surgeon-reported medications for each patient. The costs of the program in Ethiopia also also be taken into account. The cost of any additional medical care utilization (not including study visits) will include utilization based on self-reports and a price based on standard use by a clinic in Ethiopia. Costs of reoperation for TT specifically will be considered. All costs will be converted to US dollar values at the time of the study, based on the international exchange rate. The outcomes will be cases of post-operative TT averted. If the proposed intervention is more effective and more expensive, the incremental cost-effectiveness ratio will be calculated as the difference in costs divided by the difference in postoperative TT cases. If one intervention is more effective and less expensive, then it will be described as dominating and is the obvious economical choice. Quality-adjusted life years (QALYs) will also be calculated based on the change from the baseline for the intervention and control group and take the difference. The incremental cost-effectiveness ratio will be expressed as dollars spent per QALY gained. Finally, sensitivity analyses varying one variable at a time will be conducted to determine whether any reasonable variation in costs or any observed variable being at the high or low end of a confidence interval rather than at the mean would change the interpretation of the economic value of the intervention (i.e., it would go from being cost-effective to not or vice versa.) If a change in one variable at a time or a change when making all variables either the most or least favorable to fluorometholone would change the qualitative interpretation of the results, the results will be bootstrapped to characterize the level of certainty about the conclusion using a cost-effectiveness acceptability curve.

Conclusion

The FLAME Trial is a randomized, double-masked, placebo-controlled parallel sequential randomized field clinical trial to evaluate the efficacy, safety, and cost-effectiveness of fluorometholone 0.1% eye drops as an ancillary therapy for trachomatous trichiasis surgery in Ethiopia. The primary outcome is the cumulative incidence of postoperative TT. Additional outcomes include ocular surface outcomes, potential adverse outcomes of surgery or study treatment, patient-reported outcomes, health economic analysis, and presenting visual acuity. Preliminary data and theoretical suggestions have led us to hypothesize that fluorometholone 0.1% will be efficacious and sufficiently safe for programmatic use as well as cost-effective.²⁰ The FLAME Trial is expected to conclusively confirm or refute these hypotheses. If the hypotheses are supported, adding fluorometholone 0.1% eye drops to TT surgery programs would be justified and might avert large numbers of cases of post-operative TT with high risk of vision loss, chronic eye pain, and limited options. Based on the current backlog, a reduction in postoperative TT from 20% to 15% corresponds to about 75,000 fewer people with postoperative TT globally; a larger effect size would avoid more postoperative TT cases. Additional incident cases are expected for many years into the future; a lower risk of PTT also would benefit those persons with new TT cases.

Supplementary Material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/09286586.2024.2415052.