A Trial of Non-absorbable Versus Absorbable Sutures for Trichiasis Surgery

Primary outcome measure:

Recurrent trichiasis, defined as one or more eyelashes touching the globe or evidence of epilation (lash stubs) on examination, or a history of repeat trichiasis surgery since the baseline surgery, at one year.

Justification of primary outcome:

The aim of surgery is a permanent solution for trichiasis. If a patient has eyelashes touching the eyeball or there is clinical evidence that they are epilating to prevent this or they have required repeat trichiasis surgery, surgery can be deemed a failure.

Analysis of primary outcome measure

Intention to treat analysis of risk factors for recurrence.

Primary analysis of primary outcome:

It is expected that the important baseline characteristics (those that are known to affect the risk of recurrence) will be balanced between the two arms by stratified (for operating surgeon) randomisation. If this is found to be the case, the outcome of the surgery in the two arms will be compared in an unadjusted logistic regression model for recurrent trichiasis at one year. If the arms are found to be substantially imbalanced with respect to, pre-operative severity, gender or age, an appropriately adjusted logistic regression model will be used.

Individuals will be considered to be recurrent trichiasis cases if they have developed recurrent trichiasis at anytime during the one year follow-up period.

Secondary analysis of primary outcome:

• Effect modification We will assess effect modification of the intervention on recurrence at one year with the following factors by including an interaction term with treatment arm in the logistic regression model.

• Operating surgeon
• Pre-operative severity
• Sex
• Age group
• Conjunctival (papillary) inflammation
• Surgical complications
• Analysis of risk factors for recurrence:

A multivariate logistic regression model will be used to identify potential explanatory factors for recurrent trichiasis at one year, in addition to the randomised suture type allocation. Other factors which will be examined in a model of recurrent trichiasis will include

• pre-operative disease severity

• trichiasis
• entropion
• corneal opacity
• surgeon
• infection status at baseline and follow up timepoints
• inflammation status at baseline and follow up timepoints
• whether epilation is being practised
• sex
• age
• geographic location of patient's home
• literacy
• body mass index
• eye dryness

• symptoms
• signs
• Impact of intervention on time-to-first-recurrence. Kaplan-Meier analysis will be used to plot the survival curves for both treatment arms up to the final visit at 2 years. Cox regression will be used to assess the impact of the intervention on time to first recurrence. The hazard ratio will be estimated with Cox regression, adjusting for substantial baseline imbalances if appropriate.
• Recurrence at two years Intention-to-treat analysis will be used to assess the impact of the intervention on failure at 2 years. Failure will be defined as one or more eyelashes touching the globe or evidence of epilation (lash stubs) on examination, or a history of repeat trichiasis surgery since the baseline surgery, at either one year and/or two year follow-up visits.

Secondary outcome measures

• Lashes touching the eyeball The effect of the intervention on the number of lashes touching at one, and two years, respectively, will be analysed using zero inflated Poisson regression (i.e. treating number of lashes as a continuous variable)
• Entropion grade The effect of the intervention on the degree of entropion will be analysed by ordered logistic regression
• Conjunctivilisation of the lid margin grade The effect of the intervention on the degree of conjunctivilisation will be analysed by ordered logistic regression
• Clinical evidence of epilation The effect of the intervention on clinical evidence of epilation will be assessed at one and two years respectively with logistic regression. In addition, Cox regression will be used to analyse time-to-first evidence of epilation.
• Repeat surgery The effect of the intervention on the proportion with repeat surgery will be assessed at one and two years respectively with logistic regression. In addition, Cox regression will be used to analyse time-to-first evidence of first repeat surgery.
• Recurrence (lashes touching the eyeball, clinical evidence of epilation, repeat surgery) + history of epilation A secondary definition of recurrence will incorporate self-reported history of epilation. The primary analyses will be repeated with this definition of recurrence.
• Corneal opacification (CO) (categorical variable)

• Descriptive measures of the baseline severity of CO
• Description of the change in CO between baseline and one and two year follow up time-points.
• Ordered logistic regression analysis for factors associated with change in CO at one year and two years, respectively, including to identify factors associated with it, such as suture type and inflammation.
• Visual acuity (VA) (categorical variable) The effect of the intervention on the visual acuity (logMAR score) at 12 and 24 months will be analysed by linear regression.
• Surgical complications

Surgical complications other than recurrence, such as granuloma, incomplete lid closure, inflammation (secondary to either of the two sutures materials) and infection can occur. Descriptive statistics, with significance testing will be used to compare the rates of occurrence of all complications and individual complications at any time during the study period between the two arms of the study.

• Patient satisfaction

Patients are asked the following questions on the follow up questionnaires that indicate their satisfaction with the procedure:

• Do you think your vision is worse, the same or better than before your surgery?
• Do you have eye pain?
• If you have eye pain is it worse, the same or better than before the operation?
• Do your eyes water?
• Did you find the surgery painful?
• If the eyelashes have returned or will return would you seek further surgery?

The proportion answering 'yes' to each of these questions will be compared between the two arms of the trial.

Additionally the answers to questions (b) and (d) will be compared between baseline (pre-operative) and the answer given at each time point within each arm.

Inclusion criteria

• Major trichiasis: more than 5 lashes touching the eye

Exclusion criteria

• Previous eyelid surgery