A unilateral lateral rectus recession combined with a medial rectus resection in the same eye. Choice of eye at investigator discretion based on any interocular difference, position under anesthesia, fixation preference, or forced duction testing. Reason for choice of eye will be recorded.

NIH-required analysis. POM stratified by ethnicity:

A participant's intermittent exotropia (IXT) was considered to be a suboptimal surgical outcome if at any visit occurring 6 months or later, ANY of the following criteria are present by masked examiner testing:

1. Exotropia at distance OR near at any time during the exam (i.e., can be constant or intermittent; determined by a cover/uncover test) with a magnitude of ≥10Δ by SPCT, confirmed by a retest
2. Constant esotropia at distance OR near (determined by at least 3 cover/uncover tests-one must be before any dissociation) with a magnitude of ≥6Δ by SPCT, confirmed by a retest
3. Decrease in Randot Preschool near stereoacuity ≥2 octaves (≥0.6 log arcsec) from enrollment, or to nil, confirmed by a retest Participants who underwent reoperation (or treatment with botulinum toxin) without first meeting any of the above suboptimal surgical outcome criteria were also counted as suboptimal surgical outcomes in the primary analysis

Decrease in Preschool Randot near stereoacuity ≥2 octaves (≥0.6 log arcsec) from enrollment, or to nil, confirmed by a retest, by 3 years. Criteria was met before any reoperation, and regardless of whether suboptimal surgical outcome was met by another criteria.

Exotropia control at distance was assessed in all patients who completed the 3-year visit. Numeric values for exotropia control were assigned so that the following seven categories were created:

Not applicable (no exodeviation) (0) No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Mean exotropia control at distance was assessed in all patients who completed the 3-year visit. All 3-year visit data will be analyzed regardless of what treatment(s) a patient has received and regardless of whether the patient has undergone reoperation. Control at distance was analyzed as a continuous variable and compared between treatment groups using analysis of covariance (ANOVA) models that adjust for the corresponding baseline value (e.g. ANCOVA model of 3-year distance control will adjust for baseline distance control).

Numeric values for exotropia control were assigned so that the following categories were created:

Not applicable (no exodeviation) 0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement.

Numeric values for exotropia control were assigned so that the following categories were created:

Not applicable (no exodeviation) 0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Exotropia control at near was assessed in all patients who completed the 3-year visit.

Numeric values for exotropia control were assigned so that the following categories were created:

Not applicable (no exodeviation) 0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Mean exotropia control at near was assessed in all patients who completed the 3-year visit. All 3-year visit data will be analyzed regardless of what treatment(s) a patient has received and regardless of whether the patient has undergone reoperation. Control at near was analyzed as a continuous variable and compared between treatment groups using analysis of covariance (ANOVA) models that adjust for the corresponding baseline value (e.g. ANCOVA model of 3-year near control will adjust for baseline near control).

Numeric values for exotropia control were assigned so that the following categories were created:

Not applicable (no exodeviation) 0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement.

Numeric values for exotropia control were assigned so that the following categories were created:

Not applicable (no exodeviation) 0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. PACT was assessed in all patients who completed the 3-year visit.

∆ = prism diopters; eso = esodeviation; exo = exodeviation

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. Mean PACT was assessed in all patients who completed the 3-year visit. All 3-year visit data will be analyzed regardless of what treatment(s) a patient has received and regardless of whether the patient has undergone reoperation. PACT was analyzed as a continuous variable and compared between treatment groups using analysis of covariance (ANOVA) models that adjust for the corresponding baseline value (e.g. ANCOVA model of 3-year PACT at distance will adjust for baseline PACT at distance).

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement. If the 3-year PACT is an esodeviation, change in PACT from baseline is the reduction in the exodeviation plus the amount of the 3-year exodeviation.

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. PACT was assessed in all patients who completed the 3-year visit.

∆ = prism diopters; eso = esodeviation; exo = exodeviation

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. Mean PACT was assessed in all patients who completed the 3-year visit. All 3-year visit data will be analyzed regardless of what treatment(s) a patient has received and regardless of whether the patient has undergone reoperation. PACT was analyzed as a continuous variable and compared between treatment groups using analysis of covariance (ANOVA) models that adjust for the corresponding baseline value (e.g. ANCOVA model of 3-year PACT at near will adjust for baseline PACT at near).

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement. If the 3-year PACT is an esodeviation, change in PACT from baseline is the reduction in the exodeviation plus the amount of the 3-year exodeviation.

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

Stereoacuity Testing: stereoacuity was assessed in current refractive correction using the following:

Preschool Randot stereotest at near (performed at 40 cm): If stereoacuity is worse than 40 arcsec, it must be retested and the better of the 2 measurements will be used for eligibility.

Distance Randot stereotest (performed at 3 meters)

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Stereoacuity Testing: stereoacuity was assessed in current refractive correction using the following:

Preschool Randot stereotest at near (performed at 40 cm): If stereoacuity is worse than 40 arcsec, it must be retested and the better of the 2 measurements will be used for eligibility.

Distance Randot stereotest (performed at 3 meters)

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement.

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

Stereoacuity Testing: stereoacuity was assessed in current refractive correction using the following:

Preschool Randot stereotest at near (performed at 40 cm): If stereoacuity is worse than 40 arcsec, it must be retested and the better of the 2 measurements will be used for eligibility.

Distance Randot stereotest (performed at 3 meters)

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Stereoacuity Testing: stereoacuity was assessed in current refractive correction using the following:

Preschool Randot stereotest at near (performed at 40 cm): If stereoacuity is worse than 40 arcsec, it must be retested and the better of the 2 measurements will be used for eligibility.

Distance Randot stereotest (performed at 3 meters)

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement.

Health-related quality of life will be assessed using the Intermittent Exotropia Questionnaire (IXTQ). This questionnaire consists of 6 components:

1. Child questionnaire - consists of 12 items which assess how the child feels about his/her eye condition.

   • One version for children aged 5 to < 8 years has a 3-level response scale
   • The version for children aged 8 years and older has a 5-level response scale

2. Parent proxy questionnaire - consists of 12 items which assess how the parent feels the child's eye condition affects the child

3. Parental questionnaire - consists of 17 items which assess how the child's eye condition affects the parent. Has 3 sub-scales: surgical, functional, and psycho-social.

All scales ranged from 0 to 100; higher values indicated a better quality of life. Sub-scales were not combined, but rather were each evaluated individually on a scale of 0-100.

The cumulative proportion of re-operation by 3 years was compared between treatment groups using methods similar to the primary analysis (i.e. using Kaplan-Meier method). A treatment-group difference and a corresponding 95% confidence interval were also calculated. Reasons for re-operation included:

XT; XT and worsening stereo ; XT, worsening stereo and social concerns ; XT, diplopia, and headaches ; XT and squinting with one eye closed ; ET ; ET, worsening stereo, and diplopia; ET, worsening stereo and social concerns ; ET, worsening stereo, social concerns, and amblyopia ; Inferior oblique overaction

Complete or near-complete resolution was defined as meeting all of the following at the 3 year visit: 1) exodeviation <10 Δ (tropia or phoria) by both SPCT and PACT at distance and near and ≥10 Δ reduction in PACT magnitude from the largest of the distance and near angles at enrollment, 2) esotropia <6 Δ at distance and near by SPCT, 3) no decrease in Randot Preschool stereoacuity of ≥2 octaves from the enrollment stereoacuity or to nil, 4) no reoperation or treatment with botulinum toxin, and 5) no non-surgical treatment for a recurrent or residual exodeviation.

Suboptimal surgical outcome at the 3-year visit was defined as meeting any of the three suboptimal surgical outcome criteria at the 3-year visit (regardless of whether the criterion had been met at an earlier visit), or undergoing reoperation at any time.

The three criteria for suboptimal surgical outcome were:

1. Exotropia at distance OR near at any time during the exam (i.e., can be constant or intermittent; determined by a cover/uncover test) with a magnitude of ≥10Δ by SPCT, confirmed by a retest
2. Constant esotropia at distance OR near (determined by at least 3 cover/uncover tests-one must be before any dissociation) with a magnitude of ≥6Δ by SPCT, confirmed by a retest
3. Decrease in Randot Preschool near stereoacuity ≥2 octaves (≥0.6 log arcsec) from enrollment, or to nil, confirmed by a retest

Total of all reporting groups

Stereoacuity scores (seconds of arc, or arcsec) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation.

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation.

Exotropia control was measured in points ranging from 0 to 5, with each point corresponding to the following:

0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Exotropia control was measured in points ranging from 0 to 5, with each point corresponding to the following:

0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Unilateral lateral rectus recession w/ medial rectus resection surgery

Unilateral lateral rectus recession with medial rectus resection (R&R): A unilateral lateral rectus recession combined with a medial rectus resection in the same eye. Choice of eye at investigator discretion based on any interocular difference, position under anesthesia, fixation preference, or forced duction testing. Reason for choice of eye will be recorded.

Exotropia ≥10Δ by simultaneous prism and cover test (SPCT) at distance or near, confirmed by a retest, by 3 years. Criteria was met before any reoperation, and regardless of whether suboptimal surgical outcome was met by another criteria.

Constant esotropia ≥6Δ by simultaneous prism and cover test (SPCT) at distance or near, confirmed by a retest, by 3 years. Criteria was met before any reoperation, and regardless of whether suboptimal surgical outcome was met by another criteria.

Decrease in Preschool Randot near stereoacuity ≥2 octaves (≥0.6 log arcsec) from enrollment, or to nil, confirmed by a retest, by 3 years. Criteria was met before any reoperation, and regardless of whether suboptimal surgical outcome was met by another criteria.

Exotropia control at distance was assessed in all patients who completed the 3-year visit. Numeric values for exotropia control were assigned so that the following seven categories were created:

Not applicable (no exodeviation) (0) No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Includes only patients who completed the 3-year visit.

Mean exotropia control at distance was assessed in all patients who completed the 3-year visit. All 3-year visit data will be analyzed regardless of what treatment(s) a patient has received and regardless of whether the patient has undergone reoperation. Control at distance was analyzed as a continuous variable and compared between treatment groups using analysis of covariance (ANOVA) models that adjust for the corresponding baseline value (e.g. ANCOVA model of 3-year distance control will adjust for baseline distance control).

Numeric values for exotropia control were assigned so that the following categories were created:

Not applicable (no exodeviation) 0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Range of the mean was 0 to 5 for both the BLR group and for the RR group.

Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement.

Numeric values for exotropia control were assigned so that the following categories were created:

Not applicable (no exodeviation) 0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Range of the mean was -3 to 5 for the BLR group and -2 to 5 for the RR group.

Exotropia control at near was assessed in all patients who completed the 3-year visit.

Numeric values for exotropia control were assigned so that the following categories were created:

Not applicable (no exodeviation) 0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Includes only patients who completed the 3-year visit.

Mean exotropia control at near was assessed in all patients who completed the 3-year visit. All 3-year visit data will be analyzed regardless of what treatment(s) a patient has received and regardless of whether the patient has undergone reoperation. Control at near was analyzed as a continuous variable and compared between treatment groups using analysis of covariance (ANOVA) models that adjust for the corresponding baseline value (e.g. ANCOVA model of 3-year near control will adjust for baseline near control).

Numeric values for exotropia control were assigned so that the following categories were created:

Not applicable (no exodeviation) 0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Range of the mean was 0 to 5 for the BLR group and 0 to 4 for the RR group.

Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement.

Numeric values for exotropia control were assigned so that the following categories were created:

Not applicable (no exodeviation) 0: No exotropia unless dissociated, recovers <1 secs (phoria)

1. No exotropia unless dissociated, recovers 1-5 secs
2. No exotropia unless dissociated, recovers >5 secs
3. Exotropia <50% of 30-second observation
4. Exotropia >50% of 30-second observation
5. Constant exotropia

Lower scores indicate better control.

Range of the mean was -3 to 4 for both the BLR and RR groups.

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. PACT was assessed in all patients who completed the 3-year visit.

∆ = prism diopters; eso = esodeviation; exo = exodeviation

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. Mean PACT was assessed in all patients who completed the 3-year visit. All 3-year visit data will be analyzed regardless of what treatment(s) a patient has received and regardless of whether the patient has undergone reoperation. PACT was analyzed as a continuous variable and compared between treatment groups using analysis of covariance (ANOVA) models that adjust for the corresponding baseline value (e.g. ANCOVA model of 3-year PACT at distance will adjust for baseline PACT at distance).

Range of the mean was -14 to 35 for the BLR group and -6 to 30 for the RR group.

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement. If the 3-year PACT is an esodeviation, change in PACT from baseline is the reduction in the exodeviation plus the amount of the 3-year exodeviation.

Range of the mean was -10 to 44 for the BLR group and -5 to 36 for the RR group.

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. PACT was assessed in all patients who completed the 3-year visit.

∆ = prism diopters; eso = esodeviation; exo = exodeviation

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. Mean PACT was assessed in all patients who completed the 3-year visit. All 3-year visit data will be analyzed regardless of what treatment(s) a patient has received and regardless of whether the patient has undergone reoperation. PACT was analyzed as a continuous variable and compared between treatment groups using analysis of covariance (ANOVA) models that adjust for the corresponding baseline value (e.g. ANCOVA model of 3-year PACT at near will adjust for baseline PACT at near).

Range of the mean was -14 to 40 for the BLR group and -6 to 30 for the RR group.

The prism and alternate cover test (PACT) is used to measure the angle of strabismus, or deviation, in prism diopters. This is measured separately at distance and at near. Smaller numbers are better because they indicate a smaller angle of deviation. Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement. If the 3-year PACT is an esodeviation, change in PACT from baseline is the reduction in the exodeviation plus the amount of the 3-year exodeviation.

Range of the mean was -14 to 44 for the BLR group and -12 to 41 for the RR group.

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

Stereoacuity Testing: stereoacuity was assessed in current refractive correction using the following:

Preschool Randot stereotest at near (performed at 40 cm): If stereoacuity is worse than 40 arcsec, it must be retested and the better of the 2 measurements will be used for eligibility.

Distance Randot stereotest (performed at 3 meters)

Includes only patients who completed the 3-year visit.

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Stereoacuity Testing: stereoacuity was assessed in current refractive correction using the following:

Preschool Randot stereotest at near (performed at 40 cm): If stereoacuity is worse than 40 arcsec, it must be retested and the better of the 2 measurements will be used for eligibility.

Distance Randot stereotest (performed at 3 meters)

The range of the mean was -1.6 to 3.2 for both the BLR and RR groups.

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement.

Range of the mean was -1.3 to 1.0 for the BLR group and -1.4 to 1.0 for the RR group.

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

Stereoacuity Testing: stereoacuity was assessed in current refractive correction using the following:

Preschool Randot stereotest at near (performed at 40 cm): If stereoacuity is worse than 40 arcsec, it must be retested and the better of the 2 measurements will be used for eligibility.

Distance Randot stereotest (performed at 3 meters)

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Stereoacuity Testing: stereoacuity was assessed in current refractive correction using the following:

Preschool Randot stereotest at near (performed at 40 cm): If stereoacuity is worse than 40 arcsec, it must be retested and the better of the 2 measurements will be used for eligibility.

Distance Randot stereotest (performed at 3 meters)

Range of the mean was 1.8 to 2.9 for both the BLR and RR groups.

Stereoacuity scores (seconds of arc) were calculated based on the Randot Preschool stereoacuity test (scores: 800, 400, 200, 100, 60 and 40). Seconds of arc refers to the visual angle that is being measured in order to determine depth perception. Lower scores indicate better stereoacuity.

A logarithm base 10 transformation was used to convert stereoacuity scores to the log scale to calculate descriptive statistics (reported as seconds of arc, or arcsec).

Change is defined as the baseline value minus the 3-year value, therefore positive change = improvement.

Range of the mean was -0.9 to 1.1 for the BLR group and -0.8 to 1.1 for the RR group.

Health-related quality of life will be assessed using the Intermittent Exotropia Questionnaire (IXTQ). This questionnaire consists of 6 components:

1. Child questionnaire - consists of 12 items which assess how the child feels about his/her eye condition.

   • One version for children aged 5 to < 8 years has a 3-level response scale
   • The version for children aged 8 years and older has a 5-level response scale

2. Parent proxy questionnaire - consists of 12 items which assess how the parent feels the child's eye condition affects the child

3. Parental questionnaire - consists of 17 items which assess how the child's eye condition affects the parent. Has 3 sub-scales: surgical, functional, and psycho-social.

All scales ranged from 0 to 100; higher values indicated a better quality of life. Sub-scales were not combined, but rather were each evaluated individually on a scale of 0-100.

The total number of children/patients was split between two categories: younger and older, which is why the number analyzed is different between rows. 1 patient in the BLR group and 2 patients in the RR group were missing quality of life questionnaire data. Parent information was not missing which is why the total number is equal to number analyzed

The cumulative proportion of re-operation by 3 years was compared between treatment groups using methods similar to the primary analysis (i.e. using Kaplan-Meier method). A treatment-group difference and a corresponding 95% confidence interval were also calculated. Reasons for re-operation included:

XT; XT and worsening stereo ; XT, worsening stereo and social concerns ; XT, diplopia, and headaches ; XT and squinting with one eye closed ; ET ; ET, worsening stereo, and diplopia; ET, worsening stereo and social concerns ; ET, worsening stereo, social concerns, and amblyopia ; Inferior oblique overaction

Complete or near-complete resolution was defined as meeting all of the following at the 3 year visit: 1) exodeviation <10 Δ (tropia or phoria) by both SPCT and PACT at distance and near and ≥10 Δ reduction in PACT magnitude from the largest of the distance and near angles at enrollment, 2) esotropia <6 Δ at distance and near by SPCT, 3) no decrease in Randot Preschool stereoacuity of ≥2 octaves from the enrollment stereoacuity or to nil, 4) no reoperation or treatment with botulinum toxin, and 5) no non-surgical treatment for a recurrent or residual exodeviation.

Includes only those patients that completed the 3-year visit.

Suboptimal surgical outcome at the 3-year visit was defined as meeting any of the three suboptimal surgical outcome criteria at the 3-year visit (regardless of whether the criterion had been met at an earlier visit), or undergoing reoperation at any time.

The three criteria for suboptimal surgical outcome were:

1. Exotropia at distance OR near at any time during the exam (i.e., can be constant or intermittent; determined by a cover/uncover test) with a magnitude of ≥10Δ by SPCT, confirmed by a retest
2. Constant esotropia at distance OR near (determined by at least 3 cover/uncover tests-one must be before any dissociation) with a magnitude of ≥6Δ by SPCT, confirmed by a retest
3. Decrease in Randot Preschool near stereoacuity ≥2 octaves (≥0.6 log arcsec) from enrollment, or to nil, confirmed by a retest

Includes only those who completed the 3 year visit.

NIH-required analysis. Primary OM (stratified by sex):

A participant's intermittent exotropia (IXT) was considered to be a suboptimal surgical outcome if at any visit occurring 6 months or later, ANY of the following criteria are present by masked examiner testing:

1. Exotropia at distance OR near at any time during the exam (i.e., can be constant or intermittent; determined by a cover/uncover test) with a magnitude of ≥10Δ by SPCT, confirmed by a retest
2. Constant esotropia at distance OR near (determined by at least 3 cover/uncover tests-one must be before any dissociation) with a magnitude of ≥6Δ by SPCT, confirmed by a retest
3. Decrease in Randot Preschool near stereoacuity ≥2 octaves (≥0.6 log arcsec) from enrollment, or to nil, confirmed by a retest Participants who underwent reoperation (or treatment with botulinum toxin) without first meeting any of the above suboptimal surgical outcome criteria were also counted as suboptimal surgical outcomes in the primary analysis

Time-to-event analysis included all randomized participants regardless of 3-year visit completion.