Skip to main content
. Author manuscript; available in PMC: 2014 Dec 17.
Published in final edited form as: Cochrane Database Syst Rev. 2013 Dec 3;12:CD005431. doi: 10.1002/14651858.CD005431.pub3

Table 9.

Characteristics of included studies [ordered by study ID]

Bedrossian 1974
Methods Study design: Quasi-randomized controlled series.
Exclusions after allocation: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were assigned
Sample size calculations: Not reported.
Participants Country: USA.
Dates: Not reported.
Number allocated: 58 consecutive patients alternately assigned to treatment group after classification based on the size of initial hyphema
Age: Not reported.
Sex: Not reported.
Race: Not reported.
Sickle cell disease: Not reported.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Nontotal traumatic hyphema.
Interventions Cycloplegics (n = 28): 1% atropine ointment.
Miotics (n = 30): 2% pilocarpine ointment (or eserine ointment)
Treatment for both groups included:

  1. Topical anesthetic if needed;

  2. Bed rest;

  3. Head of bed elevated 30–90°;

  4. Binocular patching or pinhole glasses;

  5. No reading or watching television;

  6. Metal shield over injured eye;

  7. Soft, nonchew diet;

  8. Laxatives;

  9. Room with other individuals; and

  10. Sedation.
Outcomes Primary outcome: Time to resolution of primary hemorrhage.
Secondary outcomes:

  1. Risk of secondary hemorrhage; and

  2. Risk of iridodialysis.


Follow-up: days 1 to 7.
Notes Funding source not reported.
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) High risk Allocation was not randomized; alternately assigned patients to treatment groups based on the blood level in the anterior chamber
Allocation concealment (selection bias) High risk Allocation was assigned on an alternate basis.
Blinding (performance bias and detection bias)
Participants		 High risk Masking was not reported.
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 High risk Masking was not reported.
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk All participants were analyzed in the group to which they were assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk All participants were analyzed in the group to which they were assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Christianson 1979
Methods Study design: Randomized, double-masked, placebo-controlled clinical trial
Exclusions after randomization: None reported.
Losses to follow-up: None reported.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Not reported.
Participants Country: USA.
Dates: Not reported.
Number randomized: 45.
Age: Not reported.
Sex: Not reported.
Race: Not reported.
Sickle cell disease: Not reported.
Inclusion criteria: Traumatic hyphema.
Exclusion criteria: Not reported.
Interventions Treatment (n = 22): Oral aminocaproic acid, loading dose 75 mg/kg, followed by 60 mg/kg every 4 hours; length of treatment not reported
Control (n = 23): Placebo, presumably every 4 hours.
Outcomes Primary outcome: Risk of secondary hemorrhage, details not reported
Secondary outcomes: Time to resolution of primary hyphema, details not reported
Notes Abstract of unpublished study.
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Method of randomization not reported.
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 Low risk Authors used a placebo control and stated that the study was double-masked
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Authors used a placebo control and stated that the study was double-masked
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk Unclear if number randomized equaled the number reported and analyzed in the abstract, but no exclusions or losses to follow-up were reported
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk Unclear if number randomized equaled the number reported and analyzed in the abstract, but no exclusions or losses to follow-up were reported
Selective reporting (reporting bias) Unclear risk Few study details available in the abstract and no full version was published
Other bias Unclear risk Few study details available in the abstract and no full version was published
Crouch 1976
Methods Study design: Randomized, double-masked, placebo-controlled clinical trial
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Not reported.
Participants Country: USA.
Dates: September 1972 to October 1974.
Number randomized: 59.
Age: 83% ages 6–30 years.
Sex: 83% male.
Race: 65% black people, 35% white people.
Sickle cell disease: 8/59 (14%) had sickle cell trait.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. Penetrating injury;

  2. Total hyphema;

  3. History of a bleeding disorder; and

  4. Pregnancy.
Interventions Treatment (n = 32): Oral aminocaproic acid 100 mg/kg every 4 hours for 5 days
Control (n = 27): Placebo (200 mL of aromatic elixir (5% glucose, water, and ethanol) in 1000 mL sterile water) every 4 hours for 5 days
Treatment for both groups included:

  1. Moderate ambulation;

  2. No reading;

  3. Head of bed elevated to 45°;

  4. Patching of affected eye;

  5. No mydriatics, miotics, corticosteroids, or other topical medication; and

  6. No salicylates.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed by daily slit lamp exam, and documented by 3 observers
Secondary outcomes:

  1. Time to resolution of primary hemorrhage;

  2. Time to secondary hemorrhage;

  3. Final VA, with follow-up ranging between 6 months and 2.5 years;

  4. IOP assessed daily by applanation tonometry; and

  5. Risk of complications and adverse events.


Follow-up: 1 week, 1, 2, 3, 6, 12, 18, and 24 months.
Notes Funded by the National Eye Institute, National Institutes of Health
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Low risk Participants assigned to treatment groups using computerized randomization
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 Low risk Authors used a placebo control and stated that the study was double-masked
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Authors used a placebo control and stated that the study was double-masked
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk There were no exclusions and losses to follow-up. All participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk There were no exclusions and losses to follow-up. All participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Crouch 1997
Methods Study design: Randomized, double-masked clinical trial.
Exclusions after randomization: 1 individual assigned to oral aminocaproic acid and topical placebo excluded based on side effect of drug (vomiting)
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Sample size was determined to be 25–30 participants in each of the 3 groups based on alpha of 0.05 and power of 80%
Additional comments: The investigators also studied a control group that did not receive either topical or systemic aminocaproic acid and had refused randomization. We did not include these patients in our analyses
Participants Country: USA.
Dates: March 1990 to May 1996.
Number randomized: 64: 29 to oral aminocaproic acid plus topical placebo, 35 to oral placebo plus topical aminocaproic acid. Additional 54 participants included as control group
Age: 72% younger than 21 years.
Sex: 67% male.
Race: 50% black people, 49% white people, and 1% (1 participant) was Asian
Sickle cell disease: 2/35 (6%) of participants assigned to topical aminocaproic acid, and 2/29 (7%) of participants assigned to oral aminocaproic acid had sickle cell trait
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. Penetrating ocular injury;

  2. History of anticoagulant or antiplatelet agent within 7 days of ocular trauma;

  3. Oral or topical corticosteroid use within 48 hours of study;

  4. History of a coagulopathy;

  5. History of renal or hepatic insufficiency;

  6. Previous intraocular surgery;

  7. History of sensitivity to any component of topical aminocaproic acid;

  8. Pregnancy; and

  9. Participation in any investigational drug trial within last 4 weeks.
Interventions Treatment: 0.2 mL of 30% aminocaproic acid in 2% carboxymethylene gel applied to inferior fornix every 6 hours plus oral placebo solution every 4 hours, for 5 days
Control: Oral aminocaproic acid 50 mg/kg (up to 30 g/day) plus placebo gel every 4 hours, for 5 days
Treatment for both groups included:

  1. Moderate ambulation;

  2. Head of bed elevated to 30°;

  3. Shield on affected eye;

  4. No aspirin, corticosteroids, nonsteroidal anti-inflammatory, or antiplatelet agents; and

  5. Topical timolol maleate, apraclonidine hydrochloride, dipivefrin hydrochloride, or oral acetazolamide if IOP > 22 mmHg.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed by daily slit lamp exam, and documented by a sketch each day
Secondary outcomes:

  1. VA, measured daily and at the end of the 5 days (final VA);

  2. Cell and flare, assessed daily for 5 days;

  3. Corneal blood staining and toxicity, assessed daily by slit lamp exam for 5 days;

  4. IOP assessed daily for 5 days by applanation tonometry; and

  5. Risk of complications and adverse events.
Notes Funded in part by the Lions Medical Eye Bank and Research Center of Eastern Virginia
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Low risk Participants assigned to treatment groups using computerized randomization
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 Low risk Authors used a placebo control and stated that the study was double-masked. Placebo pills were given to the topical group and placebo gel administered to the systemic group to make both regimens similar
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Authors used a placebo control and stated that the study was double-masked. “Data were compiled by observers who did not know what patients were in the treated and untreated control groups.”
Incomplete outcome data (attrition bias)
Primary outcome		 Unclear risk 1 patient was excluded: 1 individual assigned to oral aminocaproic acid and topical placebo excluded based on side effect of drug (vomiting). The remaining participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Unclear risk 1 patient was excluded: 1 individual assigned to oral aminocaproic acid and topical placebo excluded based on side effect of drug (vomiting). The remaining participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Edwards 1973
Methods Study design: Quasi-randomized controlled series.
Exclusions after allocation: Patients over 20 years old were excluded from the study because of the small number enrolled
Losses to follow-up: None.
Intention-to-treat: Participants aged 20 years and younger were analyzed in the group to which they were assigned
Sample size calculations: Not reported.
Participants Country: USA.
Dates: 1969–1971.
Number allocated: 64 consecutive patients alternately assigned to treatment group
Age: Mean 10 years (up to 20 years).
Sex: 61 (95%) men and 3 (5%) women.
Race: Not reported.
Sickle cell disease: Not reported.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Traumatic hyphema.
Exclusion criteria: Patients over 20 years of age.
Interventions Treatment: Monocular patching (n = 35)
Control: Binocular patching (n = 29)
Treatment for both groups included:

  1. Standard regimen (including position in bed, sedation, and diet);

  2. Acetazolamide for severe secondary glaucoma; and

  3. No topical medications.
Outcomes Primary and secondary outcomes not specified.
Measured outcomes:

  1. Risk of secondary hemorrhage;

  2. Duration of rebleeding;

  3. Complication rates; and

  4. Final VA.


Follow-up: days 1–7.
Notes Funded by Research to Prevent Blindness Inc., Public Health Service Training Grant, and the National Institutes of Health
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) High risk Allocation was not randomized; an independent study director assigned patients to treatment groups on an alternate basis by turning a card. Occasionally the card was not turned each time, which led to an uneven number of patients in each group
Allocation concealment (selection bias) High risk Allocation was assigned on an alternate basis.
Blinding (performance bias and detection bias)
Participants		 High risk Masking of patients was not possible with the interventions being studied
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Unclear risk Authors reported study to be double-masked, although this statement was not clear. The study investigators seldom participated in patient care to allow other examiners with less experience in monocular patching to collect data in hopes of minimizing observation bias
Incomplete outcome data (attrition bias)
Primary outcome		 Unclear risk Patients over 20 years of age were excluded after allocation to treatment group
Incomplete outcome data (attrition bias)
Secondary outcomes		 Unclear risk Patients over 20 years of age were excluded after allocation to treatment group
Selective reporting (reporting bias) Low risk Reported results for all outcomes.
Other bias Low risk No other sources of potential bias were identified.
Farber 1991
Methods Study design: Randomized, double-masked clinical trial.
Exclusions after randomization: 6 participants in the aminocaproic acid group were excluded; 4 were administered prednisone instead of aminocaproic acid (treatment crossover), 1 participant had an unrelated seizure, and 1 developed thrombocytopenia. 1 participant in the prednisone group was administered aminocaproic acid instead of prednisone (treatment cross-over) and was excluded
Losses to follow-up: 2 participants in the aminocaproic acid group and 1 participant in the prednisone group withdrew from the study
Intention-to-treat: The participants lost to follow-up or excluded were not included in the analyses and the intention-to-treat principle was not followed in the analyses
Sample size calculations: Not reported.
Additional comments: The authors noted that there were no secondary hemorrhages in the individuals who had been excluded or withdrew from the study
Participants Country: USA.
Dates: July 1985 to March 1990.
Number randomized: 122: 64 to aminocaproic acid, 58 to prednisone
Age: Mean age in aminocaproic acid group: 23.8 ± 13.8 years (range 4–64 years); in prednisone group: 23.3 ± 13.4 years (range 1.5–62 years)
Sex: 79% male.
Race: 53% black people, 22% white people, 22% Hispanic people, and 3% of other ethnic or racial group. Study groups were not balanced by race: 57% of black people and 20% of white people in aminocaproic acid group vs. 48% of black people and 25% of white people in prednisone group
Sickle cell disease: None; excluded
Inclusion criteria: Traumatic hyphema
Exclusion criteria:

  1. Penetrating ocular injury;

  2. Need for immediate surgery;

  3. Sickle cell trait or disease;

  4. History of intravascular coagulopathy;

  5. History of gastric ulcer;

  6. History of diabetes mellitus;

  7. Pregnancy;

  8. Intoxication;

  9. Presence of detectable blood in stool.
Interventions Treatment: Oral aminocaproic acid 50 mg/kg (up to 30 g/day) every 4 hours plus 2 doses placebo, for 5 days
Control: Oral prednisone 40 mg/day in 2 doses plus 6 doses placebo; children and adults weighing less than 60 kg were given 0.6 mg/kg/day prednisone, for 5 days
Treatment for both groups included:

  1. Moderate ambulation;

  2. No reading;

  3. Head of bed elevated to 30°;

  4. Patch and shield on affected eye;

  5. Topical 1% atropine sulfate 4 times/day;

  6. Oral acetaminophen (paracetamol) up to 650 mg/day, no aspirin;

  7. Topical timolol maleate 0.25% or 0.50% with or without oral acetazolamide if IOP > 25 mmHg; and

  8. Prochlorperazine edisylate (5 or 10 mg) if vomiting or nausea.
Outcomes Primary outcome: Risk of secondary hemorrhage, recorded daily by slit lamp exam, documented by measuring height in mm and defined as a definite increase in level of presence of ’fresh’ blood visible over darker clotted blood
Secondary outcomes:

  1. VA, initial and final (5 days);

  2. IOP measured daily using applanation tonometry; and

  3. Risk of complications and adverse events.
Notes Funded by the National Eye Institute of the National Institutes of Health, Bethesda, MD, and Research to Prevent Blindness
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation not reported.
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 Low risk Authors used a double-dummy placebo design and stated that the study was double-masked
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Authors used a double-dummy placebo design and stated that the study was double-masked. “All of the treating physicians and nurses were masked to the identity of the treatment.”
Incomplete outcome data (attrition bias)
Primary outcome		 Unclear risk The participants lost to follow-up or excluded were not included in the analyses and the intention-to-treat principle was not followed in the analyses
Incomplete outcome data (attrition bias)
Secondary outcomes		 Unclear risk The participants lost to follow-up or excluded were not included in the analyses and the intention-to-treat principle was not followed in the analyses
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Karkhaneh 2003
Methods Study design: Randomized, double-masked clinical trial.
Exclusions after randomization: None.
Losses to follow-up: 23; 4 to homatropine drops plus topical aminocaproic acid gel, 5 to homatropine drops plus topical placebo gel, 14 to homatropine drops only
Intention-to-treat: The participants lost to follow-up were not included in the analyses and the intention-to-treat principle was not followed in the analyses
Sample size calculations: Not reported
Participants Country: Iran
Dates: 1998–1999
Number randomized: 155: 45 to homatropine drops plus topical aminocaproic acid gel, 44 to homatropine drops plus placebo gel, 66 to homatropine drops only
Age: 4–30 years.
Sex: 87% (not including those lost to follow-up) male.
Race: Not reported.
Sickle cell disease: Not reported.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Nonpentrating traumatic hyphema and emergency room outpatient of Farabi Eye Hospital
Exclusion criteria:

  1. Penetrating ocular injury;

  2. Total hyphema;

  3. Microscopic hyphema;

  4. More than 24 hours since trauma;

  5. History of bleeding disorder;

  6. Previous ocular surgery in affected eye;

  7. Recent aspirin or anticoagulant ingestion;

  8. Pregnancy; and

  9. Trauma to affected eye during follow-up.
Interventions Treatment 1: 2 drops of 25% aminocaproic acid in 2% carboxymethylene gel applied to inferior fornix of affected eye every 6 hours plus homatropine eyedrops 3 times/day, for 5 days
Control 1: 2 drops 2% carboxymethylene (placebo) gel applied to inferior fornix of affected eye every 6 hours plus homatropine eyedrops 3 times/day, for 5 days
Control 2: Homatropine eyedrops 3 times/day, for 5 days.
Treatment for all groups included:

  1. No reading;

  2. Head of bed elevated to 30°;

  3. Shield on affected eye;

  4. Oral acetaminophen (paracetamol);

  5. No aspirin.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily by slit lamp exam for 7 days, and then at day 14. Method for documentation and definition not reported
Secondary outcomes: All measured daily for 7 days and at day 14:

  1. Time to resolution of primary hemorrhage;

  2. Time to secondary hemorrhage;

  3. VA; final VA at day 14;

  4. IOP measured using applanation tonometry;

  5. Corneal blood staining;

  6. Drug toxicity; and

  7. Risk of complications and adverse events.
Notes Conducted with support from Sina Darou (an ophthalmic pharmaceutical company in Iran), who provided the aminocaproic acid preparation
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation was not reported.
Allocation concealment (selection bias) Low risk Allocation was concealed from investigators by use of coded bottles
Blinding (performance bias and detection bias)
Participants		 Unclear risk Authors used coded bottles to mask participants for the topical medication, but the group assigned to homatropine drops and no topical medication was not masked
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Authors used coded bottles to mask health-care providers and outcomes assessors. “The ophthalmologist who examined the patients did not know if they were treated or not.”
Incomplete outcome data (attrition bias)
Primary outcome		 Unclear risk The participants lost to follow-up were not included in the analyses and the intention-to-treat principle was not followed in the analyses. 23 participants lost to follow-up: 4 to homatropine drops plus topical aminocaproic acid gel, 5 to homatropine drops plus topical placebo gel, 14 to homatropine drops only
Incomplete outcome data (attrition bias)
Secondary outcomes		 Unclear risk The participants lost to follow-up were not included in the analyses and the intention-to-treat principle was not followed in the analyses. 23 participants lost to follow-up: 4 to homatropine drops plus topical aminocaproic acid gel, 5 to homatropine drops plus topical placebo gel, 14 to homatropine drops only
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Unclear risk Conducted with support from Sina Darou (an ophthalmic pharmaceutical company in Iran), who provided the aminocaproic acid preparation
Kraft 1987
Methods Study design: Randomized, double-masked clinical trial.
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Not reported.
Participants Country: Canada
Dates: May 1978 to December 1984
Number randomized: 49: 24 to oral aminocaproic acid; 25 to placebo
Age: 3–18 years. Mean age: aminocaproic acid group 10.6 years, placebo group 11.2 years
Sex: 73% male.
Race: 3 black participants in the aminocaproic acid group; 1 in the placebo group. The ethnicity or race of the other participants was not reported
Sickle cell disease: None; excluded.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Children with nonpenetrating traumatic hyphema treated at the Hospital for Sick Children in Toronto, Canada
Exclusion criteria:

  1. Penetrating ocular injury;

  2. More than 24 hours since trauma;

  3. Requirement for immediate surgical intervention;

  4. Positive sickle cell test or abnormal hematologic parameter;

  5. History of bleeding disorder;

  6. Ingestion of aspirin-containing medication within 7 days of admission; and

  7. Pregnancy.
Interventions Treatment: Oral aminocaproic acid 100 mg/kg every 4 hours, for 5 days
Control: Placebo every 4 hours, for 5 days.
Treatment for both groups included:

  1. Bed rest with bathroom privileges;

  2. Head of bed elevated 15°;

  3. Patch on affected eye;

  4. No topical eye medications except antibiotic ointment for corneal abrasions;

  5. Oral acetaminophen (paracetamol) (10–20 mg/kg every 4 hours, up to 650 mg/dose);

  6. No aspirin-containing medications;

  7. Up to 0.5 mg/kg per day diazepam for sedation if needed;

  8. Topical timolol maleate 0.5% if IOP > 25 mmHg;

  9. Dimenhydrinate (Gravol) 6.25–12.5 mg every 6 hours if vomiting or nausea.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily by slit lamp exam; documented by 2 observers and defined as definite increase in amount of blood compared with amount at admission or fresh red blood over darker clotted blood
Secondary outcomes: Outcomes measured daily during hospitalization (up to 5 days), then at 6 weeks, and 3, 6, 12, and 18 months after discharge

  1. Time to resolution of primary hemorrhage;

  2. VA;

  3. IOP assessed using applanation tonometry; and

  4. Risk of complications and adverse events.
Notes
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Low risk Participants assigned to treatment groups using computerized randomization
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 Low risk Authors used a placebo control and stated that the study was double-masked
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Authors used a placebo control and stated that the study was double-masked
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk There was no loss to follow-up and all participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk There was no loss to follow-up and all participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Kutner 1987
Methods Study design: Randomized, double-masked clinical trial
Exclusions after randomization: 1 participant was excluded from the aminocaproic acid group due to systemic hypotension attributable to the study drug
Losses to follow-up: None.
Intention-to-treat: The participant excluded from the study was not included in the analyses and the intention-to-treat principle was not followed in the analyses
Sample size calculations: Not reported.
Participants Country: USA.
Dates: November 1983 to January 1986.
Number randomized: 34: 21 to aminocaproic acid, 13 placebo.
Age: Mean age: aminocaproic acid 18.9 ± 7.7 years, placebo 22.8 ± 7.6 years
Sex: 88% male.
Race: 85% white people.
Sickle cell disease: None; excluded.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Nonpenetrating traumatic hyphema.
Exclusion criteria:

  1. Penetrating ocular injury;

  2. More than 48 hours since trauma;

  3. Age less than 7 years;

  4. Sickle cell anemia;

  5. History of intravascular coagulopathy, blood dyscrasia, or renal disease;

  6. History of ocular disease that could increase the susceptibility to intraocular hemorrhage;

  7. Recent anticoagulant, aspirin, or oral contraceptive use; and

  8. Pregnancy.
Interventions Treatment: Oral aminocaproic acid 100 mg/kg every 4 hours (up to 5 g/dose and 30 g/day), for 5 days
Control: Placebo every 4 hours, for 5 days.
Treatment for both groups included:

  1. Quiet activities;

  2. No reading;

  3. No patch or shield;

  4. No ocular medications;

  5. Oral acetaminophen 9paracetamol) (10–20 mg/kg every 4 hours, up to 650 mg/dose);

  6. No aspirin or alcohol;

  7. Diazepam 5 mg every 6 hours for sedation if needed;

  8. Topical timolol maleate 0.5% with IOP > 35 mmHg; and

  9. Prochloroperazine 5–10 mg if vomiting or nausea.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily by slit lamp exam, for 6 days and 1 week after discharge. Defined as a definite increase in the amount of blood in the anterior chamber compared with that noted on the previous day’s exam
Secondary outcomes:

  1. Time to resolution of primary hemorrhage;

  2. VA, measured daily for 6 days and 1 week after discharge;

  3. IOP measured daily using applanation tonometry for 6 days and 1 week after discharge; and

  4. Risk of complications and adverse events.
Notes
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Low risk Participants assigned to treatment groups using computerized randomization
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 Low risk Authors used a placebo control and stated that the study was double-masked
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Authors used a placebo control and stated that the study was double-masked. Assignment codes maintained by a central data evaluator who had no clinical contact with any patient. “Physicians caring for study patients did not have access to the cumulative data until the code was broken.”
Incomplete outcome data (attrition bias)
Primary outcome		 Unclear risk One participant was excluded from the aminocaproic acid group due to systemic hypotension attributable to the study drug. It was reported that this patient did not rebleed
Incomplete outcome data (attrition bias)
Secondary outcomes		 Unclear risk One participant was excluded from the aminocaproic acid group due to systemic hypotension attributable to the study drug. Data for this patient was analyzed until time of study withdrawal
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Liu 2002
Methods Study design: Randomized clinical trial.
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Not reported.
Participants Country: China
Dates: December 1997 to December 2000
Number randomized: 92: 60 to aminomethylbenzoic acid, 32 to control
Age: Mean age: aminomethylbenzoic acid 32.7 ± 11.25 years, control 33.4 ± 10.75 years
Sex: 75% male.
Race: Not reported.
Sickle cell disease: Not reported.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. More than 48 hours since trauma;

  2. Use of anticoagulants;

  3. History of risk of clot formation;

  4. History of diabetes.
Interventions Treatment: Oral aminomethylbenzoic acid 0.5 g plus oral vitamin B1 20 mg 3 times/day, for 6 days. For children, the dosage of aminomethylbenzoic acid was modified to “follow age-recommended dose”; the vitamin B1 dosage remained the same.
Control: Oral vitamin B1 20 mg 3 times/day, for 6 days.
Treatment for both groups included 0.3% ofloxacin eyedrops 4 times/day, for 6 days
Outcomes Primary outcome: Risk of secondary hemorrhage, details not reported
Secondary outcomes: Risk of complications and adverse events
Notes Poor description of study methods in publication.
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation not reported.
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 Unclear risk The authors do not state whether masking was used.
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Unclear risk The authors do not state whether masking was used.
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk No exclusions or loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk No exclusions or loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Unclear risk Study outcomes of interest not clearly stated.
Other bias Low risk No other sources of potential bias were identified.
Marcus 1988
Methods Study design: Randomized clinical trial.
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Not reported.
Participants Country: Israel.
Dates: Not reported.
Number randomized: 51: 23 to aspirin, 28 to observation.
Age: Mean age: 20 years.
Sex: Not reported.
Race: Not reported.
Sickle cell disease: Not reported.
Author stated that participants were balanced with respect to baseline characteristics
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. Age < 7 years;

  2. Diastolic blood pressure > 100 mmHg;

  3. Current use of anticoagulants;

  4. Current use of antihypertensive medication;

  5. Peptic ulcer;

  6. “Restless”.
Interventions Treatment: Aspirin 500 mg 3 times/day for 5 days.
Control: Observation
Treatment for both groups included:

  1. Bed rest;

  2. Topical atropine 1% and dexamycin 0.1% 4 times/day; and

  3. Topical timolol or oral acetazolamide if IOP > 25 mmHg.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily. Documented by estimating percentage involvement and plotting diagrammatically; definition not reported
Secondary outcomes:

  1. VA, assessed daily for 7 days; and

  2. IOP assessed daily for 7 days; details not reported.
Notes
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation not reported.
Allocation concealment (selection bias) Low risk Allocation was concealed from investigators by use of sequentially numbered envelopes
Blinding (performance bias and detection bias)
Participants		 High risk The participants were not masked to treatment. No placebo was given to the control group
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 High risk The healthcare providers were not masked to treatment. No placebo was given to the control group
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk No exclusions or loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk No exclusions or loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Unclear risk Only report results for secondary hemorrhage.
Other bias Low risk Poor description of study methods and results in publication
McGetrick 1983
Methods Study design: Randomized, double-masked clinical trial.
Exclusions after randomization: The chart of 1 participant in the placebo group was “lost” and this participant was excluded
Losses to follow-up: None.
Intention-to-treat: The excluded participant was not included in the analyses and the intention-to-treat principle was not followed in the analyses
Sample size calculations: Not reported.
Participants Country: USA.
Dates: August 1980 to February 1982.
Number randomized: 50: 28 to aminocaproic acid, 22 to placebo
Age: 86% ages 6–40 years.
Sex: 81% male.
Race: 69% black people, 21% Hispanic people, and 10% white people
Sickle cell disease: None; excluded.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Nonpenetrating traumatic hyphema.
Exclusion criteria:

  1. Penetrating ocular injury;

  2. Requirement for immediate surgical intervention;

  3. Sickle cell hemoglobin;

  4. History of intravascular coagulopathy;

  5. Pregnancy.
Interventions Treatment: Oral aminocaproic acid 100 mg/kg (up to 5 g/dose and 30 g/day) every 4 hours, for 5 days
Control: Placebo every 4 hours, for 5 days.
Treatment for both groups included:

  1. Quiet activities;

  2. No reading;

  3. Patch and shield on affected eye;

  4. Topical 1% atropine sulfate 4 times/day;

  5. Oral acetaminophen (paracetamol) up to 650 mg/day;

  6. No aspirin; and

  7. Topical timolol maleate 0.25% or 0.5% and oral acetazolamide, if IOP > 35 mmHg
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily by slit lamp exam. Defined as a definite increase in the amount of blood in the anterior chamber following admission
Secondary outcomes:

  1. Time to resolution of primary hemorrhage;

  2. Time to secondary hemorrhage;

  3. VA (final) with follow-up ranging from 0 to 9 months;

  4. IOP assessed daily by applanation tonometry for 5 days;

  5. Length of hospitalization; and

  6. Risk of complications and adverse events.
Notes Funded by the National Eye Institute, National Institutes of Health, Bethesda, MD and Research to Prevent Blindness, Inc
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Low risk Participants assigned to treatment groups using computerized randomization
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 Low risk Authors used a placebo control and stated that the study was double-masked
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Authors used a placebo control and stated that the study was double-masked. Assignment codes were not broken until the study was terminated
Incomplete outcome data (attrition bias)
Primary outcome		 Unclear risk The chart of 1 participant in the placebo group was “lost” and this participant was excluded. The excluded participant was not included in the analyses and the intention-to-treat principle was not followed in the analyses
Incomplete outcome data (attrition bias)
Secondary outcomes		 Unclear risk The chart of 1 participant in the placebo group was “lost” and this participant was excluded. The excluded participant was not included in the analyses and the intention-to-treat principle was not followed in the analyses
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Palmer 1986
Methods Study design: Randomized, double-masked clinical trial.
Exclusions after randomization: 2 participants were excluded: 1 from the low-dose aminocaproic acid group due to need for surgery and 1 from the standard-dose aminocaproic acid group due to severe hypotension
Losses to follow-up: None.
Intention-to-treat: The intention-to-treat principle was followed only for analyses of adverse events. The 2 excluded participants were not included in the analyses and the intention-to-treat principle was not followed in the analyses
Sample size calculations: Not reported.
Participants Country: USA.
Dates: July 1982 to December 1983.
Number randomized: 59: 26 to low-dose aminocaproic acid, 33 to standard-dose aminocaproic acid
Age: Mean age: low-dose aminocaproic acid 20 years (range 4–46 years), standard-dose aminocaproic acid 22.8 years (range 3–50 years)
Sex: 23 (88%) of low-dose aminocaproic acid and 27 (82%) of standard-dose aminocaproic acid were male
Race: 13 (50%) black people, 7 (27%) white people, and 5 (19%)Hispanic people in the low-dose aminocaproic acid group, the race of the excluded participant was not reported; and 17 (52%) black people, 7 (27%) white people, and 9 (21%) Hispanic people in the standard-dose aminocaproic acid group
Sickle cell disease: None; excluded.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Traumatic hyphema, including both primary and secondary hemorrhages
Exclusion criteria:

  1. Requirement for immediate surgical intervention;

  2. Sickle cell hemoglobin;

  3. History of intravascular coagulopathy;

  4. Pregnancy.
Interventions Treatment: Low-dose (50 mg/kg) oral aminocaproic acid (up to 5 g/dose or 30 g/day) every 4 hours, for 5 days
Control: Standard-dose (100 mg/kg) oral aminocaproic acid (up to 5 g/dose or 30 g/day) every 4 hours, for 5 days
Treatment for both groups included:

  1. Quiet activities;

  2. No reading;

  3. Head of bed elevated to 30°;

  4. Patch and shield on affected eye;

  5. Topical 1% atropine sulfate 4 times/day;

  6. Oral acetaminophen (paracetamol) up to 650 mg/day;

  7. No aspirin;

  8. Topical timolol maleate 0.25% or 0.5% and oral acetazolamide if IOP > 25 mmHg;

  9. Oral prochlorperazine edisylate (5 or 10 mg) if nausea or vomiting; and

  10. Steroids on recommendation of admitting physician.
Outcomes Primary outcome: Incidence of secondary hyphema, assessed daily by slit lamp exam. Documented by level in mm and percentage of anterior chamber filled with blood. Defined as a definite increase in the amount of fresh blood in the anterior chamber over level at admission
Secondary outcomes:

  1. Time to resolution of primary hemorrhage;

  2. Time to secondary hemorrhage;

  3. VA; “final” VA not defined;

  4. IOP assessed daily using applanation tonometry;

  5. Length of hospitalization;

  6. Incidence of complications and adverse events.
Notes Funded by the National Eye Institute, National Institutes of Health, Bethesda, MD, Research to Prevent Blindness, Inc., and Lederle-Cyanamid Laboratories for serum assays
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Low risk Assignments determined by computerized randomization in the pharmacy
Allocation concealment (selection bias) Low risk Allocation was possibly concealed from investigators by pharmacy preparation of drugs
Blinding (performance bias and detection bias)
Participants		 Low risk Participants masked by preparation of drugs by pharmacy. “The treating physicians and the patients were not told of the admission dose in order to maintain the double-masked status.”
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Healthcare providers and outcomes assessors masked by preparation of drugs by pharmacy. “The treating physicians and the patients were not told of the admission dose in order to maintain the double-masked status.”
Incomplete outcome data (attrition bias)
Primary outcome		 Unclear risk 2 participants were excluded: 1 from the low-dose aminocaproic acid group due to need for surgery and 1 from the standard-dose aminocaproic acid group due to severe hypotension. The study authors noted that excluding the patient from the standard group did not affect the statistical results
Incomplete outcome data (attrition bias)
Secondary outcomes		 Unclear risk 2 participants were excluded: 1 from the low-dose aminocaproic acid group due to need for surgery and 1 from the standard-dose aminocaproic acid group due to severe hypotension. The intention-to-treat principle was followed only for analyses of adverse events
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Pieramici 2003
Methods Study design: Randomized, double-masked, placebo-controlled clinical trial
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: 124 participants based on secondary hemorrhage rate of 15% and 3% in placebo- and aminocaproic acid-treated participants, respectively, with alpha = 0.05, power = 80%, and one-tailed test of significance; study terminated due to slow enrollment
Notes: Multicenter study with 8 centers.
Participants Country: USA.
Dates: Not reported, although study was conducted over 14 months
Number randomized: 51: 24 to aminocaproic acid, 27 to placebo
Age: Mean age: aminocaproic acid group 24 ± 4 years (range 4–73 years), placebo group
23 ± 3 years (range 6–48 years)
Sex: 21 (88%) of aminocaproic acid group and 23 (85%) of placebo group were male
Race: 15 (63%) white people, 8 (33%) black people, and 1 (1%) other in aminocaproic acid group and 13 (48%) white people, 11 (41%) black people, and 3 (11%) other in placebo group
Sickle cell disease: 2/24 (8%) of participants in aminocaproic acid group and 1/27 (4%) of participants in placebo group had sickle cell trait
Participants appeared to be balanced with respect to baseline characteristics except for race and size of primary hyphema with larger hyphemas found in the placebo group
Inclusion criteria: Traumatic hyphema
Exclusion criteria:

  1. Total hyphema or unlayered microscopic hyphema;

  2. More than 36 hours since trauma;

  3. Age less than 4 years;

  4. History of clinically significant coagulopathy, renal insufficiency, or hepatic insufficiency;

  5. Hypersensitivity or idiosyncratic reaction of proparacaine hydrochloride 0.5%, aminocaproic acid, or carboxymethylene;

  6. Evidence of any clinically significant cardiac, endocrine, gastrointestinal, hematologic, or immunologic abnormalities or disease (sickle cell disease was allowed);

  7. Ingestion of anticoagulant or antiplatelet agent within the previous 7 days or any nonsteroidal anti-inflammatory drug within previous 24 hours;

  8. Pregnancy;

  9. Participation in investigational drug trial within 4 weeks before randomization;

  10. Unable to complete trial.
Interventions Treatment: Following 1 drop of 0.05% proparacaine hydrochloride, 30% aminocaproic acid in 2% gel instilled in inferior fornix every 6 hours, for 5 days
Control: Following 1 drop of 0.05% proparacaine hydrochloride, placebo gel instilled in inferior fornix every 6 hours, for 5 days
Treatment for both groups included:

  1. No reading or video games;

  2. Head of bed elevated to 30°;

  3. Shield on affected eye;

  4. Topical 2% homatropine sulfate 3 times/day;

  5. No topical steroids; and

  6. If IOP elevated, treatment at discretion of physician.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily by slit lamp exam for 7 days; defined as increase in height of hyphema of at least 0.5 mm above darker blood, color change of blood of at least 0.5 mm, obvious new “trickle” of blood on iris, or reappearance of blood after resolution
Secondary outcomes:

  1. Time to resolution of primary hemorrhage;

  2. Time to secondary hemorrhage;

  3. VA, final VA assessed at 7 days (end of treatment);

  4. Risk of complications and adverse events.
Notes Funded by Orphan Medical Inc., Covance Inc, National Eye Institute, National Institutes of health, Bethesda, MD, and Research to Prevent Blinding
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Low risk Participants assigned to treatment groups using computerized randomization
Allocation concealment (selection bias) Low risk Allocation was concealed from investigators in that treatment assignments were based on a trial number obtained from a contract research organization
Blinding (performance bias and detection bias)
Participants		 Low risk Authors used a placebo control and stated that the study was double-masked. “The investigators and patients were masked to the treatment arm.”
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Authors used a placebo control and stated that the study was double-masked. “The investigators and patients were masked to the treatment arm.”
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk No exclusions or loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk No exclusions or loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Unclear risk “There were a number of protocol violations noted in both study groups.”
“During the course of the study, only 8 of the original 13 sites enrolled patients, and at 14 months a total of 51 patients were enrolled overall. The study was terminated at this point by Orphan Medical, the manufacturer, against the advice of the principal investigators, because of slow enrollment.”
Rahmani 1999
Methods Study design: Randomized, placebo-controlled clinical trial
Exclusions after randomization: 6; 2 participants in the tranexamic acid group, 3 in the prednisone group, and 1 in the placebo group left the hospital before the end of the study and were excluded
Losses to follow-up: None.
Intention-to-treat: The excluded participants were not included in the analyses and the intention-to-treat principle was not followed in the analyses
Sample size calculations: Not reported.
Participants Country: Iran.
Dates: January 1991 to May 1992.
Number randomized: 244: 82 to tranexamic acid, 81 to prednisone, 81 to placebo
Age: Median age: tranexamic acid 11 years (range 1–65 years); prednisone 11.5 years (range 1–50 years), placebo 12 years (range 1–58 years)
Sex: 63 (79%) of tranexamic acid group, 58 (73%) of prednisone group, and 66 (82%) of placebo group were male
Race: 100% white people.
Sickle cell disease: Not reported, but all white study population
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. Penetrating ocular injury;

  2. Total hyphema or unlayered microscopic hyphema;

  3. Definite secondary hemorrhage before entry;

  4. More than 48 hours since trauma;

  5. Requirement for immediate surgical intervention;

  6. History of renal insufficiency;

  7. Acid peptic disease;

  8. Recent ingestion of aspirin or anticoagulant;

  9. Use of topical corticosteroids after trauma;

  10. Pregnancy.
Interventions Treatment 1: Oral tranexamic acid 75 mg/kg per day, divided into 3 doses/day, for 5 days
Treatment 2: Oral prednisolone 0.75 mg/kg per day, divided into 2 doses/day, for 5 days
Control: Placebo administered 3 times/day.
Treatment for all groups included:

  1. Limited ambulation;

  2. Head of bed elevated;

  3. Patch and shield on affected eye;

  4. Topical cyclopentolate for exam of the retina if necessary;

  5. Oral acetaminophen (paracetamol) for pain;

  6. No aspirin or topical corticosteroids;

  7. Topical timolol and oral acetazolamide, if elevated IOP; and

  8. Oral promethazine if nausea or vomiting.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily by slit lamp exam for 5 days. Defined as definite increase in size of level of blood or appearance of fresh blood over darker clotted blood in the anterior chamber
Secondary outcomes:

  1. VA, measured at day 5 (discharge); and

  2. Risk of complications and adverse events.
Notes
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Low risk Randomization was based on a randomization list.
Allocation concealment (selection bias) Unclear risk Participants assigned to treatment groups using a randomization list, but not clear whether list was revealed before allocation to individuals enrolling participants
Blinding (performance bias and detection bias)
Participants		 Unclear risk Participants partially masked in that authors used a placebo control for the tranexamic acid, but not for prednisone
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Healthcare providers partially masked in that authors used a placebo control for the tranexamic acid, but not for prednisone; however, ophthalmologists and outcome assessors were masked
Incomplete outcome data (attrition bias)
Primary outcome		 Unclear risk 6 patients were excluded from the study: 2 in tranexamic acid group, 3 in prednisone group, and 1 in placebo group left the hospital before the end of the study and were excluded. The excluded participants were not included in the analyses and the intention-to-treat principle was not followed in the analyses
Incomplete outcome data (attrition bias)
Secondary outcomes		 Unclear risk 6 patients were excluded from the study: 2 in tranexamic acid group, 3 in prednisone group, and 1 in placebo group left the hospital before the end of the study and were excluded. The excluded participants were not included in the analyses and the intention-to-treat principle was not followed in the analyses
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Rakusin 1972
Methods Study design: Quasi-randomized controlled series.
Exclusions after allocation: 59 patients in the series with large hyphemas underwent surgery and were not included in the analysis
Losses to follow-up: 20.
Intention-to-treat: All participants were not accounted for in the final analyses, thus intention-to-treat analysis was not followed
Sample size calculations: Not reported.
Participants Country: South Africa.
Dates: 1966–1969.
Number allocated: 390 consecutive patients.
Age: Not reported.
Sex: Not reported.
Sickle cell disease: Not reported.
Race: 90% African origin, 10% Asiatic origin.
Inclusion criteria: Traumatic hyphema.
Exclusion criteria: Surgical treatment indicated.
Interventions Series of comparisons based on 6 variable factors:

  1. Bed rest (n = 26) vs. ambulatory treatment (n = 26);

  2. Eye pads: bilateral eye pads (n = 27) vs. single eye pads (n = 26) vs. no eye pads (n = 10);

  3. Topical antibiotics (0.5% chloramphenicol, n = 21) vs. corticosteroids, 0.5% hydrocortisone acetate (n = 13) vs. neither (n = 3);

  4. Mydriatics (1% homatropine, n = 17) vs. miotics (4% pilocarpine, n = 17) vs. neither (n = 19) vs. both (n = 17);

  5. Enzymes: oral trypsin (n = 15) vs. oral papase (n = 18) vs. neither (n = 10);

  6. Ocular hypotensive agents: acetazolamide 250 mg (n = 18) vs. oral glycerol 1 mL/kg (n = 18) vs. neither (n = 10).


Treatment and control groups followed the same regimen except even-numbered patients received the variable factor, and odd-numbered patients did not
Excluding the variable factor for each series, all patients received bed rest, single pad over the injured eye, and topical chloramphenicol or chloromycetin
Outcomes Primary outcomes:

  1. Speed of absorption of blood from the anterior chamber;

  2. Risk of secondary hemorrhage;

  3. Complications of the hyphema; and

  4. Final VA.


Follow-up: Range 1–2 weeks to 3 years
Notes Funded by the University of Witwatersrand, the South African Medical Research Council, Leo Laboratories, Mer-National, and Warner Pharmaceutical Co
In the third comparison group, antibiotics versus corticosteroids, 3 patient were assigned to receive neither treatment, but this group was discontinued after all 3 patients developed a mucous conjunctival discharge
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) High risk Method of allocation unclear, not all patients in the series were allocated to the 6 comparisons under study; 59 patients were selected for surgery. Also even and odd patient number allocation is not applicable to comparison with 3 treatment groups
Allocation concealment (selection bias) High risk Method of allocation concealment not reported, not randomized
Blinding (performance bias and detection bias)
Participants		 High risk Masking of patients was not possible for some variables (i.e. bed rest and eye patching). Use of placebo for other variables was not mentioned
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Unclear risk Masking was not reported.
Incomplete outcome data (attrition bias)
Primary outcome		 Unclear risk 79 participants were not included in the analyses and the intention-to-treat principle was not followed
Incomplete outcome data (attrition bias)
Secondary outcomes		 Unclear risk 79 participants were not included in the analyses and the intention-to-treat principle was not followed
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Unclear risk The primary interventions of interest for this study were not clear. Although the majority of the patients in the series were assigned to 1 of 6 conservative treatment comparison groups, 59 recruited patients were selected for surgery
Read 1974
Methods Study design: Quasi-randomized controlled series.
Exclusions after allocation: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the groups to which they were assigned
Sample size calculations: Not reported.
Participants Country: USA.
Dates: February 1970 to July 1972.
Number allocated: 137 consecutive patients.
Age: Mean 15.9 years.
Sex: 108 men and 29 women; 79% male.
Race: 101 (74%) African-American.
Sickle cell disease: Not reported.
Participants were similar in regards to baseline characteristics
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. Associated penetrating ocular injury;

  2. Surgical exploration for suspected rupture of the globe;

  3. Bodily injury;

  4. Recurrent ocular injury;

  5. Personal or family history of diabetes or bleeding disorders.
Interventions Medical treatment #1 (n = 66): Bed rest with elevation of head to 30°, bilateral ocular patches and shield over injured eye, and sedation
Medical treatment #2 (n = 71): Moderate ambulatory activity in the hospital, patching and shielding of the traumatized eye only, and no sedation
Eyedrops were not administered in either medical treatment regimen
On day 5, patients with remaining major primary or secondary hyphemas (n = 16) were alternately assigned to continue with medical treatment or to receive surgical intervention (ab externo corneal section with clot expression)
Outcomes Primary and secondary outcomes not specified.
Measured outcomes:

  1. Changes or presence of IOP;

  2. Duration of primary hyphema;

  3. Risk of secondary hemorrhage;

  4. Risk of corneal staining;

  5. Need for surgical intervention;

  6. Complications of the hyphema; and

  7. Final VA.


Follow-up: 1 week, 1, 3, and 6months (range 3 months to 2.5 years; mean 16.5months)
Notes Funded by a grant from the Research to Prevent Blindness, Inc
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) High risk Allocation was not randomized; alternately assigned patients to treatment groups at time of admission. Imbalance in number assigned to each group (66 vs. 71) makes it appear alternation was not systematic
Allocation concealment (selection bias) High risk Allocation was assigned on an alternate basis.
Blinding (performance bias and detection bias)
Participants		 High risk Masking of patients was not possible with the interventions being studied
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 High risk All patients were treated by the primary investigator in order to standardize therapy and record results as accurately as possible
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk All participants were analyzed in the group to which they were assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk All participants were analyzed in the group to which they were assigned
Selective reporting (reporting bias) Low risk Reported results for all outcomes.
Other bias High risk A subset of patients with major hyphema on day 5 were alternately allocated to either continue with medical treatment as originally assigned or undergo surgical intervention. Thus, the patients that had surgery were censored on day 5 from their medical treatment outcomes
Spaeth 1966
Methods Study design: Randomized, double-masked, placebo-controlled clinical trial
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Not reported.
Participants Country: USA.
Dates: 1963–1964.
Number randomized: 85: 39 to estrogen, 46 to placebo.
Age: Mean age: estrogen 16.2 years (range 2–62 years), placebo 18.9 years (range 0.5–65 years)
Sex: 80% of estrogen group, 85% of placebo group were male.
Race: 72% of estrogen group, 70% of placebo group were black people; remaining participants were white people
Sickle cell disease: Not reported
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. Penetrating ocular injury;

  2. More than 24 hours since trauma;

  3. History of ocular disease;

  4. Failure to co-operate.
Interventions Treatment: Conjugated estrogen, 5 mg intramuscularly for children < 5 years; 10 mg intramuscularly for children 5 years or older but < 10 years; and 20 mg intravenously for children 10 years or older and adults, for 5 days
Control: Placebo, for 5 days.
Treatment for both groups included:

  1. Complete bed rest;

  2. Head of bed elevated;

  3. Patches on both eyes;

  4. No ophthalmic drops; and

  5. Sedation and analgesics as needed.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily by “complete ocular examination” for 5 days. Documentation and definition not reported
Secondary outcomes:

  1. Time to secondary hemorrhage;

  2. VA measured at day 5 (discharge); and

  3. Risk of complications and adverse events.
Notes Placebo and conjugated estrogen supplied by Ayerst Laboratory
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation not reported.
Allocation concealment (selection bias) Low risk Allocation was concealed from investigators by use of coded bottles
Blinding (performance bias and detection bias)
Participants		 Low risk Authors used coded bottles to mask participants. “Neither the person administering nor the patient receiving the medications knew whether estrogen or placebo was being given.”
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Authors used coded bottles to mask health-care providers and outcomes assessors. “Neither the person administering nor the patient receiving the medications knew whether estrogen or placebo was being given.”
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Spoor 1980
Methods Study design: Randomized, double-masked, placebo-controlled clinical trial
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Not reported.
Participants Country: USA.
Dates: September 1975 to December 1977.
Number randomized: 43: 23 to prednisone, 20 to placebo.
Age: Mean age: prednisone group 20.1 years (range 5–61 years), placebo group 21.2 years (range 9–51 years)
Sex: 16 (70%) of prednisone group, 16 (80%) of placebo group were male
Race: There were 14 (61%) white people, 6 (26%) Hispanic people, and 3 (13%) black people in prednisone group. There were 11 (55%) white people, 7 (35%) Hispanic people, and 2 (10%) black people in placebo group
Sickle cell disease: Not reported.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. Penetrating ocular injury;

  2. More than 24 hours since trauma;

  3. Treated before entry;

  4. Not available for 6 months follow-up.
Interventions Treatment: Oral prednisone 40 mg/day for adults and children > 10 years; 15 mg/day for children ages 4–10 years; and 10 mg/day for children ages 18 months to 4 years, for 7 days
Control: Lactose placebo capsules administered daily for 7 days
Treatment for both groups included:

  1. Bed rest;

  2. Head of bed elevated 30–45°;

  3. Patch on affected eye;

  4. No topical medications;

  5. Sedation as needed;

  6. No aspirin; and

  7. Oral acetazolamide if IOP > 24 mmHg.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily for 7 days, using slit lamp exam, documented by drawings or photography
Secondary outcomes:

  1. Time to resolution of primary hemorrhage;

  2. Time to secondary hemorrhage;

  3. VA (followed up to 6 months);

  4. IOP assessed daily for 7 days using applanation tonometry;

  5. Risk of complications and adverse events.
Notes
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation not reported.
Allocation concealment (selection bias) Low risk Allocation was concealed from investigators by use of encoded capsules prepared by pharmacy
Blinding (performance bias and detection bias)
Participants		 Low risk Participants by use of encoded capsules prepared by pharmacy. “Neither the doctor nor the patient knew which capsule the patient was receiving until the conclusion of the course of treatment and follow-up.”
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Healthcare providers and outcomes assessors by use of encoded capsules prepared by pharmacy. “Neither the doctor nor the patient knew which capsule the patient was receiving until the conclusion of the course of treatment and follow-up.”
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Sukumaran 1988
Methods Study design: Quasi-randomized controlled series.
Exclusions after allocation: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were assigned
Sample size calculations: Not reported.
Participants Country: Malaysia.
Dates: Not reported.
Number allocated: 35 consecutive patients.
Age: 80% below 30 years old.
Sex: 35 men.
Race: Not reported.
Sickle cell disease: Not reported.
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. Other serious ocular or facial injuries;

  2. Hyphema greater than 7 mm.
Interventions Treatment (n = 17): oral tranexamic acid (cyklokapron) 25 mg/kg divided into 3 doses for 7 days in addition to routine treatment
Control (n = 18): Routine treatment.
Routine treatment for both groups included:

  1. Bilateral patching;

  2. Bed rest;

  3. Sedation;

  4. Analgesics when required; and

  5. Topical corticosteroid drops from the third day for 1 week.
Outcomes Primary outcomes:

  1. Risk of secondary hemorrhage;

  2. Speed of recovery; and

  3. Final VA


Follow-up: At least 1 week
Notes Funding source not reported.
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) High risk Method of allocation unclear, not randomized.
Allocation concealment (selection bias) High risk Method of allocation concealment not reported, not randomized
Blinding (performance bias and detection bias)
Participants		 High risk No placebo was used for the control group.
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Unclear risk Masking was not reported.
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk All participants were analyzed in the group to which they were assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk All participants were analyzed in the group to which they were assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Teboul 1995
Methods Study design: Randomized, double-masked, placebo-controlled clinical trial
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Authors reported that sample sizes were not calculated because the rate of secondary hemorrhage in children was unknown and that of other populations was too variable to estimate
Participants Country: Canada.
Dates: November 1987 to February 1994.
Number randomized: 94: 48 to aminocaproic acid, 46 to placebo
Age: Mean age: aminocaproic acid group 8.2 years, placebo group 10.6 years
Sex: 42 (88%) of aminocaproic acid group, 39 (85%) of placebo group were male
Race: 43 (90%) of aminocaproic acid group, 42 (91%) of placebo group were white
Sickle cell disease: None; excluded.
Participants appeared to be balanced with respect to baseline characteristics, except for mean age where the aminocaproic acid group was younger (8.2 to 10.6 years)
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. Penetrating ocular injury;

  2. Total hyphema;

  3. More than 24 hours since trauma;

  4. Requirement for immediate surgical intervention;

  5. History of sickle cell anemia, renal disease, hepatic disease, cardiac disease, or coagulopathy;

  6. Recent ingestion of aspirin up to 1 week before entry;

  7. Pregnancy.
Interventions Treatment: Oral aminocaproic acid 100 mg/kg every 4 hours (up to 30 g/day), for 5 days
Control: Placebo every 4 hours, for 5 days.
Treatment for both groups included:

  1. Bed rest;

  2. Head of bed elevated to 45°;

  3. Patch on affected eye;

  4. 1% atropine ointment nightly and garsone drops 2 times/day;

  5. Oral acetaminophen (paracetamol) for pain;

  6. No aspirin;

  7. Topical timolol maleate 0.5% 2 times/day and oral acetazolamide if IOP > 25 mmHg; and

  8. Dimenhydrinate (Gravol) if nausea or vomiting.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed by daily slit lamp exam for 5 days; documented by drawing of hyphema with distinction between fresh and clotted blood
Secondary outcomes:

  1. Time to resolution of primary hemorrhage;

  2. Time to secondary hemorrhage;

  3. VA at final visit (follow-up range 5 days to 3.4 years);

  4. IOP measured daily for 5 days using applanation tonometry;

  5. Length of hospitalization; and

  6. Risk of complications and adverse events.
Notes
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation not reported.
Allocation concealment (selection bias) Low risk Allocation was concealed from investigators by preparation of drugs by pharmacy; statement that investigators were unaware of next treatment assignment
Blinding (performance bias and detection bias)
Participants		 Low risk Participants by use of medications prepared by pharmacy.
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Healthcare providers and outcomes assessors by use of medications prepared by pharmacy. “The double-blind code was not broken until completion of the study.”
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk “The authors have no proprietary interest in aminocaproic acid or any competing drug.”
Vangsted 1983
Methods Study design: Randomized clinical trial.
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Not reported.
Participants Country: Sweden.
Dates: November 1978 to May 1981.
Number randomized: 112: 59 to tranexamic acid, 53 to bed rest
Age: Mean age: tranexamic acid group 23.5 years (range 9–60 years), bed rest group 23. 5 years (range 9–67 years)
Sex: Ratio of male:female 4:1.
Race: Not reported.
Sickle cell disease: Not reported.
Participants appeared to be balanced with respect to baseline characteristics
Inclusion criteria: Traumatic hyphema.
Exclusion criteria:

  1. Penetrating ocular injury;

  2. Microscopic hyphema;

  3. More than 24 hours since trauma;

  4. Younger than 8 years of age;

  5. History of renal disease with creatine > 115 micromol/L;

  6. Serious blood dyscrasia or earlier thrombotic disease;

  7. Pregnancy.
Interventions Treatment: Oral tranexamic acid 25 mg/kg 3 times/day, for 7 days
Control: Complete bed rest, for 6 days.
Treatment for both groups included:

  1. Patch on affected eye;

  2. 1% atropine once/day;

  3. Dexamethasone 3 times/day;

  4. No aspirin; and

  5. Oral acetazolamide if IOP > 25 mmHg.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily by slit lamp exam at days 2 and 7. Documentation and definition not reported
Secondary outcomes:

  1. Time to resolution of primary hemorrhage;

  2. VA measured at day 2 and 7;

  3. IOP measured using applanation tonometry at day 2 and 7;

  4. Length of hospitalization; and

  5. Risk of complications and adverse events.
Notes
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation not reported.
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 High risk Participants were not masked to treatment assignment (bed rest vs. tranexamic acid)
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 High risk Healthcare providers and outcome assessors were not masked to treatment assignment (bed rest vs. tranexamic acid)
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Varnek 1980
Methods Study design: Quasi-randomized controlled series.
Exclusions after allocation: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were assigned
Sample size calculations: Not reported.
Participants Country: Denmark.
Dates: March 1978 to November 1979.
Number allocated: 232 consecutive patients from 4 study centers
Age: Mean 24.4 years.
Sex: 188 men, 44 women; 81% male.
Race: 100% white people.
Sickle cell disease: Not reported, but all white study population
Inclusion criteria:

  1. Traumatic hyphema with sedimented hyphema or visible clots in the anterior chamber; and

  2. Admitted less than 24 hours after sustaining injury.


Exclusion criteria:

  1. Patients with hemorrhagic flare only;

  2. Pregnancy;

  3. Perforating eye injuries.
Interventions Treatment (n = 102): oral tranexamic acid 25 mg/kg divided into 3 doses for 6 days
Control (n = 130): Conservative treatment.
Treatment for both groups included:

  1. Hospitalization;

  2. Bed rest; and

  3. Stenopaeic glasses for 5 days.
Outcomes Primary outcomes:

  1. Risk of secondary hemorrhage;

  2. Speed of absorption of primary hemorrhage;

  3. Final VA; and

  4. Length of hospitalization.


Follow-up: Days 5 and 12.
Notes Funding source not reported.
Method used to calculate mean VA not reported.
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) High risk Allocation was not randomized; assigned patients to treatment groups based on date of admission
Allocation concealment (selection bias) High risk Method of allocation based on even versus odd admission dates
Blinding (performance bias and detection bias)
Participants		 High risk No placebo was used for the control group.
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 High risk Masking was not done because of the noticeable delay in resolution time between treatment groups. Tranexamic acid was considered to induce persistence of the primary clot a priori
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk All participants were analyzed in the group to which they were assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk All participants were analyzed in the group to which they were assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Wang 1994
Methods Study design: Randomized clinical trial.
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Not reported.
Participants Country: China.
Dates: Not reported.
Number randomized: 83: 45 in treatment group, 38 in control group
Age: Range 4–49 years.
Sex: 56 (67%) men, 27 (33%) women.
Race: Not reported.
Sickle cell disease: Not reported.
Participants appeared to be balanced with respect to baseline characteristics (P value > 0.05 for between-group comparisons for anterior chamber blood volume, IOP, gender, and age). Severity of hyphema not reported; however, in the treatment group, 29 (64%) participants were given the medicine within 24 hours after the trauma, 13 (29%) cases were given the medicine within 3 days after the trauma, and 3 (7%) cases were given the medicine at day 5 after the trauma; for the control group, 31 (82%) participants were given the medicine (Carbazochrome or Etamsylate) within 24 hours after the trauma, and 7 (18%) cases were given the medicine within 3 days after the trauma
Inclusion criteria: Any degree of traumatic hyphema.
Exclusion criteria: Not reported.
Interventions Treatment (n = 45): Yunnan Baiyao (a traditional Chinese medicine) was given to the participants in the treatment group. The participants were assigned to take 0.5 g of the medicine 4 times/day orally, accompanied by vitamin C and vitamin K also taken orally, and with 0.5% vinegar eyedrops [ Inline graphic]. The length of treatment was up to 5 days (until complete resolution)
Control (n = 38): participants in the control group were given medicines such as Carbazochrome or Etamsylate to help with stopping bleeding
Follow-up: 1 week.
Outcomes Primary outcome: Number of participants “cured”, defined as complete resolution within 5 days, VA of 0.7 or better, and no rebleed within 1 week
Secondary outcomes: None reported.
Notes Funding source not reported.
Poor description of study methods and outcomes in publication
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation not reported.
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 Unclear risk Masking of participants was not reported.
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Unclear risk Masking of outcome assessors was not reported.
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Unclear risk Study outcomes of interest not clearly stated.
Other bias High risk 2 different control interventions were described, but method used to decide which participants received which control intervention not stated. Why ’vinegar eye drops’ were used in the experimental group not described. Length of time between onset of hyphema and initiation of treatment differed between treatment groups
Welsh 1983
Methods Study design: Randomized, double-masked, placebo-controlled clinical trial
Exclusions after randomization: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were randomly assigned
Sample size calculations: Not reported.
Participants Country: South Africa.
Dates: Not reported.
Number randomized: 39: 19 to tranexamic acid, 20 to placebo.
Age: Mean age: tranexamic acid group 25.2 years (range 15–38 years), placebo group 25. 2 years (range 14–52 years)
Sex: 15 (79%) of tranexamic acid group, 17 (85%) of placebo group were male
Race: 100% black people.
Sickle cell disease: Not reported.
Participants appeared to be balanced with respect to baseline characteristics. 3 of 39 patients had a hyphema due to cataract surgery; 2 in the tranexamic group and 1 in the control group
Inclusion criteria: Hyphema; either nonperforated, or if perforated, then the wound was sutured and treated as closed injury
Exclusion criteria:

  1. More than 5 days since onset;

  2. Age 14 or older;

  3. Presence of hypertension;

  4. History of thrombocytic event;

  5. Diabetes;

  6. Renal impairment;

  7. Uremia;

  8. Presence of coma;

  9. Pregnancy.
Interventions Treatment: 3 × 500 mg tablets of oral tranexamic acid 3 times/day for 7 days, for an overall total of 31.5 g of tranexamic acid
Control: 3 tablets of placebo 3 times/day for 7 days.
Treatment for both groups included:

  1. Bed rest;

  2. Patch on affected eye;

  3. 1% atropine once/day;

  4. 4% pilocarpine once/day;

  5. Cortisone eyedrops once/day.
Outcomes Primary outcome: Risk of secondary hemorrhage, assessed daily by visual exam. Documentation and definition not reported
Secondary outcomes:

  1. Percentage area of hyphema, measured daily;

  2. IOP measured daily; and

  3. Risk of complications and adverse events.
Notes Tranexamic acid and placebo supplied by Adcock Ingram Laboratories
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation not reported.
Allocation concealment (selection bias) Low risk Allocation was concealed from investigators by preparation of drugs by pharmacy; statement that investigators were unaware of next treatment assignment
Blinding (performance bias and detection bias)
Participants		 Low risk Participants by use of medications prepared by pharmacy. “Neither patient nor staff knew which tablet the patient was receiving and the code was broken by the pharmaceutical firm at the end of the trial.”
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Low risk Healthcare providers and outcomes assessors by use of medications prepared by pharmacy. “Neither patient nor staff knew which tablet the patient was receiving and the code was broken by the pharmaceutical firm at the end of the trial.”
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk There were no exclusions and no loss to follow-up. All participants were analyzed in the group to which they were randomly assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Unclear risk Cyklokapron and placebo tablets were supplied by Adcock Ingram Laboratories
Zetterstrom 1969
Methods Study design: Quasi-randomized controlled series.
Exclusions after allocation: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were assigned
Sample size calculations: Not reported.
Participants Country: Sweden.
Dates: September 1967 to September 1968.
Number allocated: 117 consecutive patients.
Age: Mean: 22.0 years (range 5–57 years).
Sex: 102 men and 17 women (as reported); 86% male.
Race: Not reported.
Sickle cell disease: Not reported.
Inclusion criteria: Traumatic hyphema.
Exclusion criteria: Perforation of the eyeball.
Interventions Treatment (n = 58): Topical atropine with Decadron (cortisone) eyedrops 5 times/day and moderate ambulatory activity within hospital
Control (n = 59): Conservative treatment of complete bed rest without pinhole glasses or simultaneous local therapy
Treatment for both groups included inpatient care until VA in the injured eye was satisfactory, the hyphema was absorbed, and IOP did not deviate from normal
Outcomes Primary outcomes:

  1. Length of hospitalization;

  2. Final VA;

  3. Risk of secondary hemorrhage; and

  4. Complication rates.


Follow-up: Followed until discharge; some patients with iritis were seen as outpatients after discharge
Notes Funding source not reported.
Method used to calculate mean VA not reported.
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) High risk Allocation was not randomized; alternately assigned patients to treatment groups based on order of admission
Allocation concealment (selection bias) High risk Method of allocation based on order of admission.
Blinding (performance bias and detection bias)
Participants		 High risk Masking of patients was not possible with the interventions being studied
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 Unclear risk Masking was not reported, but unlikely because of the types of interventions being studied
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk All participants were analyzed in the group to which they were assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk All participants were analyzed in the group to which they were assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.
Zi 1999
Methods Study design: Randomized controlled series.
Exclusions after allocation: None.
Losses to follow-up: None.
Intention-to-treat: All participants were analyzed in the group to which they were assigned
Sample size calculations: Not reported.
Participants Country: China.
Dates: September 1990 to 1997.
Number randomized: 79.
Age: Mean: 24.5 years (range 7–43 years).
Sex: 70 men and 4 women (as reported); 95% male.
Race: Not reported.
Sickle cell disease: Not reported.
Inclusion criteria: Hyphema.
Exclusion criteria: Not reported.
Interventions Treatment (n = 39): Alternatively right and left lateral position
Control (n = 35): Semi-reclined position.
Outcomes Primary outcomes: Time to resolution by severity.
Secondary outcomes:

  1. Discomfort; and

  2. Complications.


Follow-up: Not reported.
Notes Funding source not reported.
Risk of bias
Bias Authors’ judgement Support for judgement
Random sequence generation (selection bias) Unclear risk Randomized, but method of allocation not reported.
Allocation concealment (selection bias) Unclear risk Method of allocation concealment not reported.
Blinding (performance bias and detection bias)
Participants		 High risk Participants were not masked to treatment assignment (laying either semi-reclining or on side)
Blinding (performance bias and detection bias)
Personnel and outcome assessors		 High risk Healthcare providers and outcome assessors were not masked.
Incomplete outcome data (attrition bias)
Primary outcome		 Low risk All participants were analyzed in the group to which they were assigned
Incomplete outcome data (attrition bias)
Secondary outcomes		 Low risk All participants were analyzed in the group to which they were assigned
Selective reporting (reporting bias) Low risk Reported results for primary and secondary outcomes.
Other bias Low risk No other sources of potential bias were identified.

IOP: intraocular pressure; n: number of participants; VA: visual acuity.