After conducting a thorough review, fifty-two studies were eliminated based on our defined criteria. The exclusions stemmed from various reasons including intervention-related issues (n = 16), comparisons not made against standard argon laser PRP (n = 15), study design constraints (n = 12), irrelevant outcomes measured (n = 3), patient populations that did not have PDR (n = 3), and comparisons not pre-specified by this review (n = 3). For more detailed information, kindly refer to the table illustrating the characteristics of the excluded studies.
For further information, please visit weiqing.
The studies assessed did not record near visual acuity or other outcomes relevant to patients, such as loss of driving privileges or quality of life related to vision. Additionally, resource use and costs were not mentioned. The follow-up time varied, ranging from one month up to two years.
Approximately 50% of the studies reported measures on the regression or progression of PDR. Loss of visual field was only referenced in one study, while pain during laser treatment was noted in five of the studies.
Most studies, except two (Bandello; Tewari), measured and reported on our primary outcome, which was the gain or loss of at least 15 ETDRS letters (equivalent to three ETDRS lines), gauged using a LogMAR chart. In instances where follow-up data was not recorded at one or five years, we used the provided final time point.
All participants in the review received a different laser PRP strategy compared to standard argon laser PRP (characterized as midperipheral scatter and panretinal photocoagulation with a 0.1-second pulse duration of moderate laser intensity). The review encompassed diverse laser PRP interventions, including double-frequency Nd:YAG lasers (532 nm) (Bandello; Brancato), diode lasers (810 nm) (Bandello; Han; Tewari), longer exposure times with 0.5-second argon laser burns (Wade), 'light intensity' lower energy treatments with standard argon laser pulses (Bandello), 'mild scatter' patterns limited to 400 to 600 laser burns in a single session (Pahor), 'central PRP' which juxtaposed a more central approach (mean number of 437 laser burns more posteriorly with sparing of a 2 DD area around the fovea) versus a peripheral PRP (mean number of 441 laser burns further peripherally), along with mid-peripheral PRP (Blankenship), 'central sparing' approaches (Theodossiadis), and 'extended targeted' argon laser PRP to cover the entire retina anterior to the equator (Nikkhah). Additional details are provided in the table on included studies.
All studies featured participants aged between 18 and 79 years, comprising both male and female adults diagnosed with type 1 or type 2 diabetes mellitus. At least one eye of each participant had to exhibit high-risk proliferative diabetic retinopathy as outlined by the ETDRS definition, with the exception of Han, who only included participants with one affected eye. Notably, none of the studies adjusted for within-person correlation in their analyses. One within-person investigation (Tewari) also lacked appropriate matched analysis. The average age of participants across all studies ranged from 40 to 58 years, while their mean visual acuity ranged from 0.12 to 0.89 LogMAR. The sizes of the studies varied between 20 and 270 eyes.
In total, we reviewed 11 studies, all of which were randomized controlled trials, conducted in the US (2), Italy (4), South Korea (1), India (1), Iran (1), Slovenia (1), and Greece (1). Documentation regarding the sources of sponsorship was inadequate, though two studies disclosed public funding (Blankenship; Wade).
The electronic searches yielded a cumulative total of records (Figure 1). The Cochrane Information Specialist scrutizined the search results, eliminating duplicates and irrelevant references in accordance with the review's focused scope. Following the screening of 491 reports, we acquired 88 full-text reports for in-depth evaluation. Ultimately, we included 13 reports from 11 studies (see the table on included studies), while rejecting 69 reports from 52 studies (refer to the table on excluded studies). Our searches of clinical trial registries did not unveil any ongoing studies. We are still awaiting classification on six studies where the full reports could not be obtained (Chaine, Kianersi; Wroblewski) or were not translatable (Uehara, Yang) or did not furnish adequate information for inclusion criteria (Salman).
Due to the extended timeline of when these studies were conducted, we did not have access to the trial protocols (only Nikkhah was listed in a clinical trial registry), inhibiting our ability to determine if selective reporting posed a problem.
For two additional studies, there wasn't sufficient information to evaluate this issue (Bandello; Theodossiadis).
In the study conducted by Pahor, we noted a high attrition rate of 38% after one-month follow-up, without details indicating the groups that encountered loss to follow-up. In Han's study, the randomized number of participants matched those analyzed; however, follow-up details remained unclear. Furthermore, Han indicated that participants with adverse events post-treatment were excluded, though the count of such exclusions was not noted.
None of the studies incorporated masking for participants, personnel, or outcome assessors, categorizing them as high-risk for bias in these respective areas.
Two studies did report on allocation concealment, with Blankenship conducting allocation post-recruitment, while Nikkhah stated that the allocation sequence remained undisclosed to the investigators.
Five studies discussed adequate random sequence generation methods: Bandello and Nikkhah used computer-generated random numbers, while Blankenship, Tewari, and Wade employed coin flips. In other studies, it was indeterminate if random sequence generation had been appropriately executed. It is suggested that Bandello may have used alternate allocation methods.
Overall, the risk of bias has been summarized, with review authors reflecting on judgements for each bias item concerning the included studies.
The outcomes from various interventions have been broadly summarized and analyzed to reflect comparative effectiveness.
Effects of Interventions
For further details, please refer to Table 1 through Table 8.
Summary of Findings for Main Comparison: Nd:YAG Laser Compared to Argon Green Laser for Proliferative Diabetic Retinopathy
Patient or Population: Individuals diagnosed with proliferative diabetic retinopathy
Setting: Eye Hospital
Intervention: Nd:YAG Laser
Comparison: Argon Green Laser
Outcomes
Anticipated Absolute Effects*
Risk Associated with Argon Green Laser
Risk Associated with Nd:YAG Laser
Best Corrected Visual Acuity (BCVA): Loss of 15 or More ETDRS Letters
Study Population:
RR 0.80 (0.30 to 2.13)
20 Participants (1 RCT)
Certainty of the Evidence: Very Low 1 2 3
BCVA: Gain of 15 or More ETDRS Letters
Study Population:
RR 0.33 (0.02 to 7.32)
20 Participants (1 RCT)
Certainty of the Evidence: Very Low 1 2 3
Progression of PDR
Follow-up: 1 Year
Study Population:
RR 1.00 (0.07 to 14.95)
42 Participants (1 RCT)
Certainty of the Evidence: Very Low 1 2 4
Regression of PDR
Follow-up: 1 Year
Study Population:
RR 1.00 (0.87 to 1.14)
42 Participants (1 RCT)
Certainty of the Evidence: Very Low 1 2 5
Pain During Laser Treatment
Study Population:
RR 1.00 (0.36 to 2.76)
62 Participants (2 RCTs)
Certainty of the Evidence: Very Low 1 6
Adverse Events
Vitreous Hemorrhage, 13% in Argon Group, RR 1.22 (0.38 to 3.94); Choroidal Detachment, 19% in Argon Group, RR 0.23 (0.04 to 1.27); Neurotrophic Keratopathy, 10% in Argon Group, RR 1.29 (0.35 to 4.75).
62 Participants (2 RCT)
Certainty of the Evidence: Very Low 1 2 7
Notes: *The risk in the intervention group is based on assumed risk in the comparison group and the relative effect of the intervention.
GRADE Working Group Grades of Evidence:
High Certainty: We are very confident that the true effect lies close to that of the effect estimate.
Moderate Certainty: We are moderately confident in the effect estimate but recognize the true effect may differ significantly.
Low Certainty: Limited confidence exists regarding the effect estimate; the true effect may vary significantly.
Very Low Certainty: Very little confidence exists regarding the estimate, suggesting the true effect may differ considerably.
Summary of Findings 2: Diode Laser versus Argon Laser for Proliferative Diabetic Retinopathy
Patient or Population: Individuals with proliferative diabetic retinopathy
Setting: Eye Hospital
Intervention: Diode Laser
Comparison: Argon Laser
Outcomes
Anticipated Absolute Effects*
Risk Associated with Argon Laser
Risk Associated with Diode Laser
BCVA: Loss of 15 or More ETDRS Letters
Follow-up: 1 Year
Study Population:
RR 0.95 (0.66 to 1.36)
134 Participants (2 RCTs)
Certainty of the Evidence: Very Low 1 2 3
BCVA: Gain of 15 or More ETDRS Letters
Study Population:
RR 0.60 (0.25 to 1.45)
134 Participants (2 RCTs)
Certainty of the Evidence: Very Low 1 2 3
Progression of PDR
Study Population:
RR 0.90 (0.41 to 2.00)
66 Participants (1 RCT)
Certainty of the Evidence: Very Low 1 3 4
Regression of PDR
Study Population:
RR 0.75 (0.35 to 1.60)
66 Participants (1 RCT)
Certainty of the Evidence: Very Low 1 3 4
Pain During Laser Treatment
Study Population:
RR 3.07 (2.15 to 4.39)
228 Participants (4 RCTs)
Certainty of the Evidence: Moderate 1
Adverse Events
Vitreous hemorrhage, 15% of the Argon group. Inconsistent results were observed across studies with RR 0.50 (0.05, 4.86) and RR 1.82 (0.76, 4.35); choroidal detachment, 8% in the Argon group RR 4.00 (0.51, 31.13); neurotrophic keratopathy (0%), RR 3.00 (0.13, 67.51); maculopathy 14%, RR 1.30 (0.54, 3.13); cataract 16%, RR 0.52 (0.17, 1.57); pre-retinal membrane 5%, RR 1.16 (0.24, 5.49).
134 Participants (2 studies)
Certainty of the Evidence: Very Low 1 5
The risk in the intervention group is based on the assumed risk in the comparison group and the relative effect of the intervention.
High certainty, moderate certainty, low certainty, and very low certainty definitions as discussed previously apply.
Summary of Findings 3: Laser Exposure Duration of 0.5 Seconds versus 0.1 Second for Proliferative Diabetic Retinopathy
Patient or Population: Individuals with proliferative diabetic retinopathy
Setting: Eye Hospital
Intervention: 0.5 Second Exposure
Comparison: 0.1 Second Exposure
Outcomes
Anticipated Absolute Effects*
Risk with 0.1 Second Exposure
Risk with 0.5 Second Exposure
BCVA: Loss of 15 or More ETDRS Letters
Follow-up: 1 Year
Study Population:
RR 0.42 (0.08 to 2.04)
44 Participants (1 RCT)
Certainty of the Evidence: Low 1 2
BCVA: Gain of 15 or More ETDRS Letters
Study Population:
RR 2.22 (0.68 to 7.28)
44 Participants (1 RCT)
Certainty of the Evidence: Low 1 2
Progression of PDR
Study Population:
RR 0.33 (0.02 to 7.14)
16 Participants (1 RCT)
Certainty of the Evidence: Very Low 1 3
Regression of PDR
Study Population:
RR 1.17 (0.92 to 1.48)
32 Participants (1 RCT)
Certainty of the Evidence: Low 1 2
Pain During Laser Treatment not reported
Adverse Events
Pre-retinal or vitreous hemorrhage, 30% of the 0.1 sec group, RR 0.56 (0.18 to 1.70); macular thickening, 2 cases in 0.1 sec group; combined rhegmatous and traction retinal detachment, 1 case in the 0.5 sec group.
44 Participants (1 RCT)
Certainty of the Evidence: Very Low 1 3
The risk in the intervention group is reflective of the assumed risk associated with the comparison group and the relative effect of the intervention.
High, moderate, low, and very low certainty definitions are applicable as highlighted previously.
Summary of Findings 4: Light PRP Compared to Classic PRP for Proliferative Diabetic Retinopathy
Patient or Population: Individuals with proliferative diabetic retinopathy
Setting: Eye Hospital
Intervention: Light PRP
Comparison: Classic PRP
Outcomes
Anticipated Absolute Effects*
Risk with Classic PRP
Risk with Light PRP
BCVA: Loss of 15 Letters or More not reported
BCVA: Gain of 15 Letters or More not reported
Progression of DR not reported
Regression of PDR not reported
Pain During Laser Treatment
Study Population:
RR 0.23 (0.03 to 1.93)
65 Participants (1 RCT)
Certainty of the Evidence: Low 1 2
Adverse Events
Vitreous hemorrhage, 19% in the classic group, RR 0.07 (0.00 to 1.20); choroidal detachment (3 cases in classic group); neurotrophic keratopathy (2 cases in classic group); clinically significant macular edema (23% of the classic group, RR 0.13 (0.02 to 1.00).
65 Participants (1 RCT)
Certainty of the Evidence: Very Low 1 3
The risk in the intervention group is based on the assumed risk in the comparison group and the relative effect of the intervention. Definitions of high certainty, moderate certainty, low certainty, and very low certainty can be referenced in previous sections.
Summary of Findings 5: Mild Scatter PRP Compared to Full Scatter PRP for Proliferative Diabetic Retinopathy
Patient or Population: Individuals with proliferative diabetic retinopathy
Setting: Eye Hospital
Intervention: Mild Scatter PRP
Comparison: Full Scatter PRP
Outcomes
Anticipated Absolute Effects*
Risk with Full Scatter PRP
Risk with Mild Scatter PRP
BCVA: Loss of 15 or More ETDRS Letters not reported
BCVA: Gain of 15 or More ETDRS Letters not reported
Progression if PDR not reported
Regression of PDR not reported
Pain During Laser Treatment not reported
Vision-related QoL not reported
Adverse Events not reported
The risk in the intervention group and its corresponding 95% confidence interval are based on the assumed risk in the comparison group and the relative effect of the intervention. As outlined previously, high, moderate, low, and very low certainty definitions apply here as well.
Summary of Findings 6: Central PRP Compared to Peripheral PRP for Proliferative Diabetic Retinopathy
Patient or Population: Individuals with proliferative diabetic retinopathy
Setting: Eye Hospital
Intervention: Central PRP
Comparison: Peripheral PRP
Outcomes
Anticipated Absolute Effects*
Risk with Peripheral
Risk with Central
BCVA: Loss of 15 or More ETDRS Letters
1 Year Follow-up
Study Population:
RR 3.00 (0.67 to 13.46)
50 Participants (1 RCT)
Certainty of the Evidence: Low 1 2
BCVA: Gain of 15 or More ETDRS Letters
Study Population:
RR 0.25 (0.03 to 2.08)
50 Participants (1 RCT)
Certainty of the Evidence: Low 1 2
Progression of DR not reported
Regression of PDR not reported
Pain During Laser Treatment not reported
Vision-related QoL not reported
Adverse Events
Vitreous hemorrhage requiring additional PRP, 1 case in each group; macular traction detachment requiring pars plana vitrectomy (3 cases in central group, 1 case in peripheral group); macular thickening associated with loss of 2 or more lines of visual acuity (2 cases in each group).
50 Participants (1 RCT)
Certainty of the Evidence: Very Low 1 3
We assessed the risk in the intervention group alongside the assumed risk in the comparison group and the relative effect of intervention.
The definitions of high certainty, moderate certainty, low certainty, and very low certainty outlined in previous sections are applicable here.
Summary of Findings 7: Centre Sparing PRP Compared to Full Scatter PRP for Proliferative Diabetic Retinopathy
Patient or Population: Individuals with proliferative diabetic retinopathy
Setting: Eye Hospital
Intervention: Centre Sparing PRP
Comparison: Full Scatter PRP
Outcomes
Anticipated Absolute Effects*
Risk with Full Scatter PRP
Risk with Centre Sparing PRP
BCVA: Loss of 15 or More ETDRS Letters
Follow-up: 1 Year
Study Population:
RR 0.67 (0.30 to 1.50)
53 Participants (1 RCT)
Certainty of the Evidence: Low 1 2
BCVA: Gain of 15 or More ETDRS Letters
Follow-up: 1 Year Not reported
Regression of PDR
Follow-up: 1 Year
Study Population:
RR 0.96 (0.73 to 1.27)
53 Participants (1 RCT)
Certainty of the Evidence: Low 1 2
Pain During Laser Treatment not reported
Vision-related QoL not reported
Adverse Events not reported
The risk in the intervention group pertains to the expected risk in the comparison group along with the relative effect of the intervention as thoroughly defined earlier.
Summary of Findings 8: Extended Targeted PRP Compared to Standard PRP
Patient or Population: Individuals with diabetic retinopathy
Setting: Eye Hospital
Intervention: Extended Targeted PRP
Comparison: Standard PRP
Outcomes
Illustrative Comparative Risks*
Assumed Risk
Corresponding Risk
BCVA: Loss of 15 or More ETDRS Letters: Follow-up 1 Year
Study Population
RR 0.94 (0.70 to 1.28)
270 Participants (1 RCT)
Certainty of the Evidence: Low 1 2
BCVA: Gain of 15 or More ETDRS Letters: Follow-up 1 Year not reported
Regression of PDR: Follow-up 1 Year
Study Population
RR 1.11 (0.95 to 1.31)
270 Participants (1 RCT)
Certainty of the Evidence: Low 1 2
Pain During Laser Treatment not reported
Vision-related QoL not reported
Adverse Events none were noted during the study period. None of the eyes developed tractional retinal detachment, and there were no ocular or non-ocular adverse events linked to the study intervention detected by the investigators or reported by participants.
We designated the evidence for this comparison as low certainty for both 'BCVA Loss of 15 or More ETDRS Letters at 1 Year' and 'Regression of PDR at 1-Year Follow-up', reducing the certainty level due to high risks of imprecision associated with broad confidence intervals and high risks of performance and detection bias.
Comments
0