A Comparison of the Efficacy and Safety of Modified Baerveldt Drainage Implants versus Ahmed Glaucoma Valves After 1 Year of Follow-Up

Introduction

The Ahmed glaucoma valve (AGV, New World Medical, Inc., Rancho Cucamonga, USA) and the Baerveldt glaucoma implant (BGI, Johnson & Johnson, New Brunswick, USA) are two commonly used surgical interventions to lower intraocular pressure (IOP) and mitigate glaucomatous progression. These two devices employ different mechanisms to achieve intraocular pressure reduction, each with distinct advantages and limitations that influence clinical decision-making.

An important distinction between these tube shunts is whether they contain a valve. Non-valved tube shunts, such as the BGI, lack a built-in mechanism to restrict aqueous outflow.^1–3 To reduce the risk of postoperative hypotony, an absorbable ligation suture is commonly placed at the time of surgery, temporarily occluding the tube for approximately six to eight weeks while a semi-permeable capsule forms around the drainage plate and provides resistance to flow.¹ In contrast, the AGV is a valved implant with a built-in flow restrictor that allows for immediate IOP reduction.^1,2 As a result, AGVs typically lower IOP immediately after implantation, whereas BGIs often have a delayed pressure-lowering effect until the ligature dissolves.^1,2

Studies have shown that BGIs have greater long-term efficacy compared to AGVs at five years.^1,2 Some theories for why AGVs often have comparatively worse IOP outcomes than BGI include early exposure to inflammatory mediators and cytokines from early aqueous flow, or from mechanical activation of myofibroblasts contributing to capsular scar formation.^1,2,4–8 Given the contradictory data on the mechanisms and effects of aqueous humor on fibroblast function and plate encapsulation, further investigation into what causes decreased longevity of the AGV in comparison to the BGI is warranted.

Due to the differences in IOP-lowering dynamics between these tube shunts, surgical selection is often influenced by baseline and target IOP. AGVs are more commonly used in patients with severely elevated IOP who cannot wait six to eight weeks for pressure reduction, despite slightly higher long-term failure rates.^1,2 Conversely, BGIs are more commonly used in patients with moderately elevated IOP who can tolerate a delayed pressure-lowering effect in exchange for lower long-term failure rates and lower final IOP.^1,2

An additional factor influencing tube shunt selection is the postoperative “hypertensive phase” commonly associated with AGV implantation, characterized by an increase in IOP within the first 1–3 postoperative months that is not attributable to mechanical obstruction or device malfunction, and is thought to result from fibrous encapsulation of the plate.^9–11 This phenomenon further motivates investigation into factors influencing postoperative pressure control and long-term surgical outcomes.

A modification to the BGI, consisting of a reversible ventilating mattress suture, has recently been described.¹² This modification stents open the ventilating slit that is traditionally made proximal to the ligating suture, which enhances aqueous outflow prior to the dissolution of the ligating suture at week 6. The ventilating suture modification may therefore expand the clinical utility of non-valved tube shunts in patients requiring early postoperative pressure reduction.

In this technique, a 10–0 nylon suture is threaded through a slit in the BGI tube and anchored to the peripheral cornea to maintain slit patency and permit controlled early aqueous egress. Unlike traditional fenestration techniques, the suture provides traction to keep the fenestration slit patent and maintain outflow. It can be removed postoperatively at the slit lamp if judged clinically necessary. The ventilating effect gradually diminishes even prior to dissolution of the ligation suture.

The goal of this study was to characterize real-world outcomes of this ventilating suture modification compared with AGV implantation. We hypothesized that the vsBGI would achieve similar IOP reduction and surgical survival to the AGV through one year of follow-up.

Materials and Methods

Study Design and Population

We performed a retrospective cohort study to compare the efficacy and safety of ventilating suture-modified Baerveldt glaucoma implants (vsBGI) to unmodified Ahmed glaucoma valves (AGV) in lowering postoperative IOP. The study utilized Baerveldt 101–350 and Ahmed FP7 tube models for all respective vsBGI and AGV procedures. We included patients aged ≥18 years with glaucoma who underwent implantation of vsBGIs or AGVs by glaucoma surgeons (AEP and EAB) at the University of Iowa between 2016 and 2024. We excluded patients less than 18 years, individuals who were incarcerated, or those with less than a year of documented follow-up. The University of Iowa’s Institutional Review Board (IRB) approved this study (IRB #202409674). Informed consent was not required due to the retrospective study design, in concordance with university policy 45 CFR 46.104 (4). Ethical approval was obtained for all protocols from the University of Iowa IRB, which complies with the Declaration of Helsinki.

Data Collection

Data was collected via manual chart review. Extracted data included IOP, visual acuity (VA), number and types of medications, and postoperative adverse events at: Day 1, Week 1, Month 1, Month 3, Month 6, and Month 12.

Adverse events recorded included: hyphema, corneal blood staining, corneal edema or keratoplasty graft compromise, shallow anterior chamber (AC), choroidal hemorrhage, clinically significant hypotony (IOP ≤5 mmHg with choroidal effusion, choroidal hemorrhage, and/or hypotony maculopathy at any time point), cystoid macular edema, retinal detachment, tube erosion, endophthalmitis, diplopia, or need for additional glaucoma surgery.

Baseline pre-operative data was also collected from each patient including age, sex, race, pachymetry, vertical cup-to-disc ratio, type of glaucoma, phakic status, history of systemic steroid use, diabetes, and prior glaucoma surgeries.

Ventilating Suture Technique

The ventilating suture modification is performed after inserting the tube into the eye through a sclerostomy. A ventilating slit is made in the proximal tube by passing a 10–0 nylon suture through the tube lumen. The same 10–0 nylon is then used to make a partial thickness pass to the peripheral cornea in a horizontal mattress technique. The suture tension is adjusted to allow for the desired amount of flow through the ventilating slit and tied off. The resulting horizontal mattress suture will have a visible loop in the peripheral cornea, which is accessible postoperatively in the case too much flow is noted and it needs to be removed in clinic. The remainder of the BGI surgery technique is unchanged.

Outcomes

The primary outcomes were IOP and number of glaucoma medications at each postoperative time point. For descriptive purposes, IOP was summarized as the mean within each treatment group at each visit. However, longitudinal comparisons of IOP and medication burden between groups were performed using linear mixed-effects models to account for repeated measures over time. Postoperative medication adjustments were not protocolized and were made at the discretion of the treating clinician, reflecting real-world clinical practice in this retrospective cohort. In general, medications were reintroduced in a stepwise fashion when intraocular pressure exceeded the individualized target range. Medication count was defined as the total number of active glaucoma medications at each time point, including both topical agents and oral carbonic anhydrase inhibitors. Perioperative medications not intended for long-term IOP control, such as topical steroids and antibiotics, were not included in the medication count. Secondary outcomes included the probability of reaching target IOP within one year, visual acuity, CDR, and probability of survival. Three definitions of surgical success were applied based on IOP thresholds (≤21 mmHg, ≤17 mmHg, and ≤14 mmHg), defined as liberal, intermediate, and conservative, respectively. Adequate IOP reduction criteria must have been present on at least 2 consecutive visits on or after 3 months to qualify as a success. In addition to the IOP criteria, success was defined as not having a serious IOP-related or tube shunt-related adverse event (defined as endophthalmitis, retinal detachment, choroidal hemorrhage), undergoing a reoperation for glaucoma (defined as cyclophotocoagulation, AGV, BGI, trabeculectomy, or tube revision), or loss of light perception vision.

Statistical Analysis

Extracted data comparing vsBGI to AGV included IOP, visual acuity (logMAR), number of glaucoma medications, CDR, and adverse events. Data normality was assessed using the Shapiro–Wilk test, which informed the choice of statistical methods. Preoperative IOP, logMAR visual acuity, CDR, and medication count were non-normally distributed based on Shapiro–Wilk testing and data visualization; therefore, Wilcoxon rank-sum tests were used to compare baseline differences between Ahmed and Modified BGI before longitudinal modeling. Categorical variables were assessed with chi-square tests when expected frequencies were sufficient, and Fisher’s exact tests were applied when the expected counts were < 5. Glaucoma types were compared across AGV and vsBGI groups using Fisher’s exact tests with Holm’s method to adjust for multiple comparisons.

A linear mixed-effects model was used to evaluate longitudinal changes in IOP between Ahmed and Modified BGI surgical groups, accounting for repeated measures within subjects. Fixed effects included timepoint (modeled as a categorical variable spanning preoperative through postoperative month 12), surgical group, and the interaction between timepoint and surgical group to assess differential trends over time. A random intercept was included for each eye (study ID) to account for within-subject correlation. The reference categories were preoperative IOP and the Ahmed group; thus, all estimated effects are interpreted relative to baseline values in the Ahmed group. Missing data were treated as missing at random, and all available observations were included using maximum likelihood estimation within the mixed-effects framework. Each model was adjusted for the corresponding baseline value (preoperative IOP for the IOP model, baseline logMAR for the visual acuity model, and baseline CDR), as well as diabetes and steroid use.

A Poisson mixed-effects model was used to evaluate longitudinal medication burden over time. Covariates were selected a priori based on clinical relevance and potential confounding. Specifically, diabetes status and steroid use were included due to their known associations with postoperative IOP control and surgical outcomes, while implant quadrant was included to account for potential anatomic and surgical variability that may influence device performance.

Kaplan-Meier survival curves were created to estimate and compare the probability of surgical success over time between the AGV and vsBGI groups. Surgical success was evaluated using three IOP thresholds (≤21, ≤17, and ≤14 mmHg), corresponding to liberal, intermediate, and conservative criteria, respectively, in addition to no loss of light perception and no repeat glaucoma operation. For the Kaplan–Meier analyses, no additional covariates were included, as these models were used to estimate unadjusted time-to-failure outcomes.

A Kaplan-Meier survival curve was also created comparing the probability of reaching target IOP between the vsBGI and AGV groups. Target IOP was defined pre-operatively by the operating physician based on multiple factors including baseline IOP, severity of glaucoma, rate of progression, and optic nerve appearance. Group comparisons were performed using Cox proportional hazards regression, and proportional hazards assumptions were evaluated using Schoenfeld residuals.

As a sensitivity analysis to address baseline imbalance and non-randomized treatment selection, we performed propensity score matching using baseline covariates selected a priori on clinical grounds, including preoperative IOP, glaucoma subtype, lens status, baseline logMAR visual acuity, diabetes, steroid use, prior glaucoma surgery, and year of surgery. Matched cohorts were assessed using standardized mean differences, with values <0.1 considered well-balanced. Longitudinal outcome analyses were then repeated in the matched cohort.

For all statistical tests performed, a p-value < 0.05 was considered statistically significant, with 95% confidence intervals applied to all predicted outcomes. All analyses were performed using R version 4.5.0 (2025–04-11) (R Foundation for Statistical Computing, Vienna, Austria).

Results

A total of 124 eyes were included, with 62 eyes in each group. A participant flow diagram summarizing patient inclusion, exclusions, and follow-up at each timepoint is shown in Supplementary Figure 1. Baseline characteristics differed between groups in several clinically relevant respects (Table 1). Compared with the vsBGI group, the AGV group had higher preoperative IOP, a higher prevalence of diabetes and steroid use, worse preoperative visual acuity, and a different distribution of implant quadrant; baseline cup-to-disc ratio was higher in the vsBGI group. These differences likely reflect non-randomized clinical selection patterns and were accounted for in adjusted analyses.

Table 1 Baseline Characteristics

Descriptively, median [IQR] IOP at baseline was 33.5 [27.2, 45.0] mmHg for the AGV group and 25 [19.0, 31.8] mmHg in the vsBGI group. At 12 months, both groups achieved similar median IOP values (AGV: 13.0 mmHg [10.2, 17.0]; vsBGI: 13.0 mmHg [9.0, 17.0]). In adjusted linear mixed-effects modeling, both groups demonstrated substantial postoperative IOP reduction over one year. Despite higher baseline IOP in the AGV group, the two groups showed similar IOP values on postoperative Day 1. There was a slight IOP rebound at Month 1 in both groups, but this returned to a lower post-surgical baseline by Month 12. The only statistically significant difference in IOP between groups occurred at Month 6, when the AGV group had an estimated 3.0 mmHg higher mean IOP compared to the vsBGI group (p = 0.036) (Figure 1).

Figure 1 Mean IOP over time by surgery type.

After adjusting for baseline visual acuity, both the AGV and vsBGI groups showed significant improvement in logMAR visual acuity over time. AGV eyes had a significantly better logMAR visual acuity at Month 1 (p = 0.002) and Month 6 (p = 0.031) compared to vsBGI. Differences at other timepoints were not statistically significant (Figure 2).

Figure 2 Estimated best corrected visual acuity over time.

We found the CDR to be consistently higher at baseline in the vsBGI group. This difference persisted descriptively throughout follow-up, although pairwise contrasts from the adjusted model were not statistically significant (Figure 3).

Figure 3 Cup:Disc ratio over time by surgery type.

Using an adjusted Poisson mixed-effects model, medication number was assessed postoperatively up to one year. Pairwise contrasts from the adjusted model indicated that the AGV required significantly more medications on Day 1 (estimate = 1.57, p < 0.001), but fewer on Week 1 (estimate = −0.49, p = 0.031). After Month 1, the two trajectories aligned with no significant differences at later timepoints (Figure 4).

Figure 4 Estimated number of glaucoma medications over time.

Reoperation-free survival was higher in the vsBGI group compared to the AGV group, with divergence in the Kaplan-Meier curves beginning around Month 4 and continuing through Month 12. Cox proportional hazards regression showed no significant differences in reoperation-free survival between groups (p = 0.278).

Success of IOP lowering was assessed using Kaplan-Meier survival curves across three pre-defined IOP thresholds (≤21, ≤17, and ≤14 mmHg). Under the liberal criterion (IOP ≤21 mmHg), there were no significant differences between groups (Cox p = 0.337, Figure 5). Under the intermediate criterion (IOP ≤17 mmHg), the vsBGI group demonstrated higher observed survival probability (Cox p = 0.037, Figure 6). Under the conservative criterion (IOP ≤14 mmHg), no significant difference was observed (p = 0.072, Figure 7). This finding at the ≤17 mmHg threshold was not consistent across other thresholds and is therefore interpreted with caution.

Figure 5 Kaplan-Meier Curve: Liberal Criteria. Kaplan-Meier survival curves of the cumulative probability of success defining adequate IOP reduction as IOP ≤21 mmHg. Inadequate IOP reduction criteria must have been present on at least 2 consecutive visits on or after 3 months and loss of light perception at any point to qualify as failure.

Figure 6 Kaplan-Meier Curve: Intermediate Criteria. Kaplan-Meier survival curves of the cumulative probability of success defining adequate IOP reduction as IOP ≤17 mmHg. Inadequate IOP reduction criteria must have been present on at least 2 consecutive visits on or after 3 months and loss of light perception at any point to qualify as failure.

Figure 7 Kaplan-Meier Curve: Conservative Criteria. Kaplan-Meier survival curves of the cumulative probability of success defining adequate IOP reduction as IOP ≤14 mmHg. Inadequate IOP reduction criteria must have been present on at least 2 consecutive visits on or after 3 months and loss of light perception at any point to qualify as failure.

A Kaplan-Meier curve was also used to assess success in reaching target IOP between the two groups postoperatively. Although the Kaplan-Meier curve using IOP target as survival criteria numerically favored the vsBGI group, it was not statistically significant (Cox p = 0.970) (Figure 8).

Figure 8 Kaplan-Meier Curve: Target IOP Criteria. Kaplan-Meier survival curve comparing probability of reaching target IOP between the Ahmed and Modified BGI groups. Target IOP was defined pre-operatively by the operating physician.

Postoperative complications were generally infrequent and comparable, with no statistical significance in adverse events between groups. (Figure 9) The most common adverse event in both groups was hyphema or heme in the anterior chamber, occurring in 14% of AGV eyes and 9% of vsBGI eyes. Corneal edema or worsening of a corneal graft was observed in 9.7% of AGV eyes and 8.1% of vsBGI eyes. Cystoid macular edema occurred more frequently in the vsBGI group (9.7%) compared to the AGV group (3.2%). Clinically significant hypotony was noted in 4.8% of vsBGI eyes and 3.2% of AGV eyes. Complications of shallow anterior chamber, choroidal hemorrhage, and diplopia were infrequent in both groups. There were no cases of retinal detachment, tube erosion, or endophthalmitis in either group. The need for additional glaucoma surgery occurred more frequently in the AGV group, with 9% of AGV eyes versus 5% of vsBGI eyes experiencing a reoperation.

Figure 9 Adverse events within 12 months by surgery type.

In a propensity score–matched sensitivity analysis, 16 eyes in each group were matched on clinically selected baseline covariates. Covariate balance improved substantially after matching, although residual imbalance remained for several variables, including surgery year, lens status, and baseline visual acuity, POAG/pseudoexfoliation glaucoma subtype, and diabetes status (Supplemental Figure 2, Supplemental Tables 1 and 2). In the matched cohort, longitudinal IOP, medication burden, visual acuity, and cup-to-disc ratio findings were overall consistent with the primary analyses, with no meaningful new between-group differences identified (Supplementary Tables 3–6).

Discussion

Both the AGV and vsBGI achieved substantial and sustained IOP reduction with broadly comparable safety profiles over one year. Although isolated differences were observed at select timepoints and under one intermediate IOP-based survival definition, overall outcomes between groups were similar.

The rationale for this modification lies in the limitations of traditional ventilating slit techniques. Experimental studies have demonstrated that slit characteristics can influence aqueous outflow dynamics; however, clinical studies suggest that slit-based flow modulation may not reliably provide early postoperative pressure control in vivo, likely due to biological factors such as fibrosis and wound healing.^13–17 Findings from the Ahmed Versus Baerveldt (AVB) Study (2011) and Ahmed Baerveldt Comparison (ABC) studies similarly demonstrated less effective early postoperative IOP reduction with Baerveldt implants compared with Ahmed valves during the first postoperative month.^1,2 These limitations support investigation of alternative strategies for improving early pressure control in non-valved glaucoma drainage devices.

The concept of modifying non-valved shunts with a traction or ventilating suture has previously been described by Akil et al, who demonstrated the feasibility of a reversible ventilating suture for controlling early aqueous flow.¹² Our study builds upon this work by extending follow-up to one year and including a larger comparative cohort, providing additional assessment of the safety and efficacy of this modification during the early and intermediate postoperative period.

Despite the AGV and vsBGI groups being broadly comparable demographically, the AGV group had higher baseline IOP. Baseline differences between groups, including higher preoperative IOP and variation in glaucoma subtype distribution, may introduce residual confounding despite statistical adjustment and should be considered when interpreting comparative outcomes. Findings from the propensity score-matched sensitivity analysis were broadly consistent with the primary analyses, although residual imbalance remained for several covariates and the matched cohort size was limited. The discrepancy in baseline IOP between groups likely reflects the retrospective nature of the study and the practice patterns of the operating surgeons. Specifically, AGVs were more frequently selected in cases of markedly elevated preoperative IOP, consistent with prior survey data demonstrating that many glaucoma surgeons preferentially select valved implants in high-IOP clinical scenarios.¹⁸

Baseline cup-to-disc ratio was significantly higher in the vsBGI group compared to the AGV group (Wilcoxon p < 0.001), although adjusted longitudinal analyses showed no significant postoperative differences between groups. This baseline difference may reflect underlying differences in disease chronicity and surgical selection patterns inherent to the retrospective study design.

Analysis of postoperative IOP revealed no significant differences between the AGV and vsBGI groups over one year, except at postoperative Month 6, where the vsBGI group demonstrated an estimated 2.9 mmHg lower mean IOP compared to the AGV group (p = 0.036). These findings are consistent with broadly comparable postoperative IOP control between the two devices. Although the AGV is commonly preferred for clinical situations of high IOP, our results suggest that it would also be reasonable to consider the vsBGI for these clinical scenarios.¹⁸ Prospective studies are needed to equalize baseline IOP characteristics between the two groups, allowing one to better understand the efficacy and limitations of the vsBGI.

Both AGV and vsBGI implants showed improvements in visual acuity over the first postoperative year. Eyes receiving AGV recovered more lines of visual acuity in the early and intermediate postoperative period compared to vsBGI, with clinically and statistically significant advantages at Month 1 and Month 6. The early visual acuity advantage observed in the AGV group may reflect differences in early postoperative IOP control and recovery dynamics between devices, which could have short-term functional implications despite similar long-term outcomes. However, these differences diminished by Month 12, suggesting that the early benefit in the AGV group may attenuate over time. While the visual acuity data from this cohort did not reveal statistically significant differences in postoperative complications between groups, there are a variety of confounding influences on this real-world retrospective data.

The finding that AGV eyes required significantly more medications than vsBGI eyes on postoperative Day 1 is consistent with clinical expectations. This trend aligns with prior practice-defining literature, notably that of Pakravan et al (2014), in which a temporarily aggressive IOP target in the postoperative phase following AGV implantation led to better IOP outcomes than a control group.⁷ These findings have influenced clinical management strategies and likely contributed to the observed postoperative medication patterns in our cohort. However, after postoperative month one, the trajectories of medication use between the two groups converged, with no significant differences observed at later time points. Both groups demonstrated similar long-term medication burden after Month 1.

The overall Kaplan-Meier survival curve for reoperation showed no statistically significant difference between the vsBGI and AGV groups within the first year. However, this limited time frame likely underestimates the long-term divergence expected between these two interventions. Prior studies, including the AVB Study and the ABC Study, have demonstrated that the BGI tends to show superior long-term outcomes, with greater durability of IOP control and lower rates of surgical failure at five years.^1,2 Based on both clinical experience and existing long-term data, the gap in success rates tends to widen over time, with the BGI often showing improved durability beyond the first postoperative year. However, the duration of analysis is one limitation of our current study. Given the limited long-term postoperative data on this modification beyond one year, our cohort should be followed longitudinally to assess 5-year reoperation-free survival and confirm durability of vsBGI results.

The Kaplan–Meier survival analysis by three different IOP-based definitions highlights how outcome definitions influence interpretation of surgical success. Using the liberal criterion (IOP ≤21 mmHg), AGV and vsBGI performed similarly, suggesting both are adequate for modest postoperative targets. However, under the intermediate criterion (IOP ≤17 mmHg), vsBGI demonstrated higher observed survival probability. At the most conservative definition (IOP ≤14 mmHg), the difference between groups did not reach significance, likely reflecting the difficulty of consistently achieving very low pressures across both procedures. Overall, these findings support broadly comparable IOP outcomes between devices. While a statistically significant difference was observed at the ≤17 mmHg threshold, this finding should be interpreted with caution, as no significant differences were observed at the liberal or conservative thresholds. This may indicate a potential advantage of the ventilating suture modification in achieving intermediate levels of IOP control; however, the overall similarity in outcomes across other measures tempers this interpretation. Additionally, baseline differences between groups, including preoperative IOP, may have influenced these results and should be considered when interpreting comparative outcomes.

When survival was instead defined by achievement of individualized target IOP, the Kaplan–Meier analysis revealed no significant difference between groups, despite numerically higher survival in the vsBGI group (Cox p = 0.970). This likely reflects differences in physician-selected target IOPs and baseline surgical selection patterns between groups, which may limit direct between-group comparison.¹⁸ While this highlights a degree of selection bias, it also emphasizes the need for a future randomized controlled trial to eliminate such confounders and more definitively compare the efficacy of these two surgical approaches.

Postoperative complication rates were low and largely comparable between the AGV and vsBGI groups, suggesting comparable safety profiles between groups. (Figure 9) While hyphema was the most common complication in both groups (14% in the AGV group and 9% in the vsBGI group), its slightly higher incidence in the AGV group did not appear to impact longer-term outcomes. Notably, cystoid macular edema occurred more frequently in the vsBGI group, warranting further exploration in future studies. Other complications such as clinically significant hypotony, corneal edema, and diplopia occurred at low rates without significant group differences. We defined complications using the precedent of the Tube Versus Trabeculectomy studies, with some modifications.^16,17,19,20 Clinically significant hypotony was defined as an IOP of ≤5 mmHg accompanied by choroidal effusion, choroidal hemorrhage, and/or hypotony maculopathy at any postoperative time point. This definition was chosen to capture cases of hypotony associated with visually significant or vision-threatening sequelae, distinguishing them from transient, clinically benign postoperative IOP reductions.²¹

Importantly, there were no cases of retinal detachment, tube erosion, or endophthalmitis in either cohort. The need for additional glaucoma surgery was higher in the AGV group, which may reflect differences in IOP control durability between devices. Overall, the adverse event profile suggests that vsBGI may be a safe and effective alternative to AGV in select clinical settings. Based on these findings, vsBGI may be a reasonable option in cases requiring both substantial pressure reduction and relatively low target IOP. In contrast, the more stable early postoperative course and early visual acuity advantage observed with AGV may be beneficial in patients where early visual recovery is a priority or where minimizing early IOP fluctuations is desired. Given the overall similarity in outcomes across most measures, these results support a patient-centered approach in which surgeons weigh early postoperative stability against the potential for achieving lower intermediate IOP targets when selecting between devices.

This study has several limitations. First, the retrospective study design introduces the potential for selection bias and residual confounding. In particular, baseline intraocular pressure differed between groups, which may have influenced comparative outcomes despite statistical adjustment. This imbalance likely reflects physician selection patterns, such as preferential use of AGV in patients with higher preoperative IOP.

Second, this was a single-institution study with a relatively small sample size, which may limit the generalizability of these findings. While the study population is representative of the local Iowan patient population, it may not reflect broader national demographics. The sample size was also constrained by the relative novelty of the ventilating suture modification, though this is expected to improve as more patients undergo the procedure.

Third, the follow-up duration of one year is sufficient to evaluate early and intermediate postoperative outcomes but may be insufficient to assess long-term device performance and durability. Prior studies, including the AVB and ABC trials, have demonstrated divergence in outcomes between AGV and BGI at five years postoperative. As such, extended follow-up of this cohort to at least five years would be important to better characterize the long-term efficacy and durability of this modification.

Taken together, these limitations highlight the need for prospective, randomized controlled studies with longer follow-up. Given the comparable adverse event rates observed between groups in this study, such trials would be both feasible and ethically appropriate, and would allow for more rigorous assessment of the efficacy of the ventilating suture modification while mitigating physician bias.

Conclusion

The vsBGI appears to be a comparable alternative to the AGV for early postoperative IOP reduction, with similar overall IOP control, medication burden, visual acuity, adverse events, and achievement of target IOP at one year. Longer-term follow-up will be important to further characterize comparative durability.