The Paul Glaucoma Implant versus the Ahmed and Baerveldt Glaucoma Implants: A Systematic Review and Bayesian Network...

To compare the efficacy and safety of the Paul Glaucoma Implant (PGI) with those of the two established glaucoma drainage devices, the Ahmed Glaucoma Valve (AGV) and the Baerveldt Glaucoma Implant (BGI).

PubMed, Embase, Scopus, Web of Science, Cochrane CENTRAL, ClinicalTrials.gov, ScienceDirect and Google Scholar were searched. Eligible studies were randomised or comparative observational studies reporting outcomes in patients who received the PGI alongside a concurrent AGV or BGI comparator. The primary outcomes were intraocular pressure (IOP) reduction, surgical success rate, and reduction in glaucoma medications. A Bayesian network meta-analysis was performed, with treatment ranking summarized by surface under the cumulative ranking curve (SUCRA). Quality assessment and evidence certainty was performed using the Cochrane Risk of Bias tool and GRADE framework respectively.

Five comparative studies (238 patients) were included. Relative to the AGV, the PGI was associated with a mean difference in IOP at 12 months of -3.344 (95% CrI -7.743 to 1.058) mmHg, and relative to the BGI of 4.387 (95% CrI -0.288 to 9.074) mmHg. Corresponding differences in medication number at 12 months were]-0.186 (95% CrI -0.858 to 0.545) and 0.444 (95% CrI -0.358 to 1.272). Credible intervals crossed the null for 29 of the 30 primary contrasts. Ranking probabilities favoured the AGV for IOP reduction from 3 to 12 months (SUCRA 0.885 to 0.976) and the PGI for early medication reduction (SUCRA 0.769 at 1 month), but these rankings were derived from imprecise estimates. Safety estimates were based on small numbers of events and several were unstable. Certainty of evidence was low to very low for most comparisons.

Within the limits of a small and predominantly non-randomised evidence base, the PGI achieved intraocular pressure and medication reduction broadly comparable to those of the AGV and BGI over the first postoperative year, with a distinct complication profile. The rankings reported here are hypothesis-generating and should not by themselves determine device selection. Adequately large randomised comparisons with follow-up beyond one year are required.