Postoperative Adverse Events, Interventions, and the Utility of Routine Follow-Up After 23-, 25-, and 27-Gauge Pars Plana Vitrectomy

Transcription

1 ORIGINAL CLINICAL STUDY Postoperative Adverse Events, Interventions, and the Utility of Routine Follow-Up After 23-, 25-, and 27-Gauge Pars Plana Vitrectomy Ryan A. Shields, MD, Cassie A. Ludwig, MD, MS, Matthew A. Powers, MD, MBA, Elaine M.T. Tran, BA, Stephen J. Smith, MD, and Darius M. Moshfeghi, MD Purpose: To evaluate the utility of standard postoperative visit (POV) intervals in pars plana vitrectomy (PPV) as a function of adverse events (AEs) identified. Design: Retrospective case review. Methods: The medical records of all patients undergoing 23-, 25-, and 27-gauge PPV from January 1, 2016 to December 31, 2016 were reviewed. Each POV was assessed as a standard (s-pov), physicianadjusted (a-pov), or patient-initiated visit (p-pov). Preoperative features, diagnoses, and surgical procedures were evaluated to determine protective and risk factors for AEs. Results: A total of 256 patients (310 PPVs) were included in this study. The most common cumulative postoperative AEs were elevated intraocular pressure (>30 mm Hg) (12.3%), cystoid macular edema (6.1%), and retinal detachment (5.8%). Patients with the diagnosis of macular hole or epiretinal membrane had the lowest relative risk of AEs [0.30; 95% confidence interval (CI), and 0.36; 95% CI, , respectively]. There was no difference in time to AE among different vitrectomy gauge sizes (P = 0.733). Patients in a-pov and p-pov groups had a statistically significant higher incidence of AEs in the POV day 5 10 window (P = 0.004). Conclusions: The utility of standard POVs in detecting AEs is dependent on the indication for PPV. Specifically patients undergoing isolated macular surgery (epiretinal membrane peel or macular hole repair) had the lowest relative risk of postoperative AEs and may warrant a less-intensive follow-up regimen. (POD1) visit, has been called into question. 1 3 The current standard of care after uncomplicated vitreoretinal surgery is POD1, week 1 (POW1), month 1 (POM1), and month 3 standard (s-pov) examinations. One of the primary goals of the POD1 visit is to assess for elevated intraocular pressure (IOP). 4 6 Additional early adverse events (AEs) can also be detected, including hypotony, choroidal effusion or hemorrhage, retinal detachment (RD), and endophthalmitis all of which are vision threatening if left untreated. 7 Recent literature indicates that the intervention rate after vitrectomy in the immediate postoperative period ranges from 0.7% to 3.7%, which is similar to the POD1 intervention rate after routine cataract surgery (2.8%). 1 3,8 Previous studies have reported postoperative outcomes on 20- and 23-gauge PPV, with fewer studies reporting postoperative outcomes of 25- and 27-gauge PPV We hypothesized that the later POVs months 1 to 3 were an anachronism from an earlier time period characterized by larger-gauge surgery, longer surgery times, and inadequate visualization of the peripheral retina In light of the improved surgical technique, we would expect a decreased AE rate at later POVs that would be both clinically and statistically significant for patients undergoing PPV. Based on these factors, we conducted a retrospective chart review of patients undergoing 23-, 25-, and 27-gauge PPV to evaluate the incidence of postoperative AEs and their treatment in an attempt to find a subset of patients who may not require later POVs. Key Words: adverse events, follow-up, macular surgery, vitrectomy, vitrectomy gauge (Asia Pac J Ophthalmology (Phila) 2019;0:0 0) With continued improvement in surgical and anesthetic techniques for pars plana vitrectomy (PPV), the utility of postoperative visits (POVs), especially the postoperative day 1 From the Byers Eye Institute, Department of Ophthalmology, Stanford University School of Medicine, Palo Alto, CA, United States. Submitted September 3, 2018; accepted October 22, Supported by an unrestricted departmental grant from the National Eye Institute (P ) and Research to Prevent Blindness, Inc. Neither of these played a role in conducting the study or the preparation of the manuscript. The authors declare no competing financial interests. This study was presented at the Association for Research in Vision and Ophthalmology Annual Meeting; April 29 to May 3, 2018; Honolulu, Hawaii. Reprints: Darius M. Moshfeghi, Horngren Family Vitreoretinal Center Byers Eye Institute, Department of Ophthalmology, Stanford University School of Medicine, 2452 Watson Ct, Palo Alto, CA, 94304, United States. dariusm@stanford.edu. Copyright 2019 by Asia-Pacific Academy of Ophthalmology ISSN: DOI: /APO METHODS This study was conducted in compliance with the Health Insurance Portability and Accountability Act, and approval was obtained from the Institutional Review Board of the Stanford Hospital s subcommittees for the protection of human subjects. A retrospective cohort study was conducted on patients who underwent PPV at the Byers Eye Institute of Stanford University School of Medicine between January 1, 2016 and December 31, Patients who underwent PPV were identified using CPT, ICD-9, and ICD-10 codes. Exclusion criteria included patients younger than 18 years, patients with incomplete records (including patients with less than 2 POVs), and patients receiving nonvitreoretinal ocular surgery that could confound follow-up results. To allow the maximum inclusion of AEs, any finding that was not seen after routine PPV or required monitoring was considered abnormal and recorded as an AE even if no intervention was undertaken. Patients were categorized into 3 groups. Group 1 included any patient that did not deviate from the traditional s-pov even if an AE occurred; group 2 included those who deviated from s-pov and adjusted based on physician discretion 1

2 Shields et al (a-pov); and group 3 included those who deviated from s-pov based on patient discretion (p-pov). In this study, an IOP above 30 mm Hg was considered a significant AE. Hypotony was defined as an IOP less than 5 mm Hg. In cases without recorded IOP, a clinical diagnosis of hypotony was considered sufficient for inclusion. Adverse events were counted once per surgical follow-up period, even if their presence was noted at sequential visits. Patient data were compiled and analyzed using Statistical Analysis Software Enterprise Guide Version 7.1 (SAS Institute, Cary, NC, US). Relative risks with corresponding 95% confidence intervals were obtained to determine the risk of AEs associated with demographic factors, preoperative diagnoses, and surgeries performed. Kaplan-Meier time-to-event analyses were used to assess the time to AEs between exposure variables. Statistical tests were performed using the appropriate parametric (χ 2 ) and nonparametric (Fisher exact, McNemar exact) tests. P values less than 0.05 were considered significant. RESULTS Study Cohort A total of 283 patients underwent PPV between January 1, 2016 and December 31, Fifteen were excluded due to insufficient records, 11 owing to younger than 18 years, and 1 due to concurrent primary ruptured globe exploration. A total of 310 PPVs in 256 patients were included in this analysis. Baseline Characteristics Baseline characteristics are listed in Table 1. Approximately TABLE 1. Baseline Characteristics Characteristic No. (%) of Patients* No. (%) of Patients with AE RR of Any AE (95% CI) Age (median, IQR), y 63 (52 71) > (41.0) 41 (32.3) 0.70 ( ) (59.0) 84 (45.9) 1.00 (ref) Sex Male 167 (53.9) 70 (41.9) 1.06 ( ) Female 143 (46.1) 56 (39.2) 1.00 (ref) Prior retinal surgery Yes 111 (35.8) 54 (48.6) 1.36 ( ) No 199 (64.2) 71 (35.7) 1.00 (ref) If yes, what was the prior retinal surgery Pars plana vitrectomy 66 (21.3) 32 (48.5) 1.36 ( ) Scleral buckle 6 (1.9) 3 (50.0) 1.39 ( ) Both 39 (12.6) 19 (48.7) 1.37 ( ) Preoperative BCVA (Snellen) 20/20 to 20/ (55.5) 59 (34.3) 1.00 (ref) 20/125 to 20/ (17.7) 29 (52.7) 1.56 ( ) Counting fingers or hand motion 62 (20.0) 25 (40.3) 1.20 ( ) Light perception 17 (5.5) 9 (52.9) 1.57 ( ) Unable 4 (1.3) 4 (100) 2.97 ( ) Preoperative IOP (any method), mm Hg (10.0) 16 (51.6) 1.44 ( ) (75.8) 85 (36.2) 1.00 (ref) (7.7) 12 (50.0) 1.40 ( ) >30 8 (2.6) 5 (62.5) 1.74 ( ) Unavailable 12 (3.9) 8 (66.7) 1.87 ( ) PVD status PVD 114 (36.8) 38 (33.3) 1.00 (ref) No PVD 60 (19.4) 18 (30.0) 0.89 ( ) No view or unknown 31 (10.0) 17 (54.8) 1.63 ( ) Vitrectomized 105 (33.9) 51 (48.6) 1.44 ( ) Lens status Phakic 157 (50.6) 55 (35.0) 1.00 (ref) Pseudophakic 138 (44.5) 58 (42.0) 1.22 ( ) Aphakic 15 (4.8) 13 (86.7) 2.52 ( ) AE indicates adverse event; BCVA, best-corrected visual acuity; CI, confidence interval; IOP, intraocular pressure; IQR, interquartile range; PVD, posterior vitreous detachment; RR, relative risk. *Unless otherwise specified. Patients without prior retinal surgery used as a reference for each prior surgery type. Not recorded, or soft to palpation. Because of rounding, percentages may not total

3 Adverse Events After Vitrectomy 54% of the surgeries were performed on male patients. The median age at the time of surgery was 63 years (interquartile range, years). Preoperative best-corrected visual acuity and IOP are shown in Table 1. Only 2.6% of patients had a preoperative IOP greater than 30 mm Hg. Over a third of patients (35.8%) had undergone prior retinal surgery including PPV (33.9%) and/ or scleral buckle (SB) (14.5%). Over half of patients were phakic (50.6%), followed by pseudophakic (44.5%), and aphakic (4.8%). With regard to posterior vitreous detachment (PVD) status, 36.8% had a documented PVD. Ten percent of patients had no view from a vitreous hemorrhage (VH) and/or no documentation of PVD status. Preoperative Diagnoses and Surgical Procedures The most common indications for vitrectomy were RD (99, 31.9%), epiretinal membrane (ERM) (65, 21.0%), and VH (53, 17.1%). Glaucoma was present in 47 (15.2%) of patients, with 13 of those having previously undergone prior glaucoma surgery. Ten eyes had active endophthalmitis, and 1 eye had a history of treated endophthalmitis. Ten eyes had a diagnosis of posterior uveitis, which included idiopathic vitritis, panuveitis, progressive outer retinal necrosis, toxoplasmosis chorioretinitis, and a history of acute retinal necrosis. Six eyes had a diagnosis of von Hippel Lindau syndrome. The remaining preoperative diagnoses are listed in Table 2. Of the 310 PPVs, 85.2% were 25-gauge, 11.9% were 27-gauge, and 2.9% were 23-gauge. One case used a combination of 25- and 27-gauge vitrectomy. Combined PPV and SB was performed in 30 (9.7%) eyes. Concurrent phacoemulsification with intraocular lens (IOL) insertion was performed in 26 (8.4%) eyes. Twenty patients underwent manipulation of IOL which included explantation, repositioning, exchange, or secondary insertion of an anterior chamber (AC) IOL or posterior chamber IOL. The complete list of surgical procedures performed is shown in Table 3. Postoperative Adverse Events and Interventions A total of 1203 POVs were reviewed. On average, a patient was seen 3.9 (median, 4) times in the first 3 months (range, 2 12 visits). The most common postoperative AEs in descending order of frequency were elevated IOP (12.3%), cystoid macular edema (CME) (6.1%), RD (5.8%), VH (5.2%), and hypotony (4.8%). All other AEs occurred at a frequency of less than 3% (Table 4). The most frequent type of AEs varied depending on the time after surgery. A total of 34 (10.9%) vitrectomies required repeated operation with the most common being for RD. Patients in group 1 (s-pov) had an overall lower rate of AEs (61/212, 28.8%) compared with group 2 (a-pov) (43/68, 63.2%, P < 0.001) and group 3 (p-pov) (21/30, 70.0%, P < 0.001) (Fig. 1). There was no statistically significant difference in the overall rate of AEs when comparing group 2 and group 3. There was a statistically significant difference (P = 0.004) in the AEs between the groups only at the POD5 10 window (Fig. 1). Adverse events TABLE 2. Adverse Events by Preoperative Diagnoses in the Operative Eye Preoperative Diagnosis No. (%) of Patients No. (%) of Patients with AE RR (95% CI) Retinal detachment 99 (31.9) 51 (51.5) 1.47 ( ) Macula-sparing RRD* 32 (10.3) 17 (53.1) 1.37 ( ) Macula-involving RRD* 39 (12.6) 22 (56.4) 1.48 ( ) Macula-sparing TRD 13 (4.2) 5 (38.5) 0.95 ( ) Macula-involving TRD 25 (8.1) 11 (44.0) 1.10 ( ) Cystoid macular edema 79 (25.5) 29 (36.7) 0.88 ( ) ERM/macular pucker 65 (21.0) 11 (16.9) 0.36 ( ) Vitreous hemorrhage 53 (17.1) 22 (41.5) 1.04 ( ) Glaucoma 47 (15.2) 23 (48.9) 1.26 ( ) Retained silicone oil or silicone oil 46 (14.8) 23 (50.0) 1.29 ( ) emulsification Proliferative diabetic retinopathy 46 (14.8) 18 (39.1) 0.97 ( ) Proliferative vitreoretinopathy 43 (13.9) 24 (55.8) 1.48 ( ) Macular hole 31 (10.0) 4 (12.9) 0.30 ( ) High myopia 20 (6.5) 8 (40.0) 0.99 ( ) Endophthalmitis 11 (3.5) 5 (45.5) 1.13 ( ) Posterior uveitis/vitritis 10 (3.2) 1 (10.0) 0.24 ( ) Ruptured globe sequelae 8 (2.6) 6 (75.0) 1.90 ( ) Intraocular lens subluxation or dislocation 8 (2.6) 6 (75.0) 1.90 ( ) Retained lens material or dropped lens 6 (1.9) 4 (66.7) 1.67 ( ) von Hippel Lindau syndrome 6 (1.9) 6 (100) 2.55 ( ) Vitreous opacity/floater 3 (1.0) 0 (0) 0.31 ( ) Choroidal detachment 2 (0.6) 2 (100) 2.50 ( ) Hypotony 2 (0.6) 2 (100) 2.50 ( ) Ocular tumor 1 (0.3) 0 (0) 0.62 ( ) AE indicates adverse events, CI, confidence interval; ERM, epiretinal membrane; RR, relative risk; RRD, rhegmatogenous retinal detachment; TRD, tractional retinal detachment. *Two patients with giant retinal tear. 3

4 Shields et al TABLE 3. Adverse Events by Intraoperative Factors Intraoperative Factor No. (%) of Patients No. (%) of Patients with AE RR (95% CI) Surgical system used 23-gauge 9 (2.9) 5 (55.6) 1.42 ( ) 25-gauge* 264 (85.2) 103 (39.0) (ref) 27-gauge 37 (11.9) 17 (45.9) 1.18 ( ) Air-fluid exchange 211 (68.1) 81 (38.4) 0.86 ( ) Endolaser 147 (47.4) 67 (45.6) 1.28 ( ) Gas tamponade 92 (29.7) 37 (40.2) 1.00 ( ) C3F8 39 (12.6) 19 (48.7) 1.43 ( ) SF6 53 (17.1) 18 (34.0) (ref) Internal limiting membrane peel 89 (28.7) 18 (20.2) 0.42 ( ) With indocyanine green 70 (22.6) 15 (21.4) 1.36 ( ) Without indocyanine green 19 (6.1) 3 (15.8) (ref) Intravitreal triamcinolone 74 (23.9) 18 (24.3) 0.54 ( ) Epiretinal membrane peel 65 (21.0) 11 (16.9) 0.36 ( ) Diathermy 62 (20.0) 30 (48.4) 1.26 ( ) Perfluorocarbon liquid 49 (15.8) 25 (51.0) 1.33 ( ) Silicone oil removal 39 (12.6) 19 (48.7) 1.25 ( ) Silicone oil injection or exchange 37 (11.9) 18 (48.6) 1.24 ( ) Intravitreal injection of medication 36 (11.6) 15 (41.7) 1.08 ( ) Scleral buckle 30 (9.7) 16 (53.3) 1.32 ( ) Phacoemulsification with IOL insertion 26 (8.4) 10 (38.5) 0.95 ( ) Anterior segment IOL manipulation 20 (6.5) 14 (70.0) 1.83 ( ) Scleral sutured IOL 15 (4.8) 10 (66.7) 1.71 ( ) Retinectomy 12 (3.9) 6 (50.0) 1.25 ( ) Posterior capsulotomy 8 (2.6) 6 (75.0) 1.90 ( ) Lensectomy 7 (2.3) 4 (57.1) 1.43 ( ) Cryotherapy 5 (1.6) 3 (60.0) 1.50 ( ) Retisert implantation 2 (0.6) 0 (0) 0.41 ( ) Explantation of Ahmed glaucoma valve 2 (0.6) 2 (100) 2.50 ( ) AE indicates adverse events; CI, confidence interval; IOL, intraocular lens; RR, relative risk. *One surgery was performed with both 25- and 27-gauge pars plana vitrectomy. One subtenon injection. Includes explantation, repositioning, exchange and removal of IOL, secondary insertion of anterior or posterior chamber IOL, and/or sulcus IOL. occurred earlier in group 2 and 3 (Fig. 2A). Vitrectomy gauge size did not result in a statistically significant difference in time to AE by Kaplan-Meier survival analysis (P = 0.733) (Fig. 2B). On the other hand, there was a statistically significant reduction of all AEs among those undergoing isolated macular surgery (P < 0.001) and time to AE by Kaplan-Meier survival analysis (P < ) (Fig. 2C). Of those undergoing macular surgery, there were 54 isolated ERM peels, 20 macular hole repairs, and 11 combined ERM peel and MH repairs. On POD1, 5 (5.9%) patients had AEs (all related to hypotony or elevated IOP). Four more patients had IOP-related AEs on their second POV. Of the 64 isolated macular PPVs with 3 or more visits and no AEs on their first 2 visits, only 2 had late AEs. One patient returned to clinic POD11 after ERM peel and was found to have worsening of CME, which resolved by the POM1 visit. The second patient developed diabetic macular edema on her POM1 visit and was treated with intravitreal bevacizumab. Elevated IOP was the most common AE throughout the entire postoperative period (Table 4). Of the 12 patients diagnosed with elevated IOP on POD1, 11 were treated with IOP-lowering medications alone, but 1 required an AC tap for an IOP of 41 mm Hg. Six patients were diagnosed with hypotony on POD1 visit 1 had a tube shunt placed at the time of the surgery, 2 were eventually diagnosed with ciliary body shutdown, and 1 was found to have a small wound leak that self-sealed. Two patients required injection of sterile air in the clinic and 1 of them was brought back to the operating room on POD2 to confirm closure of vitrectomy ports. Vitrectomy gauge was not associated with hypotony (P = 0.82). Cystoid macular edema was the most common late AE with a peak incidence of 6.8% at POD Of those with postoperative CME, 26.3% (5/19) had a preoperative diagnosis of CME and 10.5% (2/19) had a preoperative diagnosis of proliferative diabetic retinopathy. Treatment for CME was topical steroids and/ or nonsteroidal anti-inflammatory drops in 9 patients, intravitreal anti vascular endothelial growth factor agents in 7 patients, and intravitreal triamcinolone in 3 patients. Postoperative RD was seen in 18 patients. Of them, 15 had prior RDs, 2 had prior endophthalmitis (1 after ruptured globe endophthalmitis), and 1 occurred 6 weeks after PPV for nonclearing VH for Terson syndrome. 4

5 Adverse Events After Vitrectomy TABLE 4. Timing of Initial Adverse Event* Initial Adverse Event POD 0 1 (n = 310) POD 2 4 (n = 28) POD 5 10 (n = 274) No. (%) of Patients POD (n = 110) POD (n = 232) POD (n = 108) POD (n = 73) Total PPVs (n = 310) IOP >30 mm Hg 12 (3.9) 1 (3.6) 13 (4.7) 6 (5.5) 3 (1.3) 3 (2.8) 38 (12.3) Cystoid macular edema 2 (1.8) 7 (3.0) 5 (4.6) 5 (6.8) 19 (6.1) Retinal detachment 3 (1.1) 5 (4.5) 6 (2.6) 2 (1.9) 2 (2.7) 18 (5.8) Vitreous hemorrhage 7 (2.3) 1 (3.6) 3 (1.1) 2 (1.8) 2 (0.9) 1 (1.4) 16 (5.2) Hypotony 6 (1.9) 1 (3.6) 2 (0.7) 2 (1.8) 4 (1.7) 15 (4.8) Hyphema 4 (1.3) 1 (3.6) 2 (0.7) 2 (1.8) 9 (2.9) Proliferative vitreoretinopathy 1 (0.4) 2 (1.8) 3 (1.3) 2 (2.7) 8 (2.6) Choroidal detachment 3 (1.0) 1 (3.6) 1 (0.4) 2 (1.8) 1 (0.4) 8 (2.6) Subluxed/dislocated PCIOL 2 (0.6) 2 (0.7) 2 (0.9) 2 (2.7) 8 (2.6) Persistent subretinal fluid 1 (0.3) 1 (0.9) 3 (1.3) 1 (1.4) 6 (1.9) Epithelial defect 1 (0.4) 3 (2.7) 1 (0.4) - 5 (1.6) Epiretinal membrane 2 (0.9) 2 (1.9) 1 (1.4) 5 (1.6) Suture granuloma/reaction 2 (1.8) 3 (1.3) 5 (1.6) Silicone oil in anterior 1 (3.6) 2 (0.7) 1 (0.9) 4 (1.3) chamber Macular hole 1 (0.4) 1 (0.9) 2 (1.9) 4 (1.3) Anterior segment 2 (1.8) 1 (0.9) 3 (1.0) neovascularization Anterior uveitis 1 (0.9) 1 (0.4) 1 (0.9) 3 (1.0) Possible gas leak 1 (3.6) 1 (0.3) IOP indicates intraocular pressure; PCIOL, posterior chamber intraocular lens; POD, postoperative day; PPV, pars plana vitrectomy. *Adverse events were only counted once for each individual at the earliest postoperative visit even if subsequently noted at multiple visits. n represents number of patients who had at least 1 postoperative visit within this time period. One associated with tube, 1 wound leak, 2 ciliary body shutdown. Two with choroidal hemorrhage. Epithelial defects not recorded until POD7 at which point they were considered persistent. One with von Hippel Lindau syndrome. DISCUSSION The safety of vitreoretinal surgery has improved over the past several decades. However, the current standard of postoperative monitoring is still based on data from the 1980s. These studies showed that a high percentage of vitrectomy patients required inpatient care, mostly for recovery from general anesthesia. 14 Due to advances in vitrectomy equipment and the increasingly frequent use of local over general anesthesia, serious postoperative AEs have become less common. 15,16 Data from this study suggest that patients undergoing uncomplicated macular surgery may not require a POM1 visit. In addition, patients who were placed on a standard postoperative evaluation schedule were less likely to have AEs than those where the physician or patient requested a more intensive follow-up schedule. Multiple studies have reported on the incidence of early postoperative IOP elevation after PPV. Although rarely used anymore, 20-gauge sutured PPV is associated with the highest rate of IOP elevation and IOP instability, particularly in comparison with 23-gauge PPV 9,10,17,18 and 25-gauge PPV. 11,19,20 A recent study of 27-gauge PPV has reported a low incidence of postoperative IOP spikes. 13 This review of 23-, 25-, and 27-gauge PPV found an elevated IOP in 3.9% at POD1, which is similar to rates reported in current literature (0.7% 10.2%). 1-3 Notably, there were a significant proportion of patients with late IOP spikes (Table 4), although most of these were associated with secondary etiologies including steroid-induced, neovascular glaucoma and pupillary block/angle closure secondary to gas or oil tamponade. Nevertheless, a normal POD1 and POW1 IOP did not preclude IOP elevation at later visits, especially in patients undergoing surgery for proliferative diabetic retinopathy or those requiring silicone oil. In this cohort, POD1 hypotony occurred in 1.9% of patients, which is similar to other studies (1.8% 9.2%). 1,5,13,17 These studies have shown that hypotony often resolves within 1 to 2 weeks for 20-, 23-, and 27-gauge vitrectomies. 5,13,17 Two patients in the present review required reoperation for hypotony in the first week. Neither of these patients were found to have a wound leak, and ciliary body shutdown was implicated. Reports of postoperative RD after PPV have also been low and varied depending on gauge size. The rate of RD after 20-gauge vitrectomy in previous studies has ranged between 1.7% and 14% For 23-gauge vitrectomies, the incidence of postoperative RD ranges from 0.2% to 2%. 5,24,25 The rate of RD in 25-gauge PPV has been reported as 2% to 2.2%. 11,12 The incidence of RD in this review was 5.8% and peaked at POD 11 to 21 days. With the continued emphasis on cost-effective health care, there has been a push towards eliminating unnecessary POVs. 1 3 Alexander et al 1 found that POD1 examinations of their

6 Shields et al Adverse Event Frequency (%) (P = 0.06) (P = 0.84) (P = 0.004) (P = 0.28) (P = 0.23) Postoperative Day s-pov (n = 212) a-pov (n = 68) p-pov (n = 30) (P = 0.32) (P = 0.72) FIGURE 1. Comparison of the adverse event rate as a function of the follow-up group. The only significant difference was seen in postoperative visit (POV) 5 10 window. There was no difference in adverse event rate in the late postoperative period. s-pov indicates standard, a-pov physician-adjusted, and p-pov patientinitiated postoperative visit. patients culminated in changes in medication in 6 (2%) due to IOP greater than 30 mm Hg and hypotony, and reoperation in 4 (1.5%) due to a flat AC, hyphema, nuclear fragment in the AC, and intraocular foreign body. Brennan et al 2 conducted a study of gauge vitrectomies with planned POD1 follow-up and found that 8 (4.5%) had IOP greater than 40 mm Hg and 2 (1%) required surgical intervention (1 for hypotony due to a single sclerotomy leak and 1 for severe pain due to silicone oil overfill). Zick and Joondeph 3 reported on 134 PPVs where only 1 patient required an intervention (vitreous tap for gas overfill) on POD1. In response to the low intervention rate, those authors no longer conduct routine POD1 follow-up visits. Based on the present data, the need for routine POD1 evaluation remains equivocal. In this study, 3.9% of patients had IOP greater than 30 mm Hg, and 3 patients required procedural intervention [1 patient required an AC tap (IOP, 41 mm Hg) and 2 patients required an injection of sterile air for hypotony]. Our data suggest that elimination of the POW1 visit is not indicated, given the higher incidence of AEs across all 3 groups (Table 4, Fig. 1). However, our data suggest that the POM1 visit can be eliminated in patients undergoing isolated macular surgery and who do not have AEs at the POD1 or POW1 visit. Few studies have reported the AE rate after macular surgery with none indicating that it would be safe to eliminate the POM1 follow-up visit. 26,27 The comparison of standard postoperative follow-up (group 1), physician-directed follow-up periods (group 2), and patientdirected follow-up requests (group 3) provides a distinct clinical window into scheduling. The findings suggest that physicians and patients are capable of identifying early AEs or settings in which AEs are more likely. This may also reveal that patients who are deemed to require only standard follow-up may require less intensive evaluation. Clinical judgment is required in these cases, as there is a wide variability in patients ability to identify symptoms that suggest possible AEs. This report is not without limitations. The retrospective design precluded definitive establishment of causation and prevented standardized documentation of postoperative AEs. In addition, the small number of patients undergoing 23- and 27-gauge PPV precluded more robust statistical analysis of our A Probability of No Adverse Events B Probability of No Adverse Events C 1.0 Probability of No Adverse Events Schedule Type: p-pov a-pov s-pov Censored Log rank test, P = Gauge Type: Censored Log rank test, P < Surgery Type: Isolated Macular Surgery Non-Isolated Macular Surgery + Censored Log rank test, P < Time (Days) FIGURE 2. Kaplan-Meier curves of time to adverse events. A, A significant difference among standard (s-pov), physician-adjusted (a-pov), and patient-altered (p-pov) postoperative visits. B, No significant difference in time to adverse events among 23-, 25-, and 27-gauge vitrector. C, A significant difference in time to adverse events between isolated macular surgery and non-isolated macular surgery. findings. This study was conducted at a tertiary referral center and may suffer from referral bias. Specifically, there was a relatively high proportion of complex vitreoretinal surgeries and rare diagnoses, including 6 eyes with von Hippel Lindau syndrome during the study period. These findings are likely to have artificially elevated AE rates. The data in this review suggest that the current standard schedule for follow-up visits may be tailored to the patient depending on the surgery performed. Specifically, patients undergoing standard macular surgery may require less frequent postoperative evaluation, whereas those undergoing more complex vitreoretinal surgery, including RD repair and secondary lens surgery, may require more frequent evaluations. 6

7 Adverse Events After Vitrectomy REFERENCES 1. Alexander P, Michaels L, Newsom R. Is day-1 postoperative review necessary after pars plana vitrectomy? Eye (Lond). 2015;29: Brennan N, Reekie I, Ezra E, et al. The role of day one postoperative review of intraocular pressure in modern vitrectomy surgery. Br J Ophthalmol. 2017;101: Zick J, Joondeph BC. Is a postoperative day one examination needed after uncomplicated vitreoretinal surgery? Retina. 2018;38: Tranos P, Bhar G, Little B. Postoperative intraocular pressure spikes: the need to treat. Eye (Lond). 2004;18: Parolini B, Prigione G, Romanelli F, et al. Postoperative complications and intraocular pressure in 943 consecutive cases of 23-gauge transconjunctival pars plana vitrectomy with 1-year follow-up. Retina. 2010;30: Hasegawa Y, Okamoto F, Sugiura Y, et al. Intraocular pressure elevation in the early postoperative period after vitrectomy for rhegmatogenous retinal detachment. Jpn J Ophthalmol. 2012;56: Anderson NG, Fineman MS, Brown GC. Incidence of intraocular pressure spike and other adverse events after vitreoretinal surgery. Ophthalmology. 2006;113: Allan BD, Baer RM, Heyworth P, et al. Conventional routine clinical review may not be necessary after uncomplicated phacoemulsification. Br J Ophthalmol. 1997;81: Ahn SJ, Woo SJ, Ahn J, et al. Comparison of postoperative intraocular pressure changes between 23-gauge transconjunctival sutureless vitrectomy and conventional 20-gauge vitrectomy. Eye (Lond). 2012;26: Gosse E, Newsom R, Hall P, et al. Changes in day 1 post-operative intraocular pressure following sutureless 23-gauge and conventional 20-gauge pars plana vitrectomy. Open Ophthalmol J. 2013;7: Fujii GY, De Juan E Jr, Humayun MS, et al. Initial experience using the transconjunctival sutureless vitrectomy system for vitreoretinal surgery. Ophthalmology. 2002;109: Ibarra MS, Hermel M, Prenner JL, et al. Longer-term outcomes of transconjunctival sutureless 25-gauge vitrectomy. Am J Ophthalmol. 2005; 139: Yoneda K, Morikawa K, Oshima Y, et al. Surgical outcomes of 27-gauge vitrectomy for a consecutive series of 163 eyes with various vitreous diseases. Retina. 2017;37: Isernhagen RD, Michels RG, Glaser BM, et al. Hospitalization requirements after vitreoretinal surgery. Arch Ophthalmol. 1988;106: Costen MT, Newsom RS, Wainwright AC, et al. Expanding role of local anaesthesia in vitreoretinal surgery. Eye (Lond). 2005;19: Newsom RS, Wainwright AC, Canning CR. Local anaesthesia for 1221 vitreoretinal procedures. Br J Ophthalmol. 2001;85: Framme C, Klotz S, Wolf-Schnurrbusch UE, et al. Intraocular pressure changes following 20G pars-plana vitrectomy. Acta Ophthalmol. 2012; 90: Misra A, Ho-Yen G, Burton RL. 23-gauge sutureless vitrectomy and 20-gauge vitrectomy: a case series comparison. Eye (Lond). 2009;23: Hasegawa Y, Okamoto F, Sugiura Y, et al. Intraocular pressure elevation after vitrectomy for various vitreoretinal disorders. Eur J Ophthalmol. 2014; 24: Lakhanpal RR, Humayun MS, de Juan E Jr, et al. Outcomes of 140 consecutive cases of 25-gauge transconjunctival surgery for posterior segment disease. Ophthalmology. 2005;112: Banker AS, Freeman WR, Kim JW, et al. Vision-threatening complications of surgery for full-thickness macular holes. Vitrectomy for Macular Hole Study Group. Ophthalmology. 1997;104: Ramkissoon YD, Aslam SA, Shah SP, et al. Risk of iatrogenic peripheral retinal breaks in 20-G pars plana vitrectomy. Ophthalmology. 2010;117: Park SS, Marcus DM, Duker JS, et al. Posterior segment complications after vitrectomy for macular hole. Ophthalmology. 1995;102: Singh CN, Iezzi R, Mahmoud TH. Intraocular pressure instability after 23-gauge vitrectomy. Retina. 2010;30: Gupta OP, Ho AC, Kaiser PK, et al. Short-term outcomes of 23-gauge pars plana vitrectomy. Am J Ophthalmol. 2008;146: Ozkaya A, Erdogan G, Tulu B, et al. The outcomes of subtotal vitrectomy in macular surgeries: a single surgeon case series. Int Ophthalmol Feb 7. Epub ahead of print. 27. Naruse S, Shimada H, Mori R. 27-gauge and 25-gauge vitrectomy day surgery for idiopathic epiretinal membrane. BMC Ophthalmol. 2017;17: