Abstract
Purpose:
To determine the unplanned return to OR rate within 180 days and at any time postoperatively after valved and non-valved tube shunt surgery
Design:
Retrospective case-control study
Methods:
Review of 357 eyes that underwent tube shunt surgery (151 valved, 206 non-valved) at an academic glaucoma service between 01/2014–12/2016. A control eye was time-matched for each eye that underwent reoperation.
Results:
The reoperation rate within 180 days was 10.6% (16/151) for valved and 12.1% (25/206) for non-valved tube shunts and at any time postoperatively was 20.5% (31/151) for valved and 22.8% (47/206) for non-valved tube shunts. Mean postoperative follow-up was 2.8±1.1 years. The most common reoperations within 180 days and at any time postoperatively after valved tube shunt surgery were tube revision (43.8% within 180 days, 38.7% anytime) and external CPC (31.3% within 180 days, 38.7% anytime). The most common reoperations within 180 days after non-valved tube shunt surgery were tube revision (32.0%), external CPC (12.0%), and vitrectomy with AC washout (12.0%) and at any time postoperatively were tube revision (34.0%), external CPC (31.9%), and tube explant (12.8%). At last follow-up, eyes that returned to the OR and controls were similar in terms of mean IOP, proportion of eyes meeting target IOP, and change in VA.
Conclusions:
Over 20% of eyes undergoing tube shunt surgery returned to the OR at any time postoperatively with a mean follow-up of nearly 3 years, with over 10% of eyes undergoing reoperation within the first 180 days. Rates of reoperation were similar between valved and non-valved tube shunts.
Graphical Abstract
This retrospective case-control study examined the rate of unplanned return to the operating room after attending-performed and attending-supervised tube shunt surgery at a single academic center. A considerable number of patients who underwent tube shunt implant required additional surgery; reoperation rates were similar between eyes that received valved versus non-valved tube shunts. These findings provide information on the postoperative course after tube shunt surgery, which may be important in setting appropriate surgeon and patient expectations.
Introduction
Studies of unplanned return to the operating room (OR) within ophthalmology have almost exclusively focused on resident-performed surgeries, most commonly cataract and vitreoretinal surgeries1-4. Among glaucoma surgeries, Hsia and colleagues examined complication and reoperation rates after resident-performed glaucoma procedures5, and our group has recently reported on unplanned return to the operating room after attending- and fellow-performed trabeculectomy6.
Although relatively new to the ophthalmic literature, unplanned return to OR has been investigated as a quality metric across multiple surgical fields, including general surgery, colorectal surgery, vascular surgery, and pediatric neurosurgery7-10. Unplanned return to OR has been shown to be associated with prolonged hospitalization and mortality, both of which are used as quality measures in surgical fields7,8. Furthermore, unplanned return to OR is a metric that can be broadly applicable to a wide array of surgeries, is an objective, non-discretionary measure, and can be easily identified from electronic health records, making it an attractive potential quality measure10.
Understanding the patient-, surgeon-, and procedure-specific factors that are associated with unplanned return to OR in glaucoma surgeries can inform efforts to improve upon the current standard of care. We set out to identify factors that are associated with unplanned return to OR after tube shunt surgery at an academic glaucoma service. The results of this study may help predict which patients are at increased risk for postoperative complications and unplanned reoperation.
Methods
This was a retrospective case-control study examining the outcomes of tube shunt surgery performed by glaucoma specialists at the Wilmer Eye Institute from January 2014 through December 2016. The study protocol was reviewed and approved by the Johns Hopkins University School of Medicine Institutional Review Board. The study adhered to the Declaration of Helsinki.
Patients included in the study were identified by searching billing records for procedures linked to CPT code 66179 or 66180. Inclusion criteria included postoperative follow-up of at least 180 days, age 18 years or older at the time of surgery, and no other glaucoma-related surgery (trabeculectomy, tube shunt, bleb needling, or trabeculotomy) in the same eye within the year preceding the index surgery. If both eyes of the same patient underwent tube shunt surgery within the study period, the first eye that met inclusion criteria was selected for the study. In the few cases in which a patient had multiple tube shunt surgeries in the same eye within the study period, only the first tube shunt surgery that met inclusion criteria was included. For each patient who was identified as having had an unplanned reoperation at any time postoperatively related to a direct complication from their tube shunt surgery or progression of their glaucoma, a control patient that had undergone tube shunt surgery and was time-matched within one month was selected by a web-based random number generator (Google, Alphabet Inc., Menlo Park, CA) for comparison. In this study, an unplanned return to the OR was defined as a second glaucoma surgery in the study eye that was not already scheduled prior to or at the time of the original tube shunt surgery. Multi-phase tube shunt implantations were not counted as unplanned reoperations.
Clinical data were extracted from the electronic medical record: age at the time of surgery, sex, race, eye laterality, type of glaucoma, history of incisional eye surgery, highest recorded intraocular pressure (IOP), visual acuity (VA), number of glaucoma eyedrops, and use of oral IOP-lowering medications at the visit prior to surgery. Combination eyedrops were counted as 2 glaucoma medications. The target IOP most recently documented by the treating surgeon was recorded. Surgical details recorded included surgeon, whether the surgery was combined with a cataract extraction, model of tube implant, and location of tube implant. Details documented in the medical record from the postoperative course that were recorded included any postsurgical complications (choroidal effusion, tube-corneal touch, endophthalmitis, flat anterior chamber, elevated IOP refractory to medical therapy, hyphema, hypotony, hypotony maculopathy, extraocular motility problems, pupillary/ciliary block, suprachoroidal hemorrhage, tube exposure, tube migration, wound leak, blebitis at the site of a prior trabeculectomy, or other), any postoperative treatment for the complication, and whether the complication resolved spontaneously or required intervention. Finally, whether a reoperation was performed at any time in the postoperative period, the time frame in which it was performed, reason for reoperation, type of reoperation surgery, number of additional reoperations, and the final VA, IOP, number of eye drops, and use of oral IOP-lowering medication at the last follow-up visit were recorded. Last follow-up was defined as the last clinical examination recorded on or before March 31, 2019. Throughout the analysis of postoperative complications, eyes experiencing multiple postoperative complications at once were counted once in the “any complications” category and counted multiple times among the further breakdown of specific postoperative complications. Postoperative hypotony was based on clinician assessment as recorded in the medical record, rather than an IOP cutoff value and included keywords such as “hypotony” or “IOP too low.” Postoperative hypotony maculopathy was similarly based on clinician judgement, and typically included retinal changes affecting the patient’s vision. For this paper, we did not require a macular optical coherence tomography image to confirm maculopathy. The majority of IOP measurements on the glaucoma service were performed using applanation tonometry, most often by experienced technicians, although some measurements were with the iCare tonometer (Icare USA, Raleigh, NC). Snellen VA was converted to its logarithm of the minimum angle of resolution (logMAR) equivalent to assess changes in VA11.
The tube shunt implants were either the Baerveldt BG 101-350 (B350; Johnson & Johnson Vision, Santa Ana, CA), the Baerveldt BG 103-250 (B250; Johnson & Johnson Vision), or the Ahmed FP7 glaucoma valve (AGV; New World Medical Inc., Rancho Cucamonga, CA). Most tubes were covered with an irradiated, split-thickness corneal patch graft prior to conjunctival closure. Baerveldt tubes were variably modified by being tied off, having a ripcord suture inserted, and being fenestrated anterior to the tube ligation. The tubes were classified as either “valved” (AGV) or “non-valved” (B350 and B250) for analysis in this study. Tube revision surgery included a variety of techniques aimed at modifying the existing tube shunt to improve its function or address complications, including tying the tube shunt with suture, flushing, repositioning, lengthening or shortening the tube, creating new fenestrations, and/or resuturing the conjunctiva.
Statistical Methods
The primary outcome was whether an unplanned reoperation, related either to the patient’s glaucoma or as a result of a postoperative complication, was performed within 180 days of the original surgery. A second measure was whether a reoperation occurred at any time postoperatively. Student’s t-test was used to compare the mean follow-up time, mean baseline IOP between the reoperation and control cases for each type of tube shunt and between eyes undergoing valved versus non-valved tube shunt, and the mean change in logMAR between the clinic visit just prior to tube shunt surgery and the last follow-up visit. The Wilcoxon rank-sum test was used to compare mean age at surgery, mean target IOP, mean number of glaucoma eyedrops being used prior to surgery, and mean number of additional surgeries following reoperation, as well as the mean IOP and mean number of glaucoma eyedrops at the last follow-up visit between the reoperation and control groups for each type of tube shunt. Pearson’s chi-square test was used to compare eye laterality, sex, race, type of glaucoma, prior incisional and glaucoma surgery, lens status, surgeon, location of the tube shunt implant, and whether eyes attained their target IOP level at last follow-up between the reoperation and control cases for each type of tube shunt, as well as eye laterality, sex, race, type of glaucoma, lens status, history of prior incisional and glaucoma surgery, and reason for reoperation between eyes undergoing valved versus non-valved tube shunt. Kaplan-Meier survival analysis was performed, with failure defined as an unplanned return to the OR. The log-rank test was used to compare Kaplan-Meier survival curves between the valved and non-valved tube shunts.
To identify factors associated with unplanned reoperation both within 180 days and at any time in the postoperative period, 31 factors for tube shunt surgery were initially tested for inclusion in Cox proportional hazards models, and factors with univariate significance p ≤ 0.10 were used to construct multivariable models. All types of tube shunt implants were combined in these models. These models were systematically tested by dropping the variables with the largest p-values, and the most parsimonious model, identified as the model with the lowest Akaike information criterion, was chosen as the final model. A hazard ratio greater than 1 indicates a higher chance of failure compared with the reference group.
The 31 factors initially tested for inclusion in the models were age at surgery, sex, race, eye laterality, type of glaucoma, maximum IOP prior to initiation of therapy, target IOP, number of preoperative glaucoma eye drops, use of preoperative oral IOP-lowering medications, prior incisional eye surgery (trabeculectomy, tube shunt, cataract surgery, complex cataract surgery, or pars plana vitrectomy), whether cataract extraction was combined with the tube shunt surgery, whether the tube implant was valved or not, tube implant location, surgeon, postsurgical complications (choroidal effusion, flat anterior chamber, hyphema, hypotony, suprachoroidal hemorrhage, tube migration, wound leak, bleb leak or blebitis at the site of a previous trabeculectomy, or other complications), and postsurgical treatment to address the complication (bandage contact lens). Postoperative endophthalmitis was not included in the model given that it was a rare complication.
A p-value < 0.05 was considered statistically significant. In situations in which multiple statistical comparisons were made, the p-value threshold of 0.05 was adjusted by dividing the threshold by the number of statistical comparisons. These situations include the comparisons of follow-up time, age at surgery, eye laterality, sex, race, type of glaucoma, prior incisional surgery, prior glaucoma surgery, baseline IOP, target IOP, and number of preoperative glaucoma eyedrops between the reoperation and control cases for each type of tube shunt and between eyes that underwent valved versus non-valved tube shunt, as well as comparisons between the reoperation and control cases of change in logMAR, IOP, and number of glaucoma eyedrops at the last follow-up visit. Study data were collected and managed using REDCap12,13. All statistical analyses were performed in STATA (StataCorp. 2017. Stata Statistical Software: Release 15. College Station, TX: StataCorp LLC).
Results
Rate of Unplanned Return to the OR
Of the 357 included tube shunt surgeries performed during the study period, 151 surgeries were valved tube shunts and 206 surgeries were non-valved tube shunts. Forty-seven (22.8%) of the non-valved tube shunts and 31 (20.5%) of the valved tube shunts returned to the OR at any time postoperatively for an unplanned reoperation related either to glaucoma or a complication of tube shunt surgery (mean follow-up 2.8 ± 1.1 years). Twenty-five (12.1%) eyes with non-valved tube shunts and 16 (10.6%) eyes with valved tube shunts underwent reoperation within the first 180 postoperative days. The cumulative rate of return to OR stratified early in the postoperative period is shown in Figure 1.
Figure 1. Cumulative rate of unplanned return to OR after tube shunt surgery.
This graph shows stratification of the cumulative rates of reoperation early in the postoperative period. Forty-seven (22.8%) eyes that underwent non-valved tube shunt and 31 (20.5%) eyes that underwent valved tube shunt returned to the operating room at any time postoperatively for an additional surgery related either to the patient’s glaucoma or a complication from the tube shunt surgery. Of the 25 (12.1%) non-valved tube shunts and 16 (10.6%) valved tube shunts that returned to the OR within the first 180 days, roughly three-quarters (9.7% non-valved, 7.9% valved) of the reoperations occurred within the first 90 days.
For both valved and non-valved tube shunts, reoperation and control eyes did not vary in terms of mean age at the time of surgery, eye laterality, race, sex, type of glaucoma, history of prior incisional or glaucoma surgery, lens status, mean baseline IOP, mean target IOP, or mean number of glaucoma eyedrops being used prior to surgery (Table 1).
Table 1.
Demographics of subjects and time-matched controls who returned to the OR at any time postoperatively.
| Non-valved (Baerveldt) tubes | Valved (Ahmed) tubes | |||
|---|---|---|---|---|
| Reoperations | Controls | Reoperations | Controls | |
| n | 47 | 44 | 31 | 34 |
| Mean (SD) follow-up time in years | 2.7 (1.2) | 3.1 (1.1) | 2.6 (1.1) | 2.9 (1.1) |
| Mean (SD) age in years* | 63.3 (13.1) | 64.9 (16.2) | 54.2 (15.9) | 57.1 (17.5) |
| Eye (% R eye) | 49% | 56% | 45% | 59% |
| Sex (% male) | 55% | 32% | 52% | 38% |
| Race (n, %) | ||||
| White | 18 (38%) | 19 (43%) | 13 (42%) | 19 (56%) |
| Black | 22 (47%) | 21 (48%) | 15 (48%) | 14 (41%) |
| Asian | 5 (11%) | 2 (5%) | 2 (6%) | 0 (0%) |
| Hispanic | 1 (2%) | 0 (0%) | 0 (0%) | 1 (3%) |
| Other | 2 (4%) | 2 (5%) | 1 (3%) | 0 (0%) |
| Type of Glaucoma (n,%) | ||||
| OAG* | 31 (66%) | 25 (57%) | 8 (26%) | 8 (24%) |
| ACG | 2 (4%) | 2 (5%) | 2 (6%) | 4 (12%) |
| Uveitic/Inflammatory | 7 (15%) | 8 (18%) | 3 (10%) | 5 (15%) |
| PXF | 0 (0%) | 1 (2%) | 1 (3%) | 1 (3%) |
| NVG* | 4 (9%) | 2 (5%) | 13 (42%) | 9 (26%) |
| Pigmentary | 1 (2%) | 0 (0%) | 0 (0%) | 0 (0%) |
| Traumatic | 1 (2%) | 2 (5%) | 1 (3%) | 1 (3%) |
| Suspect | 0 (0%) | 3 (7%) | 0 (0%) | 1 (3%) |
| Other | 1 (2%) | 1 (2%) | 3 (10%) | 5 (15%) |
| Prior Incisional Surgery (n,%) | 39 (83%) | 33 (75%) | 20 (65%) | 22 (65%) |
| Prior Glaucoma Surgery (n,%) | ||||
| Any* | 25 (53%) | 26 (59%) | 7 (23%) | 11 (32%) |
| Prior trab* | 22 (47%) | 23 (52%) | 6 (19%) | 7 (21%) |
| Prior tube | 6 (13%) | 4 (9%) | 2 (6%) | 4 (12%) |
| Other (trab revision, tube revision, CPC) | 3 (6%) | 6 (14%) | 1 (3%) | 0 (0%) |
| Mean (SD) Baseline IOP | 37.5 (11.5) | 35.9 (9.8) | 47.5 (12.3) | 46.6 (10.3) |
| Lens Status (n,%) | ||||
| Phakic | 21 (45%) | 18 (41%) | 16 (52%) | 15 (44%) |
| Pseudophakic | 22 (47%) | 26 (59%) | 14 (45%) | 17 (50%) |
| Aphakic | 4 (9%) | 0 (0%) | 1 (3%) | 2 (6%) |
| Mean (SD) Target IOP* | 14.9 (3.1) | 15.9 (3.1) | 18.4 (3.3) | 16.7 (2.9) |
| Mean (SD) number of glaucoma eyedrops prior to surgery | 3.1 (1.3) | 3.3 (0.8) | 2.5 (1.5) | 2.8 (1.4) |
OAG = open angle glaucoma, ACG = angle closure glaucoma, PXF = pseudoexfoliation glaucoma, NVG = neovascular glaucoma, CPC = cyclophotocoagulation, SD = standard deviation.
An asterisk is used to denote statistically significant differences in that characteristic between eyes that underwent non-valved tube shunt (both reoperation cases and controls) and valved tube shunt (both reoperation cases and controls) after adjustment for multiple comparisons.
In comparing the demographics of patients who underwent valved versus non-valved tube shunt, patients who received a valved tube shunt were on average younger than those who received a non-valved tube shunt (55.7±16.7 vs. 64.1±14.6 years, p < 0.001). A higher proportion of patients who underwent non-valved tube shunt surgery had open-angle glaucoma as compared to patients who had valved tube shunt surgery (0.615 vs. 0.246, p < 0.001), whereas a higher proportion of patients who had valved tube shunt surgery had neovascular glaucoma as compared to patients who underwent non-valved tube shunt surgery (0.338 vs. 0.066, p < 0.001). There was a higher proportion of eyes with a history of any prior glaucoma surgery (0.560 vs. 0.277, p < 0.001) and with a history of previous trabeculectomy (0.495 vs. 0.200, p < 0.001) among eyes that underwent non-valved tube shunt as compared to valved tube shunt. The proportions of eyes with a prior tube shunt (0.092 for valved vs. 0.110 for non-valved) or other type of glaucoma surgery (trabeculectomy revision, tube revision, external cyclophotocoagulation (CPC), or Trabectome; 0.015 for valved vs. 0.099 for non-valved) were not statistically different between eyes that underwent valved and non-valved tube shunts. The proportions of eyes that were phakic and pseudophakic were similar between eyes that underwent valved versus non-valved tube shunt. Mean target IOP was higher in eyes that underwent valved tube shunt (17.5 ± 0.5 mmHg vs. 15.4 ± 0.4 mmHg, p < 0.001); notably, the mean maximum IOP recorded preoperatively was not statistically different between eyes that underwent valved and non-valved tube shunt implant (47.1 ± 11.3 mmHg vs. 36.9 ± 10.8 mmHg, respectively). Eyes that underwent non-valved versus valved tube shunt surgery were using similar numbers of glaucoma eyedrops prior to surgery (3.2 ± 1.1 eyedrops vs. 2.7 ± 1.5 eyedrops, respectively).
Among non-valved tube shunts, 75.8% (69/91) of tubes were placed superiotemporally, 22.0% (20/91) were placed inferonasally, 1.1% (1/91) were placed inferotemporally, and 1.1% (1/91) did not have the location of the tube recorded in the medical record. Among valved tube shunts, 89.2% (58/65) were placed superiotemporally, 6.2% (4/65) were placed inferonasally, 3.1% (2/65) were placed inferotemporally, and 1.5% (1/65) were placed superionasally. Eyes that later underwent reoperation and control eyes did not significantly differ in terms of tube location for both non-valved and valved tube shunt implants.
The rate of return to the OR was not significantly different across surgeons, ranging from 17.6-31.8% for non-valved tube shunt implants and 10.0-33.3% for valved tube shunt implants. From Kaplan-Meier survival analysis, 85.0%, 80.4%, and 78.6% of non-valved tube shunt implants and 86.1%, 83.9%, and 80.4% of valved tube shunt surgeries avoided unplanned return to the OR at 1, 2, and 4 postoperative years, respectively (Figure 2). The two survival curves did not significantly differ (p = 0.758).
Figure 2. Kaplan-Meier survival curve for unplanned return to the OR after tube shunt surgery.
Failure is defined as an unplanned return to the OR due to a progression of the patient’s glaucoma or a complication arising from the tube shunt surgery. At 1, 2, and 4 years postoperatively, 85.0%, 80.4%, and 78.6% of non-valved tube shunt implants and 86.1%, 83.9%, and 80.4% of valved tube shunt surgeries avoided reoperation. The survival curves for the non-valved and valved tube shunts did not significantly differ (p = 0.758).
Postoperative Complications
The most common postoperative complications in eyes that underwent reoperation at any time in the postoperative period after placement of a non-valved tube shunt were elevated IOP refractory to medical therapy (27 eyes, 13.1% of all non-valved tube shunts), hypotony (7 eyes, 3.4% of all non-valved tube shunts), tube exposure (7 eyes, 3.4% of all non-valved tube shunts), and conjunctival retraction (6 eyes, 2.9% of all non-valved tube shunts). Vitreous hemorrhage was present in 3 eyes (1.5% of all non-valved tube shunts); one of these eyes had neovascular glaucoma, and none had proliferative diabetic retinopathy. The most common postoperative complications in eyes that underwent reoperation at any time postoperatively after placement of a valved tube shunt were elevated IOP refractory to medical therapy (21 eyes, 13.9% of all valved tube shunts), hyphema (4 eyes, 2.6% of all valved tube shunts), hypotony (3 eyes, 2.0% of all valved tube shunts), and tube exposure (3 eyes, 2.0% of all valved tube shunts). Among the 4 eyes with hyphema that returned to the OR after valved tube shunt surgery, 3 eyes had neovascular glaucoma. Two eyes that underwent valved tube shunt (1.3% of all valved tube shunts) had vitreous hemorrhage; one of these eyes had neovascular glaucoma with proliferative diabetic retinopathy. Further detail outlining the breakdown and time-course of postoperative complications for eyes that returned to the OR and control eyes is presented in Table 2.
Table 2.
Counts of postoperative complications after tube shunt surgerya.
| Non-valved Tube Shunt | Valved Tube Shunt | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Reoperation Within: (in days) | Reoperation Within: (in days) | |||||||||||
| Control | Any | 0-30 | 31-90 | 91-180 | 181+ | Control | Any | 0-30 | 31-90 | 91-180 | 181 + | |
| Any | 13 | 46 | 8 | 12 | 5 | 21 | 9 | 29 | 4 | 7 | 4 | 14 |
| Blebitis | 0 | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Choroidals | 4 | 5 | 0 | 4 | 1 | 0 | 3 | 1 | 0 | 1 | 0 | 0 |
| Corneal touch | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 1 |
| Dysesthesia | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Endophthalmitis | 0 | 3 | 2 | 0 | 0 | 1 | 0 | 1 | 0 | 0 | 0 | 1 |
| Flat AC | 2 | 4 | 1 | 2 | 1 | 0 | 1 | 2 | 0 | 1 | 0 | 1 |
| Elevated IOP | 3 | 27 | 2 | 6 | 4 | 15 | 3 | 21 | 3 | 6 | 2 | 10 |
| Hyphema | 1 | 4 | 1 | 1 | 1 | 1 | 1 | 4 | 1 | 2 | 1 | 0 |
| Hypotony | 8 | 7 | 0 | 4 | 2 | 1 | 1 | 3 | 1 | 0 | 0 | 2 |
| Hypotony maculopathy | 0 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 |
| Motility problems | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
| Pupillary/ciliary block | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 2 | 1 | 1 | 0 | 0 |
| Suprachoroidal hemorrhage | 0 | 2 | 1 | 0 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 |
| Tube exposure | 0 | 7 | 2 | 3 | 0 | 2 | 0 | 3 | 0 | 1 | 1 | 1 |
| Tube migration | 0 | 2 | 0 | 0 | 1 | 1 | 0 | 1 | 0 | 1 | 0 | 0 |
| Tube wound leak | 0 | 5 | 1 | 4 | 0 | 0 | 1 | 1 | 0 | 0 | 0 | 1 |
| Otherb | 0 | 15 | 5 | 6 | 1 | 3 | 1 | 7 | 1 | 2 | 2 | 2 |
Eyes experiencing multiple postoperative complications at once were counted once in the “any complications” category and counted multiple times among the further breakdown of specific postoperative complication.
Other complications for non-valved tube shunts include vitreous hemorrhage, malignant glaucoma, conjunctival retraction, progression of the patient’s glaucoma despite meeting IOP goals, a thickened capsule, retrobulbar hemorrhage, and development of lens opacity due to tube shunt position; “other” complications for valved tube shunts include vitreous hemorrhage, vitreous blocking the tube shunt, inflammation, temporal dellen, a granuloma, iris prolapse, tube erosion, and infection at a suture site.
Reoperation Surgery
The most common reason for unplanned return to the OR both within the first 180 days and at any time postoperatively for both valved and non-valved tube shunts was persistently elevated IOP refractory to medical therapy, due to either a postoperative complication resulting in inadequate IOP control such as pupillary or ciliary block, or a failure of the surgery. There was not a statistically significant difference in the proportion of eyes that underwent reoperation due to elevated IOP refractory to medical therapy between valved and non-valved tube shunts (p = 0.36). Among eyes that underwent non-valved tube shunt implant, other reasons for reoperation included choroidal effusion, endophthalmitis, and tube exposure, mainly early in the postoperative period. Further stratification of the reasons for unplanned return to OR for valved and non-valved tube shunts is presented in Supplemental Figure 1.
The most common reoperation surgeries in eyes that returned to the OR within the first 180 postoperative days after a non-valved tube shunt implant were tube revision in 8 eyes (3.9% of all non-valved tube shunts), external CPC in 3 eyes (1.5% of eyes), and vitrectomy with anterior chamber washout in 3 eyes (1.5% of eyes). The most common reoperation surgeries at any time postoperatively after non-valved tube shunt implant were tube revision in 16 eyes (7.8% of all non-valved tube shunts), external CPC in 15 eyes (7.3% of eyes), tube explant in 6 eyes (2.9% of eyes), and vitrectomy with anterior chamber washout in 3 eyes (1.5% of eyes). Six eyes (2.9% of all non-valved tube shunts) underwent a new trabeculectomy or tube shunt implant.
The most common reoperation surgeries in eyes that returned to the OR within the first 180 postoperative days after a valved tube shunt implant were tube revision in 7 eyes (4.6% of all valved tube shunts) and external CPC in 5 eyes (3.3% of all valved tube shunts). Tube revision (12 eyes, 7.9% of valved tube shunts) and external CPC (12 eyes, 7.9% of valved tube shunts) were again the most common reoperation surgeries at any time postoperatively after valved tube shunt implant. Additional details showing the types and timeframe of reoperations is presented in Supplemental Table 1.
Among the 47 eyes that returned to the OR after non-valved tube shunt implant, 15 eyes (31.9%) required additional surgery beyond the first return to OR, including 10 eyes (40.0%) that had originally returned to the OR within 180 days. Among the 31 eyes that returned to the OR after valved tube shunt implant, 10 (32.3%) required additional surgery beyond the first return to OR, including 6 eyes (37.5%) that had originally returned to the OR within 180 days. The mean number of additional surgeries after the original non-valved tube shunt implantation was 1.40 ± 0.65 surgeries and after valved tube shunt implantation was 1.58 ± 1.12 surgeries (p = 0.344).
Last Follow-Up
At the last available follow-up visit, mean IOP was similar between the reoperation and control groups that underwent non-valved tube shunt surgery: 11.2 ± 4.1 mmHg for control eyes and 12.0 ± 4.8 mmHg for eyes that had returned to the OR (p = 0.641). The proportion of eyes at or below their target IOP were similar between reoperation and control eyes: 25 (53.2%) and 32 eyes (72.7%), respectively (p = 0.105). Eyes that had returned to the OR after non-valved tube shunt surgery were on average using more glaucoma eyedrops than control eyes that had undergone non-valved tube shunt surgery: 2.3 ± 1.5 eyedrops versus 1.5 ± 1.3 eyedrops (p = 0.006). The two groups did not differ in their change in VA between the original non-valved tube shunt surgery and last follow-up visit: eyes that eventually returned to the OR lost 0.22 ± 0.72 logMAR units of VA while control eyes lost 0.23 ± 0.60 logMAR units of VA during the same time period (p = 0.920).
Among eyes that underwent valved tube shunt surgery, mean IOP was similar between the reoperation and control groups at the last follow-up visit: 13.9 ± 7.2 mmHg for control eyes and 14.1 ± 11.6 mmHg for eyes that had returned to the OR (p = 0.644). The proportion of eyes at or below their target IOP were similar between reoperation and control eyes: 12 (38.7%) and 12 eyes (35.3%), respectively (p = 0.706). Eyes that had returned to the OR and control eyes were on average using similar numbers of eyedrops at the last follow-up visit: 1.4 ± 1.5 eyedrops versus 2.0 ± 1.6 eyedrops, respectively (p = 0.107). The two groups did not differ in their change in VA between the original surgery and last follow-up visit: eyes that eventually returned to the OR lost 0.15 ± 1.35 logMAR units of VA, while control eyes gained 0.06 ± 0.72 logMAR units of VA during the same time period (p = 0.441).
Factors Associated with Unplanned Return to the OR
The valved and the non-valved tube shunts were pooled for multivariable regression analysis. Whether or not the original tube shunt implant was valved was not significantly associated with an increased risk of return to the operating room both within the first 180 postoperative days (p = 0.601) and at any time postoperatively (p = 0.694) in univariate analysis.
In multivariable regression analysis for unplanned return to OR within the first 180 postoperative days, the presence of postoperative hyphema (HR = 3.52, 95% CI: 1.51-8.22, p = 0.004), choroidal effusion (HR = 2.49, 95% CI: 1.03-6.06, p = 0.044), and other postoperative complications (vitreous hemorrhage, conjunctival retraction, malignant glaucoma, thickened capsule, temporal dellen, and a granuloma; HR = 5.15, 95% CI: 2.63-10.07, p < 0.001) were associated with an increased risk of reoperation (Table 3A). Asian race was not significantly associated with return to OR within 180 days although it remained in the final model. The factors with univariate p-values ≤ 0.10 for reoperation within 180 days that were initially entered into the model were sex, Asian race, glaucoma diagnosis, maximum IOP prior to surgery, postoperative choroidal effusion, flat anterior chamber, hyphema, pupillary/ciliary block, wound leak, and other complications, listed above.
Table 3.
Multivariable regression analysis of factors associated with unplanned return to OR A) within the first 180 postoperative days and B) at any time postoperatively
| Characteristic: | Hazard Ratio: | Std. Error: | P > ∣z∣ | (95% CI) |
|---|---|---|---|---|
| A | ||||
| Post-surgical complication: hyphema | 3.519 | 1.523 | 0.004 | (1.507, 8.218) |
| Post-surgical complication: choroidal effusion | 2.493 | 1.130 | 0.044 | (1.025, 6.059) |
| Post-surgical complication: other | 5.145 | 1.763 | < 0.001 | (2.629, 10.070) |
| Asian race | 2.319 | 1.128 | 0.084 | (0.894, 6.014) |
| B | ||||
| Post-surgical complication: other | 5.287 | 1.443 | < 0.001 | (3.097, 9.026) |
| Maximum IOP prior to surgery | 0.984 | 0.009 | 0.071 | (0.967, 1.001) |
| Post-surgical complication: hyphema | 1.967 | 0.762 | 0.081 | (0.921, 4.201) |
In multivariable regression analysis for unplanned return to OR at any time postoperatively, only the presence of “other” postoperative complications (vitreous hemorrhage, malignant glaucoma, conjunctival retraction, progression of the patient’s glaucoma despite meeting IOP goals, a thickened capsule, retrobulbar hemorrhage, development of lens opacity due to tube shunt position, vitreous blocking the tube shunt, inflammation, temporal dellen, a granuloma, iris prolapse, and tube erosion) was significantly associated with increased risk of reoperation (Table 3B; HR = 5.29, 95% CI: 3.10-9.03, p < 0.001). Both maximum IOP prior to tube shunt surgery and the presence of postoperative hyphema remained in the final model but were not significantly associated with return to OR at any time postoperatively. The factors with univariate p-values ≤ 0.10 for reoperation at any time postoperatively that were initially entered into the model were sex, Asian race, glaucoma diagnosis, maximum IOP prior to surgery, postoperative hyphema, flat anterior chamber, pupillary/ciliary block, suprachoroidal hemorrhage, wound leak, and other complications, listed above. Postoperative endophthalmitis was not included in the model given that it was a rare complication.
Discussion
In a large academic glaucoma practice, over 20% of eyes undergoing tube shunt surgery returned to the OR for further surgery over a mean follow up of nearly three years. Approximately 13% of eyes required reoperation for persistently uncontrolled glaucoma and nearly 10% of eyes returned to the OR for a complication associated with their surgery. Reoperation within six months of surgery was required in over 10% of the cases, with those undergoing non-valved tube shunt surgery requiring this more frequently for postoperative complications, whereas eyes with valved tube shunts required surgery mostly for persistently elevated IOP. Over 30% of eyes that had a reoperation required at least one additional surgery. These findings point to the need to set appropriate patient expectations prior to tube shunt surgery; reoperation is common and can be needed soon after the initial procedure.
The most common reason for unplanned return to OR both within the first 180 days and at any time postoperatively for valved and non-valved tube shunts was persistently elevated IOP refractory to medical therapy, due to either a postoperative complication resulting in inadequate IOP control, such as tube occlusion by the iris or vitreous, or a failure of the surgery. This finding may be explained by the fact that over half of eyes that underwent non-valved tube shunt implant and over a quarter of eyes that underwent valved tube shunt implant during the study period had previously undergone another glaucoma surgery, most commonly trabeculectomy, but occasionally a tube shunt or external CPC. Thus, a substantial portion of the eyes included in the present study were at high risk of failure owing to previous surgery and may be predisposed to scarring owing to a more aggressive healing process.
Of note, over 60 eyes that underwent tube shunt surgery at our institution during the study period were excluded for postoperative follow-up less than 180 days. Many of these eyes were referred to our institution for their surgery, and after the initial postoperative care, followed up with their referring eye care professional for long-term care. It seems probable, though not certain, that those patients would have returned to the original surgeon should they have had serious postoperative complications or required further surgery. Thus, the actual return to OR rates may be lower than reported here given the likely existence of a selection bias with those managed elsewhere and not needing reoperation being excluded. Another 44 eyes were excluded due to having a prior glaucoma surgery in the study eye within the year prior. These cases were excluded due to the fact that eyes that have undergone glaucoma surgery recently are likely at higher risk of requiring further surgery.
In a similarly structured study, Hsia and colleagues reported a reoperation rate of 4.9% within the first 90 postoperative days among 61 valved tube shunt surgeries performed by residents; this in comparison to the 12 eyes (7.9%) in the present study that returned to the OR within the first 90 postoperative days5. The indications for revision surgery after resident-performed surgeries were tube revision, wound leak, and tube extension. In the current study, 5 eyes (3.3%) underwent tube revision surgery within the first 90 days after valved tube shunt implant; the remaining 7 eyes underwent external CPC, external CPC with tube explant, external CPC with cataract extraction, and anterior chamber washout. Accordingly, the higher rate of return to the OR among attending surgeons in the present study appears to be largely driven by interventions for persistently elevated IOP refractory to medical therapy or a progression of the patient’s glaucoma. Attending surgeons may have a lower threshold for returning to the OR early in the postoperative period to address elevated IOP surgically as compared to residents, who may opt to attempt to control IOP medically.
In one study that reported the outcomes of 48 resident-performed non-valved tube shunt surgeries, 3 eyes (6.3%) returned to the operating room within the first postoperative year: 2 eyes that required ripcord removal in the OR and another eye that returned to the OR due to dehiscence. A fourth eye was lost to follow-up prior to the planned second tube14. Unfortunately, the time frame of these reoperations is not provided, limiting comparison of rate of return to OR with the present study.
We were encouraged by the relatively small decline in VA for patients from the time of the original tube shunt surgery to the last follow-up visit, a mean loss of slightly more than 0.2 logMAR for eyes that had undergone non-valved tube shunt surgery and 0.15 logMAR for eyes that had undergone valved tube shunt surgery. Given the average follow-up of nearly 3 years, multiple factors unrelated to the trabeculectomy surgery and the patient’s glaucoma could have contributed to this mild loss of vision, including the development of age-related retinal pathology as well as progression of cataracts in phakic eyes.
The factors most strongly associated with unplanned return to the OR within 180 days were postoperative hyphema, choroidal effusion, and “other” complications – most commonly vitreous hemorrhage and conjunctival retraction – and at any time postoperatively were “other” complications, again most commonly vitreous hemorrhage and conjunctival retraction. While many of these complications may resolve on their own without surgical intervention, the persistence of these complications over the first few weeks or months of postoperative follow-up may be associated with other severe complications prompting urgent surgical intervention. With the exception of vitreous hemorrhage, which itself often resulted in a reoperation for vitrectomy, the other complications associated with return to the OR often were not the primary indication for reoperation.
An important limitation of this retrospective review is our limited ability to determine if certain surgical techniques were more prone to early return to OR, as the type of graft material, use of a ripcord versus tying the tube with suture, and location of the tube (anterior chamber, sulcus, or pans plana) were not consistently recorded across all the operative notes. Other limitations of this study are mainly related to its retrospective nature. Intraoperative complications were not consistently recorded in the medical record and so were not included. Additionally, since all surgeries were performed at a teaching hospital, some surgeries reviewed here may have been partly performed by assistant surgeons, including other attending surgeons, glaucoma fellows, and residents. All surgeries were performed under the direct supervision of the attending surgeon, however. Finally, while the surgeries reported here were performed by multiple experienced senior surgeons, our surgeons represent only a single center.
In the fields of general surgery, pediatric neurosurgery, and vascular surgery, unplanned return to the OR within 30 days is used as a metric of quality care7,9,10. However, for many of these surgeries, the patients are often admitted to the hospital and have much closer follow-up care than after ambulatory surgery in ophthalmology. To our knowledge, there is no validated quality care indicator for return to OR within a specific time period within the field of ophthalmology. Some studies of resident-performed cataract, vitreoretinal, and glaucoma surgeries report return to OR within 30 days2, whereas others report return to OR within 90 days1,5. In the present study, we chose to report return to OR within 180 postoperative days while also providing further stratification of the reasons for return to OR and types of reoperation surgeries within 30 and 90 days for comparison to the literature. We felt that reporting reoperation within 180 days, at any time postoperatively, and with further stratification allows us to examine both the early reoperations as well as the trends that may occur relatively later in the postoperative period to provide the best possible estimates of postoperative course both to patients and other surgeons.
Patients who ultimately returned to the OR and their matched controls had similar outcomes in terms of mean IOP, proportion of eyes at or below the target IOP, and change in VA since the time of the original tube shunt surgery, although nearly a third of eyes required additional surgery beyond the first reoperation to reach similar clinical endpoints. In fact, knowing that the IOP and VA outcomes were similar among eyes that underwent reoperation and control eyes may prompt surgeons to return to the OR sooner to address postoperative complications. Of note, patients who underwent reoperation after implantation of a non-valved tube shunt were using a greater number of glaucoma eyedrops at the last follow-up visit than control eyes to attain to similar IOP outcomes. As compared to matched controls, who could achieve their target IOP level with less medical therapy after surgery, patients who underwent unplanned return to OR may have a greater need for IOP control requiring both medical and surgical intervention.
Eyes that underwent valved and non-valved tube shunt surgery fared similarly in terms of unplanned return to OR rates both within the first 180 days and at any time postoperatively. However, the demographics of the patient populations that received valved versus non-valved tube shunts show important differences. Eyes that underwent non-valved tube shunt surgery tended to have higher rates of prior glaucoma surgery and lower mean target pressures than eyes that underwent valved tube shunt surgery, while eyes that underwent valved tube shunt surgery had higher rates of neovascular glaucoma. This may indicate that the eyes undergoing non-valved tube shunt surgery in this study were “sicker” or had more severe glaucoma than eyes that underwent valved tube shunt, yet the reoperation rates were close to identical for the two groups. These observations may show a role for the use of non-valved tube shunts in eyes with progression of severe glaucoma despite prior glaucoma surgery and/or prior trabeculectomy.
We believe that undergoing multiple surgeries to achieve similar IOP and VA outcomes for what can often be done with a single surgery is a suboptimal outcome. Monitoring rates of unplanned return to OR and subsequent reoperations may be a reasonable metric to assess surgical quality, although further evaluation regarding the correlation of unplanned return to OR and other quality metrics is required prior to its use as a surgical quality metric. Several factors may make unplanned return to the OR somewhat complicated to implement as a metric of surgical quality, including loss of patients to follow-up, differing surgeon preferences in starting with medical versus surgical management for postoperative complications, variability in complexity and severity of disease across different surgeons’ patient panels, and concerns about providers delaying or altogether avoiding reoperation to improve their quality numbers, even when it may negatively impact the patient. An important discussion in the context of this study is what the term “unplanned” return to the OR represents in the setting of a chronic, progressive disease like glaucoma. Both surgeons and patients often recognize that the ongoing treatment of glaucoma may include several surgical interventions over time. In examining unplanned return to the OR after tube shunt surgery, we have been able to characterize both these adjunctive therapies that may be performed over the course of glaucoma surgical management along with reoperations that arise from complications and failures of tube shunt surgery.
In summary, this retrospective case-control study of valved and non-valved tube shunt surgeries performed by attending glaucoma surgeons determined that more than 1 in 10 eyes undergoing tube shunt surgery returned to the OR within the first 180 postoperative days, and more than 1 in 5 eyes ultimately returned to the OR with a mean follow-up of nearly 3 years. This study provides new information on the postoperative course following tube shunt surgery that leads to reoperation, while also serving to provide a potential benchmark for the unplanned return to OR rate among attending glaucoma surgeons, both early in the postoperative period as well as over the course of years of postoperative follow-up.
Supplementary Material
Highlights.
Many patients who underwent tube shunt surgery required further operation
Valved and non-valved tube shunts had similar rates of reoperation
Reoperation and control eyes fared similarly in terms of IOP and VA endpoints
Unplanned reoperation needs further study prior to widespread use in ophthalmology
Acknowledgements:
a. Funding/Support:
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Footnotes
Financial Disclosures: Ms. Cardakli: no financial disclosures; Dr. Friedman: consulting for Bausch and Lomb, consulting for W L Gore and Associates, consulting for Center of Biomedical Research, University of Vermont, consulting for National Opinion Research Center, consulting for Life Biosciences, consulting for Thea; Dr. Boland: consulting for Carl Zeiss Meditec.
Presented as a poster at the Association for Research in Vision and Ophthalmology Annual Meeting, 2019
Supplemental Material available at AJO.com
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Menda SA, Driver TH, Neiman AE, Naseri A, Stewart JM. Return to the operating room after resident-performed cataract surgery. JAMA Ophthalmol. 2014;132(2):223–224. doi: 10.1001/jamaophthalmol.2013.5675 [DOI] [PubMed] [Google Scholar]
- 2.Schmidt CM, Sundararajan M, Biggerstaff KS, Orengo-Nania S, Coffee RE, Khandelwal SS. Indications and outcomes of resident-performed cataract surgery requiring return to the operating room. J Cataract Refract Surg. 2016;42(3):385–391. doi: 10.1016/j.jcrs.2015.11.043 [DOI] [PubMed] [Google Scholar]
- 3.Menda SA, Driver TH, Neiman AE, Blumberg S, Naseri A, Stewart JM. Risk Factors for Return to the Operating Room after Resident-Performed Cataract Surgery. Semin Ophthalmol. 2018;33(2):210–214. doi: 10.1080/08820538.2016.1208760 [DOI] [PubMed] [Google Scholar]
- 4.Stewart JM, Estrada MM, Porco TC. Incidence of unplanned return to the operating room following vitreoretinal surgery at a public teaching hospital. Curr Eye Res. 2013;38(8):886–888. doi: 10.3109/02713683.2013.780626 [DOI] [PubMed] [Google Scholar]
- 5.Hsia YC, Lee JH, Cui QN, et al. Early Reoperation Rate, Complication, and Outcomes in Resident-performed Glaucoma Surgery. J Glaucoma. 2017;26(2):87–92. doi: 10.1097/IJG.0000000000000533 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Cardakli N, Friedman DS, Boland MV. Unplanned Return to the Operating Room after Trabeculectomy. Am J Ophthalmol. Published online June 19, 2020. doi: 10.1016/j.ajo.2020.06.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Birkmeyer JD, Hamby LS, Birkmeyer CM, Decker MV, Karon NM, Dow RW. Is unplanned return to the operating room a useful quality indicator in general surgery? Arch Surg. 2001;136(4):405–411. [DOI] [PubMed] [Google Scholar]
- 8.Morris AM, Baldwin L-M, Matthews B, et al. Reoperation as a quality indicator in colorectal surgery: a population-based analysis. Ann Surg. 2007;245(1):73–79. doi: 10.1097/01.sla.0000231797.37743.9f [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Ploeg AJ, Lange CPE, Lardenoye J-W, Breslau PJ. The incidence of unplanned returns to the operating room after peripheral arterial bypass surgery and its value as indicator of quality of care. Vasc Endovascular Surg. 2008;42(1):19–24. doi: 10.1177/1538574407306793 [DOI] [PubMed] [Google Scholar]
- 10.Mukerji N, Jenkins A, Nicholson C, Mitchell P. Unplanned reoperation rates in pediatric neurosurgery: a single center experience and proposed use as a quality indicator. J Neurosurg Pediatr. 2012;9(6):665–669. doi: 10.3171/2012.2.PEDS11305 [DOI] [PubMed] [Google Scholar]
- 11.Holladay JT. Proper method for calculating average visual acuity. J Refract Surg. 1997;13(4):388–391. [DOI] [PubMed] [Google Scholar]
- 12.Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–381. doi: 10.1016/j.jbi.2008.08.010 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Harris PA, Taylor R, Minor BL, et al. The REDCap consortium: Building an international community of software platform partners. J Biomed Inform. 2019;95:103208. doi: 10.1016/j.jbi.2019.103208 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rothman AL, Rosdahl JA, Hunter TG, Challa P, Muir KW. Treatment Outcomes Following Resident Performed Nonvalved (Baerveldt 350) Glaucoma Drainage Device Implantation. J Glaucoma. 2019;28(11):958–964. doi: 10.1097/IJG.0000000000001354 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.