Although trainee cases demonstrated a 1.5- to 2-fold higher rate of intraoperative complications, postoperative visual outcomes did not differ significantly between trainees and independent surgeons.
Abstract
Purpose:
To compare complication rates and visual outcomes of cataract surgery performed by trainees vs independent surgeons.
Setting:
8 UK clinical centers.
Design:
Retrospective multicenter clinical database study.
Methods:
15-year data of 35 558 cataract surgeries undertaken by trainees and 77 131 by independent surgeons were evaluated for the rate of operative complications with emphasis on posterior capsule rupture (PCR) and logMAR distance visual acuity (VA) (defined as best available value of uncorrected or corrected VA) at 4 to 12 weeks. Recording of intraoperative complications was robust using a predetermined list of cataract surgery complications in the electronic medical record.
Results:
Trainees operated on less complex eyes, with lower rates of advanced cataracts and poor pupillary dilation (P < .001). Trainee surgeries had a significantly higher rate of PCR (2.4% vs 1.3%) compared with independent surgeons. Junior trainees had the highest PCR rate at 3.87%, compared with senior trainees at 2.12%. Using the funnel plot methodology, most surgeons seemed to approach the overall mean of PCR, 1.9%, at a surgical volume of approximately 150 cases. At 4 to 12 weeks postoperatively, there was no significant difference in the mean logMAR VA between both groups (0.197 vs 0.200, Snellen equivalent ∼20/30, P = .095).
Conclusions:
PCR should be the primary metric for assessing the quality of trainee-performed cataract surgery rather than VA outcomes. Funnel plot representation of PCR provides an equitable approach for monitoring trainees' surgical progress and peer-to-peer comparisons.
Cataract surgery is the most common surgery worldwide, with an estimated 20 million annual operations.1 This highlights its significance in ophthalmology training programs because phacoemulsification emerges as the foundational surgery technique. Ophthalmology trainees must complete a minimum number of cataract surgeries for independent practice qualification. In the United Kingdom, the General Medical Council and the Royal College of Ophthalmology require 350 independent cataract procedures for training completion.2,3 The Accreditation Council for Graduate Medical Education mandates 86 trainee cataracts in the United States.4 Trainees often surpass these thresholds, although they serve as surgical education benchmarks.
The literature across surgical disciplines has examined trainee vs independent surgeon operative experiences.5–7 These studies revealed that trainee-performed surgeries result in longer operations and higher rates of minor complications; however, surgical outcomes were comparable with those of independent surgeons.5–7 In ophthalmology, trainees' performance has shown similar intraocular pressure control rates and complications in glaucoma surgery compared with independent surgeons.8–10 For cataract surgery, notable disparities exist when comparing trainee intraoperative complication rates with independent surgeons between institutions (2.0%-13.7% vs 1.7%-4.8%, respectively).11–13 Although previous reports have discussed these rates, studies on trainee surgery are limited by their sample size or suboptimal recording of pivotal surgical complications.11–16
This study aimed to assess complication rates and visual outcomes associated with cataract surgery performed by trainees vs independent surgeons, using a large, comprehensive, multicenter database where operative complications are rigorously recorded with high accuracy.
METHODS
Data Extraction
We extracted data from 217 557 eyes that underwent cataract surgery at 8 UK National Health Service centers between 2000 and 2015. Each center used the electronic medical record (EMR) Medisoft Ophthalmology (Medisoft Ltd.), representing a broad population base. All surgeries were day-case phacoemulsification with standardized postoperative care as detailed in previous reports.17,18
The extracted data included patients' sex, laterality, visual acuity (VA), surgeon grade with identification code, risk factors including pseudoexfoliation/lens subluxation, advanced cataract, and small pupil, presence of coexisting ocular comorbidities, surgery complications, postoperative cystoid macular edema (CME), subsequent surgeries, and follow-up duration. Regarding complications, the EMR prompted surgeons to choose from a predefined list of well-recognized phacoemulsification surgery complications, and in their absence, surgeons could select “none” during surgical encounters. The study adhered to the tenets of the Declaration of Helsinki. As the extracted patient information was deidentified at the time of extraction, the study was not classified as human participant research, waiving the need for institutional review board approval.
Data Categorization and Selection Criteria
We included cataract phacoemulsification surgery performed by trainees and independent surgeons, excluding nonphacoemulsification techniques. We excluded combined surgeries with other procedures, including trabeculectomy or pars plana vitrectomy, as well as those with incomplete data (Supplementary Figure 1, available at http://links.lww.com/JRS/B497). We only included first-eye surgery to avoid the correlation that can occur from including both eyes of the same patient. We categorized eyes based on the surgeon's preoperative grade and divided them into 2 groups: trainees and independent surgeons. The trainee group included junior (senior house officer and specialty trainee year 1-2) and senior trainees (fellows, registrar, specialist registrar, specialty registrar, and specialty trainee year 3-7). The independent surgeon group included consultants, locum consultants, associate specialists, staff grade, and trust doctors.
Study Outcome Measures
Our primary outcome was the incidence of intraoperative complications, focusing on posterior capsule rupture (PCR) by subgroup. Secondary outcomes included postoperative complications and VA outcomes. We recorded VA as Snellen fractions or logMAR units as the best recorded value of uncorrected or corrected distance VA at each period. Preoperative VA was the value closest to the surgery date, no more than 3 months prior. The postoperative period was divided into 3 intervals: 0 to 4 weeks, 4 to 12 weeks, and 12 to 24 weeks for analysis. We used the 4 to 12 period as our primary visual outcome when vision stabilizes after phacoemulsification. We additionally analyzed the proportion of vision gain, defined as an increase of ≥0.3 logMAR units (∼3 Snellen lines) postoperatively when compared with preoperative VA, and the percentage of eyes that reached an absolute VA of ≤0.3 logMAR units (Snellen VA ≥20/40).
We defined CME as the documentation of CME diagnosis occurring within 90 days of surgery. The clinical centers performed imaging studies, including optical coherence tomography or fluorescein angiography, at the clinicians' discretion, typically in patients with unexpected VA outcomes after cataract surgery. Therefore, our study only reflects visually significant CME rather than subclinical disease.
Statistical Analysis
We used the SPSS software (v. 27.0.0, SPSS, Inc.). Since the study PCR and VA outcomes were normally distributed, we used t tests for analysis of continuous variables and the chi-square test for categorical comparisons. We fitted a logistic regression to analyze the predictive factors of intraoperative PCR development and calculated adjusted odds ratios (ORs). To account for the variation in the number of cataract surgeries (sample size) between surgeons, we explored individual unadjusted PCR rates using a funnel plot, with 99.8% CI boundaries or “control” limits representing 3 SDs.19 We fitted a logistic regression model to analyze predictive factors for postoperative VA outcomes of ≤0.3 logMAR units (Snellen VA ≥20/40) at 4 to 12 weeks and calculated adjusted ORs. A Bonferroni correction was applied to account for multiple VA comparisons. Results were considered significant at P < .05. Means were reported with ±1 SD or 95% CI boundaries when appropriate.
RESULTS
Baseline Characteristics
There were 112 689 phacoemulsification operations performed by 613 surgeons: 35 558 eyes (31.55%) by 477 trainees and 77 131 eyes (68.45%) by 136 independent surgeons (Supplementary Figure 1, available at http://links.lww.com/JRS/B497). The mean age of patients in the trainee and independent surgeon groups was 75.8 ± 9.1 years and 75.7 ± 9.5 years, respectively (P = .11). There was a female predominance of 21 201 (59.8%) and 45 426 (58.9%) patients, respectively. The rate of diabetes mellitus was 5875 (16.5%) for the trainees and 13 159 (17.06%) for the independent surgeons, with further subclassification in Table 1.
Table 1.
Baseline and preoperative characteristics of the study eyes among both trainees and independent surgeons
| Preop characteristic | Trainee group | Independent surgeon group | P value |
| Total eyes, n (%) | 35 558 (31.55) | 77 131 (68.45) | |
| Age (y), mean (SD) | 75.76 (9.1) | 75.66 (9.5) | .11 |
| Preop AL, mean (SD) | 23.4 (1.5) | 23.4 (1.6) | .97 |
| Laterality, n (%) | |||
| Right | 19 186 (53.95) | 41 610 (53.95) | .976 |
| Left | 16 372 (46.05) | 35 521 (46.05) | .976 |
| Sex, n (%) | |||
| Male | 14 251 (40.2) | 31 575 (41.01) | .01* |
| Female | 21 201 (59.8) | 45 426 (58.99) | .01* |
| Diabetes status, n (%) | |||
| Total diabetes | 5875 (16.5) | 13 159 (17.06) | <.001* |
| Type 1 | 472 (1.3) | 1129 (1.5) | .078 |
| Type 2 | 5402 (15.2) | 12 023 (15.6) | .114 |
| No diabetes | 24 994 (70.3) | 50 362 (65.3) | <.001* |
| Unknown | 4239 (11.9) | 12 784 (16.6) | <.001* |
| Uveitis, n (%) | 279 (0.8) | 932 (1.2) | <.001* |
| Advanced cataract, n (%) | 1065 (3) | 3530 (4.6) | <.001* |
| Pseudoexfoliation/phacodonesis, n (%) | 257 (0.7) | 1126 (1.5) | <.001* |
| Poor pupillary dilation, n (%) | 1193 (3.4) | 3213 (4.2) | <.001* |
| Corneal pathology, n (%) | 903 (2.54) | 2354 (3.05) | <.001* |
| Diabetic retinopathy, n (%) | 1552 (4.4) | 3602 (4.7) | .02* |
| ARMD, n (%) | 3420 (9.62) | 6948 (9.01) | <.001* |
| Retinal vein occlusion, n (%) | 335 (0.9) | 737 (1) | .83 |
| >10 intravitreal injections, n (%) | 36 (0.1) | 99 (0.13) | .22 |
| Glaucoma, n (%) | 2313 (6.5) | 6453 (8.4) | <.001* |
| Preop use of prostaglandin eyedrops, n (%) | 1740 (4.9) | 4937 (6.4) | <.001* |
| Previous trabeculectomy, n (%) | 74 (0.21) | 392 (0.51) | <.001* |
AL = axial length; ARMD = age-related macular degeneration
Statistically significant
Preoperative Characteristics
Patients operated by trainee surgeons had better preoperative logMAR VA of 0.656 (95% CI 0.652-0.659, ∼Snellen 20/90) compared with independent surgeons' patients (0.693, 95% CI 0.690-0.697, ∼Snellen 20/100) (P < .001) (Table 2). Preoperative axial lengths of eyes operated by trainees and independent surgeons were 23.4 ± 1.5 mm and 23.4 ± 1.6 mm, respectively (P = .97).
Table 2.
Preoperative and postoperative logMAR VA and visual gains both trainees and independent surgeons after cataract surgery
| Parameter | Trainee group (n) | Independent surgeon group (n) | P value |
| Preop VA, mean ± SD [95% CI] (n) | 0.656 ± 0.336 [0.652-0.659] (35,510) | 0.693 ± 0.495 [0.690-0.697] (76,821) | <.001* |
| Postop VA, mean ± SD [95% CI] (n) | |||
| 0-4 wk | 0.302 ± 0.138 [0.299-0.304] (11,607) | 0.326 ± 0.271 [0.323-0.329] (31,254) | <.001* |
| 4-12 wk | 0.197 ± 0.141 [0.195-0.199] (18,986) | 0.200 ± 0.208 [0.198-0.202] (41,710) | .095 |
| 12-24 wk | 0.209 ± 0.136 [0.207-0.212] (11,298) | 0.238 ± 0.192 [0.235-0.240] (22,704) | .01 |
| Eyes gaining ≥0.3 logMAR units (∼3 Snellen lines), n (%) | |||
| 0-4 wk | 6598 (56.8) | 17 579 (56.2) | <.001* |
| 4-12 wk | 12 266 (64.6) | 27 500 (65.9) | <.001* |
| 12-24 wk | 8355 (74.0) | 12 695 (55.9) | <.001* |
| Eyes with ≤0.3 logMAR vision (Snellen VA ≥20/40), n (%) | |||
| 0-4 wk | 8397 (72.3) | 21 515 (68.8) | <.001* |
| 4-12 wk | 15 863 (83.6) | 34 512 (82.7) | <.001* |
| 12-24 wk | 11 008 (97.4) | 16 328 (71.9) | <.001* |
Statistically significant
Trainee eyes had lower incidences of preoperative ocular comorbidities including uveitis (0.8% vs 1.2%), pseudoexfoliation (0.7% vs 1.5%), poor pupillary dilation (3.4% vs 4.2%), corneal pathology (2.54% vs 3.05%), diabetic retinopathy (4.4% vs 4.7%), glaucoma (6.5% vs 8.4%), preoperative use of prostaglandin eyedrops (4.9% vs 6.4%), and previous trabeculectomy surgery (0.21% vs 0.51%), respectively (all P < .02). Of note, incidence of prior retinal vein occlusion (0.9% vs 1%; P = .83) and the number of eyes receiving more than 10 intravitreal injections (0.1% vs 0.13%; P = .22) before cataract surgery were similar. The only comorbidity that was more frequent in trainee eyes was age-related macular degeneration (9.62% vs 9.01%; P < .001) (Table 1).
Intraoperative Complications
Across all surgeon groups, the overall PCR rate was 1.67%. The rate of PCR for trainees was 2.4% compared with 1.3% for independent surgeons (P < .001) (Table 3). Within independent surgeons, the PCR rate of consultants was 1.3% compared with 1.46% for others (associate specialists, staff grade, and trust doctors) (P < .001). Senior trainees had a lower PCR rate (2.12%) compared with junior trainees (3.87%) (P < .001). Our logistic regression model of PCR showed that trainees had an increased risk of PCR (OR 1.9, 95% CI 1.7-2.1), as compared with independent surgeons (Figure 1). In addition, the following patients' characteristics were predictive: A female sex also had reduced risk (OR 0.87, 95% CI 0.79-0.97), older age (60-80 years old; OR 1.4, 95% CI 1.1-1.8, and >80 years old; OR 1.8, 95% CI 1.4-2.3), diabetes (OR 1.2, 95% CI 1.0-1.3), advanced cataracts (OR 2.8, 95% CI 2.4-3.3), poor pupillary dilation (OR 1.3, 95% CI 1.04-1.6), zonular dialysis (OR 5.7, 95% CI 4.4-7.4), pseudoexfoliation (OR 2.1, 95% CI 1.6-2.8), previous pars-plana vitrectomy (OR 3.4, 95% CI 2.4-4.8), and eyes that had greater than 10 preoperative intravitreal injections (OR 4.1, 95% CI 2.0-8.4) (Supplementary Table 1, available at http://links.lww.com/JRS/B498).
Table 3.
List of intraoperative and postoperative complications among trainees and independent surgeons and surgical grade subgroup PCR rates
| Parameter | Trainees, n (%) | Independent surgeons, n (%) | P value |
| Total operative complicationsa | 2089 (5.9) | 3119 (4.0) | <.001 |
| Zonular dialysis with or without vitreous loss | 289 (0.8) | 464 (0.6) | <.001 |
| Dropped lens fragments | 107 (0.3) | 122 (0.16) | <.001 |
| Corneal edema | 64 (0.18) | 68 (0.09) | <.001* |
| Descemet membrane tear | 85 (0.24) | 92 (0.12) | <.001* |
| Phacoemulsification wound burn | 37 (0.1) | 40 (0.05) | .002* |
| Choroidal/suprachoroidal hemorrhage | 19 (0.05) | 39 (0.05) | .84 |
| Hyphema | 10 (0.03) | 33 (0.04) | .24 |
| Iris trauma | 187 (0.53) | 346 (0.45) | .08 |
| CMEb | 419 (1.18) | 863 (1.11) | .38 |
| Lens exchange at the surgery or within 3 mo postop | 10 (0.03) | 30 (0.04) | .37 |
| PCR | 862 (2.4) | 1022 (1.3) | <.001* |
| Total independent surgeon PCR rate (n = 77 131) | 1022 (1.33) | ||
| Consultant (n = 65 201) | 853 (1.31) | ||
| Locum consultant (n = 1690) | 20 (1.18) | ||
| Associate specialist (n = 7385) | 83 (1.12) | ||
| Staff grade (n = 2496) | 52 (2.08) | ||
| Trust doctor (n = 359) | 14 (3.9) | ||
| Total senior trainee PCR rate (n = 29 392) | 622 (2.12) | ||
| Fellow (n = 8663) | 204 (2.35) | ||
| Registrar (n = 35) | 4 (11.43) | ||
| Specialist registrar (n = 16 702) | 317 (1.9) | ||
| Specialty registrar (n = 542) | 19 (3.5) | ||
| Specialty trainee (year 7) (n = 10) | 1 (10) | ||
| Specialty trainee (year 6) (n = 220) | 4 (1.82) | ||
| Specialty trainee (year 5) (n = 962) | 17 (1.77) | ||
| Specialty trainee (year 4) (n = 1521) | 33 (2.17) | ||
| Specialty trainee (year 3) (n = 737) | 23 (3.12) | ||
| Total junior trainee PCR rate (n = 6018) | 233 (3.87) | ||
| Senior house officer (n = 4445) | 188 (4.23) | ||
| Specialty trainee (year 2) (n = 805) | 16 (1.99) | ||
| Specialty trainee (year 1) (n = 768) | 29 (3.78) | ||
CME = cystoid macular edema; PCR = posterior capsule rupture
Statistically significant
Total intraoperative and postoperative complications, percentages based on total number of surgeries
CME defined as the presence of the diagnosis recorded in the notes within 90 days of the phacoemulsification surgery
Figure 1.
Odds ratios from binary logistic regression for risk factors of PCR. PCR = posterior capsule rupture
Figure 2 is a funnel plot delineating the association between PCR incidents and the volume of cataract operations, adjusting for individual variations in surgical numbers. The horizontal dashed line indicates the average PCR rate across all surgeons. Notably, there existed a greater degree of variance in PCR rates among trainees than independent surgeons; nevertheless, most surgeons in both groups remained within 3 SD control limits. We observed that surgeons conducting ≥150 cases tended to aggregate closer to the mean line on the plot, implying a trend toward alignment with the overall average PCR rate, 1.9%, at this threshold of surgical volume.
Figure 2.
Funnel plot methodology for visual representation of PCR rates per surgeon, both trainees (green) and independent surgeons (black). Each dot represents 1 surgeon, with the x-axis displaying the number of cataract surgeries performed and the y-axis representing the incidence rate of PCR. PCR = posterior capsule rupture
Regarding other intraoperative complications, there were significantly higher rates of complications in trainee cases including zonular dialysis with or without vitreous loss at 0.8% in trainees vs 0.6% in independent surgeons (P < .001), dropped lens fragments at 0.3% vs 0.16% (P < .001), corneal edema at 0.18% vs 0.09% (P < .001), Descemet membrane tears in 0.24% vs 0.12% of independent surgeon cases (P < .001), and wound burns at 0.1% vs 0.05% (P = .002), respectively. For all other complications, there were no statistical differences between groups, including suprachoroidal hemorrhage (trainees: 0.05% vs independent surgeons: 0.05%), hyphema (0.03% vs 0.04%), or iris trauma (0.53% vs 0.45%) (Table 3).
Postoperative Complications
CME occurred in a small number of trainee-performed cases at 419 (1.18%), and in the independent surgeon group in 863 cases (1.11%) (P = .38). There was also no difference between the trainee and the independent surgeon group for intraocular lens exchange, 0.03% vs 0.04% (P = .37) (Table 3).
VA Outcomes
Table 2 presents the visual outcomes throughout the study period. We observed that eyes operated by trainees achieved better VA earlier, at 0 to 4 weeks postoperatively, with a logMAR VA of 0.302 ± 0.138 (Snellen equivalent ∼20/40) (95% CI 0.299-0.304), compared with 0.326 ± 0.271 (∼20/45) (95% CI 0.323-0.329) in eyes operated by independent surgeons (P < .001). However, by 4 to 12 weeks, visual outcomes between groups were similar, with trainees achieving 0.197 ± 0.141 logMAR (∼20/30) (95% CI 0.195-0.199) and independent surgeons achieving 0.200 ± 0.208 logMAR (∼20/30) (95% CI 0.198-0.202) (P = .095). At 12 to 24 weeks, trainees maintained slightly better VA with 0.209 ± 0.136 logMAR (∼20/30) (95% CI 0.207-0.212) vs 0.238 ± 0.192 (∼20/30) (95% CI 0.235-0.240) for independent surgeons (P = .01) (Figure 3).
Figure 3.
Visual representation of VA in logMAR from preoperative (represented by 0 weeks) to 12 weeks postoperatively. Represented as a curve for trainee surgeons in green and independent surgeons in black. By 12 weeks, VA was comparable between the 2 groups.
Our logistic regression model identified several predictive factors for achieving a postoperative VA of ≤ 0.3 logMAR units (Snellen VA ≥20/40) at 4 to 12 weeks. Patients aged 60 to 80 years had significantly higher odds of reaching this vision level (OR 1.4, 95% CI 1.3-1.6), followed by those older than 80 years (OR 1.3, 95% CI 1.1-1.6), compared with a reference group of patients aged 60 years or younger. Patients with better preoperative VA (OR 2.63, 95% CI 2.56-2.78), good pupillary dilation (OR 1.37, 95% CI 1.23-1.54), advanced cataract (OR 1.4, 95% CI 1.2-1.5), and absence of diabetic retinopathy (OR 2.0, 95% CI 1.92-2.33) also had higher odds of achieving this vision level. The surgeon grade was not predictive of postoperative VA.
DISCUSSION
To the authors' knowledge, this is the largest study comparing cataract surgery outcomes undertaken by trainees and independent surgeons. Despite a more complex case mix for independent surgeons and a 2-fold increase in the PCR complications in trainees' operated eyes, there was no difference in VA outcomes between the different grades of surgeons. These findings highlight the importance of effective case allocation and surgical supervision in optimizing visual outcomes during trainee surgery. It also underscores that the direct measurement of PCR rate, rather than VA outcome, should be the primary metric for assessing the quality of trainee-performed cataract surgery.
PCR correlates with significant postoperative risks, including CME, endophthalmitis, and retinal detachment, making it a critical metric for monitoring surgery quality and trainees' progression.20 Our trainees had a nearly 2-fold increase in PCR compared with independent surgeons, which aligned with previous studies.11,12,21 Rates of PCR vary from 0.8% to 10% among trainees and from 0.2% to 7.9% independent surgeons.11–14,21–26 The UK Cataract National Database found a 1.92% PCR rate across 55 000 cases from 400 surgeons.21,27 Johnston et al. showcased the lowest PCR rate in the independent surgeon group at 1.41%, while senior and junior trainees had rates of 2.48% and 5.1%, respectively.21
Using the funnel plot methodology, we adjusted for the varying case volumes of individual surgeons to analyze PCR rates. This relationship is visually represented in the funnel plot and control limits, which expand for smaller case volumes because of higher sampling error and converge as the number of cases increases. The significance of this pattern is 2-fold: First, it validates that trainees' complication rates were within expected limits, considering their professional development stage and smaller case volumes. Second, and perhaps most importantly, it provides data to bolster the morale of early-career trainees regarding their surgical outcomes, demonstrating progressive improvement in surgical safety as trainees handle an increasing number of cases. Furthermore, the funnel plot analysis provides a persuasive argument for the value of cumulative surgical experience, where most surgeons seem to approach the overall mean for PCR at a surgical volume of approximately 150 cases. As surgeons progress beyond their training and begin practicing independently, their surgical proficiency improves, leading to a reduced incidence of complications—an effect well documented across other surgical specialties.6,21 Perhaps an even higher rate of PCR in trainee surgery was mitigated by offering them cases with lower complexity. This approach for case allocation is important in the early stages of trainees' surgical training to promote skill acquisition while ensuring patient safety.11,13,21
Regarding visual recovery, the notably better VA in the early postoperative period among the trainee group may be attributed to a less complex case mix, as indicated by better preoperative VA and lower rates of advanced cataracts and small pupils. These factors likely contributed to a faster stabilization of VA postoperatively. Despite the increased risk of surgical complications in trainee-performed procedures, 6-month VA outcomes did not differ significantly between surgeon grades. Our findings build on previous studies on resident-performed cataract surgery, including those by Finn et al., who analyzed outcomes in 2487 eyes, and Xiao et al., who assessed 14 537 eyes with up to 3 months of postoperative VA data—both demonstrating comparable VA outcomes between residents and experienced surgeons.15,16
Similar to visual outcomes, CME rates remained comparable between trainees and experienced surgeons. This finding aligns with previous literature, including a single-center study of 1931 eyes by Moustafa et al.14 The similar CME rates across different surgeon grades may reflect the effective management of complications under supervision.
The limitations of this study include the retrospective nature and the limited follow-up of patients after 6 months. Although this window is unlikely to affect intraoperative complications, it may affect the visual outcomes; however, we did not identify a significant difference between the 2 groups as early as 4 to 12 weeks. An additional limitation is the nonrandomized design, which may confer inherent biases because of different ocular copathologies at baseline that we could not account for. Our VA data included a combination of uncorrected and corrected distance measurements, rather than consistently using corrected distance VA at all timepoints. This heterogeneity introduces bias; however, it is expected to be nondifferential between the trainee and independent surgeon groups. Furthermore, the data available to us date from 2000 to 2015, and examination technology and treatment of macular disease have advanced significantly in recent years. In addition, it is challenging to determine from our data how frequently trainees handed over their operations to supervising surgeons when difficulties arose.
Despite these limitations, our study has several advantages. We present a substantial dataset comprising eyes operated on by surgeons of different experience levels, with data collected as part of routine care across multiple centers, thus reflecting a nonselective and pragmatic approach. Conducting a randomized controlled trial, although the gold standard for level I evidence, would be logistically challenging and time-intensive on the scale required to compare different surgeon grades. A unique strength of this database study is the robust recording of intraoperative complications of cataract surgery using a predetermined list in the EMR, ensuring complete and accurate documentation. The mandatory requirement for surgeons to document complications before closing operative records minimizes the risk of underreporting bias. Although our study is not the first to examine trainee cataract surgery outcomes, we believe it is a definitive one. Previous studies have been hindered by relatively small sample sizes, including fellow eyes of the same patient, a lack of access to indicators of cataract surgery complexity, or reliance on indirect surrogate markers for assessing complications, such as return to the operating room. These limitations have contributed to uncertainties regarding the rates of trainee complications and the key variables for monitoring the quality of their surgical care.
In summary, our multicenter analysis of cataract surgery demonstrates that the funnel plot analysis is a valuable tool for assessing trainees' PCR rates in relation to their accumulated surgical volumes. Although trainees experienced higher intraoperative complication rates than their independent surgeon counterparts, VA outcomes were comparable across surgeon grades, indicating the safety and efficacy of trainee-performed surgery when cases are distributed appropriately, surgical supervision is provided, and complications are adequately managed. Such information could be important when consenting patients with cataract in teaching institutions and for the strategic development of surgical training curricula.
WHAT WAS KNOWN
Complications and visual outcomes in cataract surgery performed by trainees have been studied in smaller studies, and these results vary significantly among institutions.
Previous studies have been limited by their sample sizes, lacked access to indicators of cataract surgery complexity, or used imprecise surrogate markers to analyze surgical complications.
WHAT THIS PAPER ADDS
This is the most extensive comparison of trainee-performed cataract surgeries to independent surgeons, with a rigorous recording and analysis of the rate of posterior capsule rupture.
A funnel plot representation of posterior capsule rupture rates provides an equitable approach for monitoring trainees' surgical progress and facilitating peer-to-peer comparisons.
Footnotes
Disclosures: None of the authors have any financial or proprietary interest in any material or method mentioned.
First author:
Kaersti L. Rickels, BA
Department of Ophthalmology, Harvey and Bernice Jones Eye Institute, University of Arkansas for Medical Sciences (UAMS) Medical Center, Little Rock, Arkansas
Contributor Information
Kaersti L. Rickels, Email: klmclellan@uams.edu.
Abdelrahman M. Elhusseiny, Email: amelhusseiny@uams.edu.
Muhammad Z. Chauhan, Email: mzchauhan@uams.edu.
Joseph Toma, Email: joseph_touma@med.asu.edu.eg.
Abdallah A. Ellabban, Email: ellabbanabdallah@gmail.com.
REFERENCES
- 1.Cicinelli MV, Buchan JC, Nicholson M, Varadaraj V, Khanna RC. Cataracts. Lancet. 2023;401(10374):377–389. doi: 10.1016/S0140-6736(22)01839-6 [DOI] [PubMed] [Google Scholar]
- 2.General Medical Council. Ophthalmology: Specialty Specific Guidance. Accessed January 29, 2024. https://www.gmc-uk.org/-/media/documents/sat---ssg---ophthalmology---kse_pdf-103434336.pdf [Google Scholar]
- 3.Royal College of Ophthalmologists. RCOpth Guidance on Cataract Surgery Training in High Volume Settings. Accessed February 1, 2024. https://www.rcophth.ac.uk/news-views/rcophth-guidance-on-cataract-surgery-training-in-high-volume-settings/ [Google Scholar]
- 4.ACGME. ACGME Ophthalmology Case Log Information. Accessed January 29, 2024. https://www.acgme.org/globalassets/pfassets/programresources/oph_caseloginfo.pdf [Google Scholar]
- 5.D'Souza N, Hashimoto DA, Gurusamy K, Aggarwal R. Comparative outcomes of resident vs attending performed surgery: a systematic review and meta-analysis. J Surg Educ. 2016;73(3):391–399. doi: 10.1016/j.jsurg.2016.01.002 [DOI] [PubMed] [Google Scholar]
- 6.Escorel Neto AC, Sá MP, Van den Eynde J, Rotbi H, Do-Nguyen CC, Olive JK, Cavalcanti LRP, Torregrossa G, Sicouri S, Ramlawi B, Hussein N. Outcomes of cardiac surgical procedures performed by trainees vs. consultants: a systematic review with meta-analysis. J Thorac Cardiovasc Surg. 2023;166(2):612–627. doi: 10.1016/j.jtcvs.2021.12.029 [DOI] [PubMed] [Google Scholar]
- 7.Bath MF, Awopetu AI, Stather PW, Sadat U, Varty K, Hayes PD. The impact of operating surgeon experience, supervised trainee vs. trained surgeon, in vascular surgery procedures: a systematic review and meta-analysis. Eur J Vasc Endovasc Surg. 2019;58(2):292–298. doi: 10.1016/j.ejvs.2019.03.029 [DOI] [PubMed] [Google Scholar]
- 8.Thangamathesvaran L, Crane E, Modi K, Khouri AS. Outcomes of resident-versus attending-performed tube shunt surgeries in a United States residency program. J Curr Glaucoma Pract. 2018;12(2):53–58. doi: 10.5005/jp-journals-10008-1245 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Kam JP, Zepeda EM, Ding L, Wen JC. Resident-performed laser peripheral iridotomy in primary angle closure, primary angle closure suspects, and primary angle closure glaucoma. Clin Ophthalmol. 2017;11:1871–1876. doi: 10.2147/OPTH.S148467 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Zheng CX, Copparam S, Lin MM, Moster SJ, Sanvicente CT, Katz LJ, Razeghinejad MR, Myers JS, Lee D. Outcomes of trabecular microbypass surgery: comparison of resident trainees and attending surgeons. J Cataract Refract Surg. 2019;45(12):1704–1710. doi: 10.1016/j.jcrs.2019.07.021 [DOI] [PubMed] [Google Scholar]
- 11.Oliveira-Ferreira C, Leuzinger-Dias M, Tavares Ferreira J, Macedo JP, Falcão-Reis F. Cataract phacoemulsification performed by resident trainees and staff surgeons: intraoperative complications and early postoperative intraocular pressure elevation. J Cataract Refract Surg. 2020;46(4):555–561. doi: 10.1097/j.jcrs.0000000000000105 [DOI] [PubMed] [Google Scholar]
- 12.Magyar M, Sándor GL, Ujváry L, Nagy ZZ, Tóth G. Intraoperative complication rates in cataract surgery performed by resident trainees and staff surgeons in a tertiary eyecare center in Hungary. Int J Ophthalmol. 2022;15(4):586–590. doi: 10.18240/ijo.2022.04.10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Low SAW, Braga-Mele R, Yan DB, El-Defrawy S. Intraoperative complication rates in cataract surgery performed by ophthalmology resident trainees compared to staff surgeons in a Canadian academic center. J Cataract Refract Surg. 2018;44(11):1344–1349. doi: 10.1016/j.jcrs.2018.07.028 [DOI] [PubMed] [Google Scholar]
- 14.Moustafa GA, Borkar DS, McKay KM, Eton EA, Koulisis N, Lorch AC, Kloek CE; PCIOL Study Group. Outcomes in resident-performed cataract surgeries with iris challenges: results from the perioperative care for intraocular lens study. J Cataract Refract Surg. 2018;44(12):1469–1477. doi: 10.1016/j.jcrs.2018.08.019 [DOI] [PubMed] [Google Scholar]
- 15.Finn AP, Borboli-Gerogiannis S, Brauner S, Peggy Chang HY, Chen S, Gardiner M, Greenstein SH, Kloek C, Miller JW, Chen TC. Assessing resident cataract surgery outcomes using Medicare physician quality reporting system measures. J Surg Educ. 2016;73(5):774–779. doi: 10.1016/j.jsurg.2016.04.007 [DOI] [PubMed] [Google Scholar]
- 16.Xiao G, Srikumaran D, Sikder S, Woreta F, Boland MV. Assessing resident cataract surgical outcomes using electronic health record data. Ophthalmol Sci. 2023;3(2):100260. doi: 10.1016/j.xops.2022.100260 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Chu CJ, Johnston RL, Buscombe C, Sallam AB, Mohamed Q, Yang YC; United Kingdom Pseudophakic Macular Edema Study Group. Risk factors and incidence of macular edema after cataract surgery: a database study of 81984 eyes. Ophthalmology. 2016;123(2):316–323 [DOI] [PubMed] [Google Scholar]
- 18.Fouad YA, Soliman MK, Elhusseiny AM, Yang YC, Sallam AB. Visual outcomes and complications of combined versus sequential pars plana vitrectomy and phacoemulsification for epiretinal membrane. Eye (Lond). 2024;38(9):1707–1713. doi: 10.1038/s41433-024-03004-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Spiegelhalter DJ. Funnel plots for comparing institutional performance. Stat Med. 2005;24(8):1185–1202. doi: 10.1002/sim.1970 [DOI] [PubMed] [Google Scholar]
- 20.Desai P, Minassian DC, Reidy A. National cataract surgery survey 1997-8: a report of the results of the clinical outcomes. Br J Ophthalmol. 1999;83(12):1336–1340. doi: 10.1136/bjo.83.12.1336 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Johnston RL, Taylor H, Smith R, Sparrow JM. The Cataract National Dataset electronic multi-centre audit of 55,567 operations: variation in posterior capsule rupture rates between surgeons. Eye (Lond). 2010;24(5):888–893. doi: 10.1038/eye.2009.195 [DOI] [PubMed] [Google Scholar]
- 22.Joshi RS, Raza N, Wadekar P, Patil N, Tamboli S, Surwade T, Bansode N, Turankar A. To study the learning curve of capsulorhexis in manual small incision cataract surgery among postgraduate residents in Central India. Saudi J Ophthalmol. 2023;37(1):15–19. doi: 10.4103/sjopt.sjopt_135_21 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Day AC, Donachie PHJ, Sparrow JM, Johnston RL. The Royal College of Ophthalmologist's National Database: a study of cataract surgery. Eye (Lond). 2015;29(4):1711–1715 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Segers MHM, Behndig A, van den Biggelaar F, Brocato L, Henry YP, Nuijts RMMA, Rosen P, Tassignon MJ, Young D, Stenevi U, Lundström M, Dickman MM. Outcomes of cataract surgery complicated by posterior capsule rupture in the European Registry of Quality Outcomes for Cataract and Refractive Surgery. J Cataract Refract Surg. 2022;48(8):942–946 [DOI] [PubMed] [Google Scholar]
- 25.Zare M, Javadi MA, Einollahi B, Baradaran-Rafii AR, Feizi S, Kiavash V. Risk factors for posterior capsule rupture and vitreous loss during phacoemulsification. J Ophthalmic Vis Res. 2009;4(4):208–212 [PMC free article] [PubMed] [Google Scholar]
- 26.Vajpayee RB, Sharma N, Dada T, Gupta V, Kumar A, Dada VK. Management of posterior capsule tears. Surv Ophthalmol. 2001;45(6):473–488 [DOI] [PubMed] [Google Scholar]
- 27.Jaycock P, Johnston RL, Taylor H, Adams M, Tole DM, Galloway P, Canning C, Sparrow JM; UK EPR user group. The Cataract National Dataset electronic multi-centre audit of 55,567 operations: updating benchmark standards of care in the United Kingdom and internationally. Eye (Lond). 2009;23(1):38–49. doi: 10.1038/sj.eye.6703015 [DOI] [PubMed] [Google Scholar]