The Impact of Surgeon Experience on Surgical Parameters and Complication Rates for the Surgical Management of Adult Spinal Deformities: A Systematic Review and Meta-Analysis

Abstract

The surgical management of adult spinal deformities (ASDs) involves a wide variety of complex and technically challenging operative techniques. Despite numerous publications examining the relationship between surgeon experience and outcomes in ASD, no systematic review or meta-analysis exists. This first-time systematic review and meta-analysis examines the impact of surgeon experience on the surgical parameters and complication rates for the surgical management of ASD. Four databases were used for the initial search of this study from database inception until 22 September 2023. The inclusion criteria required articles that examined the outcomes for surgery for ASD, stratified outcomes by surgeon experience and/or the learning curve as a proxy for surgeon experience, and utilized adult patients (>18 years of age). Seven articles met the criteria for final inclusion. Patients in the Experienced Surgeon group had statistically significantly lower levels of EBL with no significant difference in operative time after surgery for ASD compared to patients in the Inexperienced Surgeon group via a meta-analysis of three articles. Patients in the Experienced Surgeon group had a statistically significantly lower total complication rate compared to patients in the Inexperienced Surgeon group via a meta-analysis. Increased surgeon experience resulted in lower levels of EBL, without a significant difference in the operative time after surgery for ASD.

1. Introduction

Adult spinal deformity (ASD) is a condition that represents a spectrum of degenerative spinal disorders with multi-planar scoliosis that results in significant economic, medical, and individual burden in adult patients [1,2,3,4,5,6,7]. ASD increases in prevalence with age, and an estimated twenty million adults in the United States are estimated to have some degree of spinal deformity [5,6,7]. Due to the significant impacts of ASD on patient health and quality of life, surgical management is frequently warranted for severe cases of ASD, with surgical concepts and techniques continuing to evolve over the past decade [5,6,7]. The surgical techniques utilized to treat ASD include long-segment posterolateral fusions in combination with corrective osteotomies, the placement of interbody cages, and/or the release of the anterior longitudinal ligament [8,9,10,11,12]. Indeed, because of the complexity inherent to the surgical management of ASD, there has been recent interest on the impact of surgeon experience on the outcomes and complications after surgery for ASD, with numerous studies being published in the last five years [8,10,11,12,13]. The inherent learning curve that comes with learning a new surgical technique is incredibly important to study for numerous reasons. First, this study will place increased emphasis on both resident training programs and surgical workshops to increase training for new techniques, and, secondly, acknowledging where inexperienced surgeons struggle in terms of complications (hardware, infection, etc.) will allow for increased attention on those skills in the aforementioned training programs.

While numerous systematic reviews and/or meta-analyses have described the impact of surgeon experience and learning curve on outcomes following other disciplines within orthopedic surgery, no systematic review and meta-analysis currently exists that examines this crucial topic in relation to the surgical management of ASD [14,15,16,17]. Various individual articles examine different facets of this clinical question but are either too focused on one aspect of clinical outcome or are comprehensive but limited by sample size [9,18]. Therefore, we sought to assess multiple common parameters associated with clinical outcomes for the surgical management of ASD in a pooled analysis of the available existing literature. Specifically, the purpose of this systematic review and meta-analysis was to examine how surgeon experience impacts various surgical parameters (i.e., operating room (OR) time, estimated blood loss (EBL)) and complication rates (i.e., dural tears, neurological complications) after surgery for ASD in order to guide surgeon decision making and shed light on the impact of surgeon education.

2. Materials and Methods

2.1. Study Creation and Initial Search

This study is a systematic review and meta-analysis of the clinical outcomes and complications based on surgeon experience when operating on patients with ASD. This systematic review and meta-analysis was performed under the guidance of the most recent Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Ref. [19]). Four databases—PubMed, CINAHL, MEDLINE, and Web of Science—were used for the initial search of this study from database inception until 22 September 2023. The search terms used in each database were the following: (“adult spinal deformity” OR “adult scoliosis” OR “spine deformity”) AND (“learning curve” OR “physician* experience*” OR “surgeon* experience*”).

2.2. Inclusion and Exclusion Criteria

The inclusion criteria comprised articles that examined the outcomes for surgery for ASD, stratified the outcomes by surgeon experience and/or the learning curve as a proxy for surgeon experience, utilized adult patients (>18 years of age), were comparative studies for the purpose of our meta-analysis, and had their full text in English. The exclusion criteria encompassed articles that did not examine surgery for ASD, used pediatric patients (<18 years of age), were not in English, were not comparative studies, compared the learning curves of two different types of surgery in the same article, did not have a full text, and did stratify by surgeon experience. Articles were also excluded if they focused on robotic use, such as the article by Avrumova et al. (2021), and non-surgeon experience in traditional non-robotic methods of surgery for ASD (Ref. [20]). An example of an excluded article due to comparing two different types of surgeries for ASD was Buric et al.’s paper [21].

2.3. Study Definitions

For the purposes of this study, the patients were stratified into the “Experienced Surgeon” group or the “Inexperienced Surgeon” group based on the experience level of their surgeon. All the surgeons were board-certified orthopedic surgeons, regardless of their group in this study. This was determined in various ways in each individual article, but each article represented a comparison between experienced and inexperienced surgeons. Some articles, such as the article by Choi et al. (2017), examined surgeon experience in a single surgeon over a span of four years, with earlier surgeries representing inexperience and later surgeries representing experience [9]. Other articles, such as the one by Skovrlj et al. (2015), examined multiple experienced and inexperienced surgeons by proxy via membership to the Scoliosis Research Society [18]. Finally, the article by Schupper et al. (2020) directly compared cases between a junior surgeon and a senior surgeon [13]. When studies reported multiple groups along a continuum of surgeon experience (i.e., surgeries completed from 2008 and 2015), only the first and last groups were used for comparison in this study via a meta-analysis. The total complications represented all the complication types listed in each article. Examples of neurological complications included spinal cord injury, paralysis, transient neurological deficits, and stroke. Examples of hardware complications included instrumentation failure, bilateral rod failure, implant migration, and implant malposition.

2.4. Primary Outcome Measures

The primary outcome measures for this study were surgical parameters and complication rates after surgery for ASD based on surgeon experience. The surgical parameters included EBL and operative time. The complication rates included the total complication rate, the dural tear rate, the neurological complication rate, and the hardware failure rate. Study data were too heterogenous on other forms of clinical outcomes (i.e., postoperative radiographic correction or function) for our meta-analysis. Therefore, qualitative reporting was undertaken for these variables.

2.5. Article-Sorting Process

Once all the retrieved articles were gathered from the four databases used in this study, the retrieved articles were uploaded into Rayyan, a software commonly used in the literature for article sorting in systematic reviews and meta-analyses [22]. Duplicate articles were then removed manually, and the articles were sorted by title and abstract based on the eligibility criteria. Next, full-text articles were retrieved for final full-text inclusion based on the inclusion criteria. Article sorting was completed by multiple authors.

2.6. Data Extraction Process

Data extraction was performed by a single author for this study. The data extracted included the first author, the year of publication, the type of study (comparative), the years of study inclusion, the group description (Inexperienced Surgeon group or Experienced Surgeon group), the description of surgery type for ASD, the number of patients, the average patient age, the proportion of female patients, the follow-up time, EBL, the operative time, the radiographic outcomes, the functional outcomes, and the complications (total complication rate and specific complications).

2.7. Article Quality Grading and Bias Assessment

All the included articles were graded via the Methodological Index for Non-Randomized Studies (MINORS) scale [23]. The MINORS scale has 12 criteria for comparative studies, with each criteria worth 0–2 points. A score of 24 would indicate a study with a strong methodology [23]. Based on an example in the literature, this study considered a MINORS score of 24 points to be high quality, 15–23 points to be moderate quality, and less than 15 points as low quality in terms of graded article quality [24]. Article grading was performed by a single author. Funnel plots were not created for this study due to the limitation of a small sample size of articles [25].

2.8. Statistical Considerations

The Statistical Package for the Social Sciences (SPSS) version 29.0 (Armonk, NY, USA: IBM Corp) was used for the statistical analysis in this study. Descriptive statistics and the frequency-weighted mean (FWM) were used to represent patient demographics and outcome data for ease of reading, without statistical inference. For the meta-analysis, a random-effect continuous model was utilized for the continuous variables and reported as the unstandardized mean difference (MD) for effect size with a 95% confidence internal (CI). For the binary outcomes, a random-effect binary model was used and reported as relative risk (RR) for effect size with a 95% CI. The significance level was set to 0.05 for this study. Specific data were not used in the meta-analysis if they did not have means with standard deviations for the continuous variables. Forest plots for the main outcomes were generated to visually depict the statistical relationships.

3. Results

3.1. Initial Search and Article Grading Results

A total of seven articles met the final inclusion criteria from the 122 articles initially retrieved from the four databases used in this systematic review and meta-analysis [8,9,10,11,12,13,18]. Refer to Figure 1 for the PRISMA diagram outlining the search process for this study. All the seven included articles were graded according to the MINORS scale, and the mean MINORS score (n = 7 articles) was 15.3 ± 2.6 points out a possible 24 points for comparative studies. Three articles were considered “low quality”, and four articles were considered “moderate quality”, without any “high quality” articles included in this study. Refer to

Table 1. The Methodological Index for Non-Randomized Studies (MINORS) grading for each of the included articles in this systematic review and meta-analysis.

Author (Year)	Study Type	Total MINORS Score	Clearly Stated Aim	Inclusion of Consecutive Patients	Prospective Collection of Data	End Points Appropriate to Study Aim	Unbiased Assessment of Study End Point	Follow-Up Period Appropriate to Study Aim	Less Than 5% Lost to Follow Up	Prospective Calculation of the Study Size	Adequate Control Group	Contemporary Groups	Baseline Equivalence of Groups	Adequate Statistical Analysis
Bourghli (2018)	Comparative	14	2	2	0	2	2	0	0	0	2	0	2	2
Choi (2017)	Comparative	16	2	2	0	2	2	2	0	0	2	0	2	2
Lau (2020)	Comparative	14	2	2	0	2	2	2	0	0	0	0	2	2
Raad (2019)	Comparative	17	2	2	0	2	2	2	2	0	2	0	1	2
Schupper (2020)	Comparative	19	2	2	0	2	2	2	2	0	2	2	1	2
Skovrlj (2015)	Comparative	11	2	2	0	1	1	0	0	0	1	2	1	1
Wang (2019)	Comparative	16	2	2	0	2	2	2	2	0	0	0	2	2

for the individual article grading via the MINORS scale.

3.2. Patient Demographics

The total included patients in all seven articles (n = 6403) had an FWM age of 52.7 ± 2.7 years (n = 5844; 68.9% of the patients reported), and all the patients underwent surgery for ASD. Only two studies (n = 120) reported on the follow-up time, with an FWM follow-up time of 38.2 ± 10.7 months. Of the 6403 patients, only 5290 patients (82.6% of the patients) were placed into one of two groups, with 1644 patients in the Inexperienced Surgeon group and 4276 patients in the Experienced Surgeon group. The Experienced Surgeon group (n = 4117; 96.3% of the patients reported) had an FWM age of 51.7 ± 1.6 years, whereas the Inexperienced Surgeon group (n = 1494; 90.0% of the patients reported) had an FWM age of 53.9 ± 2.2 years. Refer to

Table 2. Information on the included articles and patient demographics. The data recorded included the first author, the year of publication, the type of study, the years of study inclusion, the group description, the study grouping (Inexperienced Surgeon group versus Experienced Surgeon group), the surgery description, the number of patients, the average age, the proportion of female patients, and the follow-up time.

Author (Year)	Study Type	Study Period	Group Description	Study Group	Description of Surgery	Number of Patients (n)	Age (Mean)	Percent Female Gender	Mean Follow-Up (Months)
Bourghli (2018)	Comparative	05/2005–7/2013	First Series	Inexperienced	Lumbar Pedicle Subtraction Osteotomy (PSO) cases 1–34	34	62	74%	-
			Second Series	-	Lumbar Pedicle Subtraction Osteotomy (PSO) cases 35–68	34	60	68%	-
			Last Series	Experienced	Lumbar Pedicle Subtraction Osteotomy (PSO) cases 69–102	34	55	62%	-
Choi (2017)	Comparative	2/2012–6/2016	Group 1—the first 20 patients	Inexperienced	PSO for fixed sagittal imbalance, performed by a single surgeon	20	64.8	85%	14.4
			Group 2—the last 20 patients	Experienced		20	67.3	90%	-
Lau (2020)	Comparative	50 3COs Performed	-	Inexperienced	Adult spinal deformity surgery with 3 column osteotomies	50	-	-	-
		100 3COs Performed		-	Adult spinal deformity surgery with 3 column osteotomies	50	-	-	-
		150 3COs Performed		-	Adult spinal deformity surgery with 3 column osteotomies	50	-	-	-
		200 3COs Performed		-	Adult spinal deformity surgery with 3 column osteotomies	50	-	-	-
		250 3COs Performed		-	Adult spinal deformity surgery with 3 column osteotomies	50	-	-	-
		300 3COs Performed		-	Adult spinal deformity surgery with 3 column osteotomies	50	-	-	-
		362 3COs Performed		Experienced	Adult spinal deformity surgery with 3 column osteotomies	62	-	-	-
Raad (2019)	Comparative	2005–2014	Cases 1–100	Inexperienced	Three column osteotomies for adult spinal deformity, performed by one surgeon	100	-	70%	43
			Cases 101–197	Experienced		97	-	74%	-
Schupper (2020)	Comparative	1/1/2008–11/30/2016	Operated on by senior surgeon	Experienced	Adult spinal deformity surgery	216	55.7	62%	-
			Operated on by junior surgeon	Inexperienced		147	57.3	65.30%	-
Skovrlj (2015)	Comparative	2004–2007	Candidate members performing spinal fusion for scoliosis	Inexperienced	Scoliosis Research Society candidate membership is open to orthopedic surgeons and neurosurgeons whose practices devote 20% or more to the treatment of spinal deformities, including scoliosis, kyphosis, spondylolisthesis, and fractures. Candidate members stay in this category for 5 years, during which they must meet specific membership requirements, including submitting yearly online M&M reports and a complete 11-month case list (non-operative and operative) during the fifth year as candidate members. Although the submission of these data is a requirement to remain in good standing as a candidate member and ultimately progress to a fully active membership, whether active membership is granted is not influenced by the number or types of complications reported	1281	53.1	-	-
			Active members performing spinal fusion for scoliosis	Experienced		3836	51.4	-	-
Wang (2019)	Comparative	2008–2015	2008	Inexperienced	Less-invasive adult spinal deformity surgery; combined data from 8 tertiary spine care centers around the U.S.	12	50.67	-	-
			2009	-		26	60.81	-	-
			2010	-		37	58.99	-	-
			2011	-		41	61.77	-	-
			2012	-		35	64.26	-	-
			2013	-		31	65.74	-	-
			2014	-		29	60.07	-	-
			2015	Experienced		11	62.36	-	-

for more information on the included articles and patient demographics.

3.3. Estimated Blood Loss by Surgeon Experience

Patients in the Experienced Surgeon group (n = 298) had statistically significantly lower levels of EBL after surgery for ASD compared to patients in the Inexperienced Surgeon group (n = 217) via a meta-analysis of three articles (p = 0.005; MD = −510.53 mL; 95% CI: −869.29 mL, −151.76 mL; Figure 2). The FWM EBL during surgery for ASD for the patients in the Experienced Surgeon group (n = 298) was 1332.0 ± 273.8 mL, while the FWM EBL during surgery for ASD for the patients in the Inexperienced Surgeon group (n = 217) was 1798.3 ± 73.0 mL. Refer to

Table 3. Information on the surgical parameters and complications after surgery for adult spinal deformity. The data recorded include the first author, the year of publication, the study group, the number of patients, the estimated blood loss (EBL), the operative time, and the complications (number and type specified where reported).

Author (Year)	Study Group	Number of Patients	Total Case EBL—mL (Mean)	Total OR Time—min (Mean)	Complications
Bourghli (2018)	Inexperienced	34	2255	270	n = 8 dural tears; n = 1 operation incomplete due to severe bleeding; n = 4 deep infections; n = 2 paralyses; n = 1 PE
	-	34	2100	240	n = 1 dural tear; n = 2 hematomas; n = 3 deep infections
	Experienced	34	1600	220	n = 3 dural tears; n = 1 intraoperative death; n = 2 hematomas; n = 2 deep infections
Choi (2017)	Inexperienced	20	1777.5 ± 910.1	569.6 ± 162.4	n = 7 intraoperative complications; n = 7 perioperative complications; n = 6 late- onset complications; n = 4 additional operations
	Experienced	20	949.5 ± 629.6	392 ± 106.6	n = 1 intraoperative complications; n = 4 perioperative complications; n = 5 late- onset complications; n = 3 additional operations
Lau (2020)	Inexperienced	50	1930.8 ± 1727.5	326.1 ± 72.7	n = 15 overall complications; n = 6 surgical complications; n = 7 neurological deficits
	-	50	2093.2 ± 1516.4	316.9 ± 67	n = 15 overall complications; n = 3 surgical complications; n = 2 neurological deficits
	-	50	1759.4 ± 1151	327.5 ± 67.2	n = 13 overall complications; n = 2 surgical complications; n = 3 neurological deficits
	-	50	1966 ± 1016.6	324.8 ± 65.8	n = 10 overall complications; n = 2 surgical complications; n = 1 neurological deficits
	-	50	1975 ± 975.4	306.7 ± 67.8	n = 9 overall complications; n = 4 surgical complications; n = 1 neurological deficits
	-	50	2342 ± 1209.7	296.1 ± 53.1	n = 18 overall complications; n = 5 surgical complications; n = 3 neurological deficits
	Experienced	62	1849.2 ± 1338	283.4 ± 61	n = 8 overall complications; n = 5 surgical complications; n = 4 neurological deficits
Raad (2019)	Inexperienced	100	-	-	n = 11 new postoperative neurological deficits; n = 10 instrumentation failures; n = 6 bilateral rod failures
	Experienced	97	-	-	n = 2 new postoperative neurological deficits; n = 5 instrumentation failures; n = 5 bilateral rod failures
Schupper (2020)	Experienced	216	1219 ± 1227	378 ± 145	n = 105 cases of postprocedural hemorrhagic anemia; n = 8 AKI; n = 4 MI; n = 5 cardiac arrests; n = 4 cases of pneumonia; n = 4 PE; n = 3 wound dehiscence; n = 5 surgical site infections; n = 5 sepsis; n = 3 septic shock; n = 3 UTIs; n = 1 death; n = 4 incidental durotomies
	Inexperienced	147	1756 ± 2051	414 ± 146	n = 3 airway complications; n = 52 cases of postprocedural hemorrhagic anemia; n = 5 AKI; n = 3 MIs; n = 3 cardiac arrests; n = 3 strokes; n = 5 pneumonia; n = 4 PEs; n = 2 wound dehiscence; n = 4 surgical site infections; n = 4 sepsis; n = 3 septic shock; n = 3 UTIs; n = 3 death; n = 7 incidental durotomies
Skovrlj (2015)	Inexperienced	1281	-	-	n = 14 SCIs, n = 23 SSI superficial; n = 36 SSI deep; n = 49 dural tears; n = 4 PE; n = 6 pulmonary (other); n = 5 hematological problems; n = 9 implant failures; n = 8 implant migration; n = 10 implant malposition; n = 16 all other causes; n = 3 deaths
	Experienced	3836	-	-	n = 21 SCIs; n = 33 SSI superficial; n = 82 SSI deep; n = 125 dural tears; n = 15 PE; n = 46 pulmonary (other); n = 18 hematological problems; n = 18 implant failures; n = 23 implant migration; n = 26 implant malposition; n = 79 all other causes; n = 12 deaths
Wang (2019)	Inexperienced	12	1254.17	525.42	33.3% complications, 25.0% reoperations
	-	26	983.27	546.77	42.3% complications, 26.9% reoperations
	-	37	1153.92	577.42	59.5% complications, 40.5% reoperations
	-	41	713.78	447.2	43.9% complications, 22.0% reoperations
	-	35	809.41	541.82	42.9% complications, 25.7% reoperations
	-	31	878.87	490.81	58.1% complications, 16.1% reoperations
	-	29	673.62	401.48	48.3% complications, 20.7% reoperations
	Experienced	11	423.64	350.18	18.2% complications, 0% reoperations

for more information on EBL from individual articles.

3.4. Operative Time by Surgeon Experience

There was no significant difference in the operative time during surgery for ASD between the patients in the Experienced Surgeon group (n = 298) and the patients in the Inexperienced Surgeon group (n = 217) via our meta-analysis of three articles (p = 0.062; MD = −76.34 min; 95% CI: −156.42 min, 3.75 min; Figure 3). The FWM operative time during surgery for ASD for the patients in the Experienced Surgeon group (n = 298) was 359.26 min ± 39.10 min, while the FWM operative time during surgery for ASD for the patients in the Inexperienced Surgeon group (n = 217) was 408.09 min ± 63.31 min. Refer to Table 3 for more information on operative time from individual articles.

3.5. Radiographic Outcomes by Surgeon Experience

Bourghli et al. (2017) and Choi et al. (2017) examined pedicle subtraction osteotomies for ASD and observed that increased surgeon experience was associated with statistically significant improvements in the sagittal vertical axis (SVA) (Refs. [8,9]). Bourghli et al. (2017) reported an increase in the mean postoperative SVA of 24 ± 27 degrees to 40 ± 37 degrees (p > 0.05), whereas Choi et al. (2017) reported a significant decrease in the mean SVA of 40.1 ± 54.2 degrees to −3.6 ± 43 degrees (p = 0.008) with increased surgeon experience [8,9]. Similarly, increased surgeon experience was associated with an increased correction angle by a single-level pedicle subtraction osteotomy in the article by Choi et al. (2017) (Ref. [9]). Regarding pelvic tilt (PT) and pelvic incidence minus lumbar lordosis (PI-LL), Bourghli et al. (2017) reported non-significant increases in the PT of 15 ± 13 degrees to 18 ± 9 degrees and non-significant decreases in the PI-LL of −3 ± 16 degrees to 3 ± 9.1 degrees (Ref. [8]). Furthermore, Wang et al. (2019) found radiographic improvements for PT, SVA, pelvic incidence, PI-LL, lumbar lordosis, and the maximum Cobb angle with increases in training, but the results were not statistically significant [12].

3.6. Functional Outcomes by Surgeon Experience

Choi et al. (2017) reported significant improvements in the Scoliosis Research Society—22 (SRS-22) scores after pedicle subtraction osteotomy procedures [9]. The average preoperative SRS-22 score was 2.41 ± 0.79 points compared to 2.29 ± 0.58 points (p = 0.963), and the postoperative SRS-22 score was 2.76 ± 0.84 points compared to 2.90 ± 0.79 points (p = 0.735) in the inexperienced group and the experienced group, respectively. The improvement in the postoperative SRS-22 score was not significantly different between the groups based on surgeon experience (p = 0.395) [9].

3.7. Total Complication Rate by Surgeon Experience

The patients in the Experienced Surgeon group (n = 4276) had a statistically significantly lower total complication rate compared to patients in the Inexperienced Surgeon group (n = 1644) via our meta-analysis of seven articles (p = 0.029; RR: 0.88; 95% CI: 0.78, 0.99; Figure 4). The total complication rate for the patients in the Experienced Surgeon group was 16.2% (n = 692 total complications out of 4276 total cases), and the total complication rate for the patients in the Inexperienced Surgeon group was 22.4% (n = 369 total complications of 1644 total cases). When the large study by Skovrlj et al. (2015) was removed for sensitivity analysis due to the disproportionally large sample size, the patients in the Experienced Surgeon group (n = 440) continued to have a significantly lower total complication rate (RR: 0.83; 95% CI: 0.75, 0.92) compared to the patients in the Inexperienced Surgeon group (n = 363) via our meta-analysis of six articles (p < 0.001). The total complication rate for the Experienced group was 44.1% (n = 194 total complications out of 440 total cases), and the total complication rate for the Inexperienced group was 51.2% (n = 186 total complications of 363 total cases). Refer to Table 3 for more information on the complications.

3.8. Dural Tear Rate by Surgeon Experience

There was no statistically significant difference in the rate of dural tears after surgery for ASD between the patients in the Experienced Surgeon group (n = 4106) compared to the patients in the Inexperienced Surgeon group (n = 1482) via our meta-analysis of four articles (p = 0.201; RR: 0.98; 95% CI: 0.95, 1.01). The dural tear rate for the patients in the Experienced Surgeon group was 3.2% (n = 133 cases out of 4106 total cases), and the dural tear rate for the patients in the Inexperienced Surgeon group was 4.7% (n = 69 cases of 1482 total cases).

3.9. Hardware Complication Rate by Surgeon Experience

There was also no significant difference in the rate of hardware complications after surgery for ASD between the patients in the Experienced Surgeon group (n = 3964) compared to the patients in the Inexperienced Surgeon group (n = 1413) via our meta-analysis of four articles (p = 0.479; RR: 0.99; 95% CI: 0.95, 1.02). The hardware complication rate for the patients in the Experienced Surgeon group was 2.0% (n = 79 cases of 3964 total cases), and the hardware complication rate for the patients in the Inexperienced Surgeon group was 3.3% (n = 47 cases out of 1413 total cases).

3.10. Neurological Complication by Surgeon Experience

There was also no statistically significant difference in the rate of neurological complications after surgery for ASD between the patients in the Experienced Group (n = 4276) compared to the patients in the Inexperienced Surgeon group (n = 1644) via our meta-analysis of seven articles, despite a trend towards significance (p = 0.051; RR: 0.97; 95% CI: 0.94, 1.00). The neurological complication rate for the patients in the Experienced Surgeon group was 0.7% (n = 28 cases of 4276 total cases), and the neurological complication rate for the patients in the Inexperienced Surgeon group was 2.4% (n = 39 cases out of 1644 total cases).

4. Discussion

This study is the first systematic review and meta-analysis evaluating the impact of surgeon experience on the surgical outcome parameters and complication rates after the operative management of ASD in the literature to date. Increased attention and insight into the treatment of ASD, both operative and non-operative, is essential, as ASD can impact up to 68% of the geriatric population and is only expected to increase [5]. Furthermore, ASD can lead to significant individual disability as well as collective economic burden, thus necessitating optimized care [3,4,5,6,7]. As rapid advances in the surgical management of ASD continue to attempt to improve patient care, it is interesting to explore how individual surgeon experience is able to keep up with these advances, as ASD operations are often lengthy, complex, and technically challenging [3,8,11]. Overall, the results of this systematic review and meta-analysis indicate that certain surgical parameters, such as the EBL, and the overall complication rates are inversely associated with surgeon experience.

Our subgroup analysis showed that specific neurological complications, such as paralysis and spinal cord injury (SCI), were just outside the threshold for statistical significance as a function of surgeon experience in this study (0.7% experienced group versus 2.4% inexperienced; p = 0.051; RR: 0.97). We hypothesize that this association was underpowered due to the relatively low incidence of neurological complications after ASD correction. This observation may also be true for the other subgroup analyses of specific complications after surgery for ASD, such as the rate of dural tears and the hardware complications after surgery. In this study, the patients treated by experienced surgeons had absolute lower rates of dural tears than those treated by inexperienced surgeons (3.2% versus 4.7%; p = 0.201), as well as fewer hardware complications (2.0% versus 3.3%; p = 0.479), but these associations lacked statistical significance. It is unclear whether any significant association between these specific complications and surgeon experience found in future studies with larger sample sizes would be clinically meaningful, given the low incidence of these complications. This remains to be elucidated in future studies, as complications after surgery for ASD can be devastating [3,18].

In terms of the surgical parameters, it is noteworthy that this study found that EBL, but not operative time, was associated with surgeon experience. Experienced surgeons had significantly lower EBL (mean difference 510.53 mL) than inexperienced surgeons following ASD correction surgery. As with the trend in specific complications such as neurological complications after surgery for ASD, this study also demonstrated a trend in decreased operative time (MD = −76.34 min; 95% CI: −156.42 min, 3.75 min) in the experienced surgeons compared to the inexperienced surgeons, although this association did not quite reach statistical significance (p = 0.062). However, it should be noted that, while the meta-analysis for the total complication rates had relatively large sample sizes (between 1000 and 5000 patients in each group), the meta-analysis for EBL and the operative time had much smaller sample sizes (less than 300 patients in each group). Therefore, while this study builds on the individual power limitations of each included article, this meta-analysis itself has a power limitation in terms of these parameters but does show trends that may reach statistical significance in future studies. In terms of other parameters, such as radiographic outcomes and function via the SRS-22, the results were too heterogenous to allow for a meta-analysis in this study. The radiographic outcomes were mixed in the individual articles, with only some of the studies showing significant improvements with increased surgeon experience. Similarly, Choi et al. (2017) reported no significant improvement in function with increased surgeon experience [9].

Overall, this systematic review and meta-analysis appears to indicate, based on low- and moderate-quality evidence observational studies, that relative surgeon inexperience is associated with increased EBL and total complication rates after surgery for ASD. Surgeons must progress through the learning curve in their practice when operating on patients for ASD. Adequate fellowship training as well as mentorship practices with junior and senior surgeons are possible ways to help alleviate the impact of the learning curve in spinal surgery [26]. Furthermore, strategies such as dual-surgeon teams have been utilized in other surgical interventions, such as during pediatric spinal deformity surgical interventions, and may need consideration [27,28,29].

There are multiple limitations that impact the generalizability of the results of this study. First, this study only included retrospective studies, as there are no prospective studies or randomized clinical trials in the literature to date. In order to improve the quality of these studies, only comparative studies evaluating the impact of surgeon experience on various outcomes after surgery for ASD were included. This study also sought to build upon the results of the large study included in this meta-analysis by Skovrlj et al. (2015), in which the Scoliosis Research Society registry for complications after surgery for ASD was examined as a function of surgeons’ experience levels via the proxy of active or candidate membership. The current systematic review and meta-analysis examined the total complication rate with and without this large study via a sensitivity analysis and determined that an increased total complication rate was consistently associated with decreased surgeon experience. This sensitivity analysis is a strength of this study, because large registry studies, such as the one by Skovrlj et al. (2015), can be prone to underreporting [18]. This fact is evident, as Skovrlj et al. (2015) reported a total complication rate of 14.2% in the patients treated by inexperienced surgeons and a total complication rate of 12.9% in the patients treated by experienced surgeons [18]. However, the pooled total complication rates were 22.4% and 16.2% with the above-mentioned study included and 44.1% and 51.2% without it for the patients treated by inexperienced and experienced surgeons, respectively [18].

Another limitation of this study was the variability in the types of data collection in the included articles. Some articles focused on a single surgeon over a longer period of time as they moved from inexperienced to experienced, while other articles focused on multiple surgeons with different experience levels. Furthermore, the studies reported surgeons’ experience levels in different ways, despite presenting these as binary demographics. These metrics of assessing the learning curve and subsequent surgeon experience, while similar, are not identical, which could lead to the differing interpretation of results in this study. Furthermore, another limitation is that multiple different types of surgical interventions were examined in this study, and surgical heterogeneity, as well as differing techniques between individual surgeons, could contribute to discrepancies in the data. Overall, more high-powered studies examining the surgeon learning curve for ASD are necessary to understand these procedural subset associations with the clinical outcomes.

5. Conclusions

Increased surgeon experience was significantly associated with decreased levels of total complications after surgery for ASD via a meta-analysis of seven low-to-moderate-quality observational studies. Furthermore, increased surgeon experience resulted in lower levels of EBL without any significant difference in operative time after surgery for ASD. Importantly, this study may be underpowered regarding operative time and specific complication rates, such as the dural tear rate, hardware complications, and neurological complications, which did not show statistical significance but had absolute lower values for the patients treated by experienced surgeons. This study represents the first systematic review and meta-analysis on the topic of surgeon experience and surgical outcomes after surgery for ASD. More research is needed to further explore the learning curve as well as ameliorate the necessary fact of inexperience that occur early during the practice of a spine surgeon.