Are all high-volume surgeons equally good outcome? A survival analysis of patients with oral cavity squamous cell carcinoma in relation to surgical margin status based on propensity score matching analysis

doi:10.21203/rs.3.rs-3642108/v1

Background

In this retrospective study, we assessed whether Taiwanese high-volume surgeons performing oral cavity squamous cell carcinoma (OCSCC) removal may differ in terms of margin status, and examined how this variable – as a quality standard – could have an impact on clinical outcomes after adjustment for clinicopathological risk factors and treatment modalities.

Methods

On analyzing a nationwide dataset, margins < 5 mm (including positive margins) were identified in 49.5% (6927/13984) of patients with OCSCC. We subsequently identified the surgeon with the highest absolute volume (number of operated patients = 560), who was located below the mean value (49.5%). Among surgeons above the mean, we identified the two surgeons with the highest volumes (termed as Surgeon 2 and Surgeon 3). The number of patients and survival operated by Surgeon 2 and Surgeon 3 were similar (number 229 and 221, respectively, totaling 450 patients; 5-year overall survivals [OSs] 64% and 65%, respectively) and thus they were grouped together for the purpose of analysis (Surgeons 2–3).

Results

The patient proportion of margins ≥ 5 mm was markedly higher in Surgeon 1 than Surgeons 2 − 3 (75.4% and 22.5%, respectively). Compared to Surgeons 2 − 3, the tumor severity was higher in Surgeon 1 (mainly a higher frequency of pT4a status, p-Stage IV, and poorly differentiated tumor). The clinical outcomes of patients treated by Surgeon 1 were more favorable than those treated by Surgeons 2 − 3 and these survival differences were even more pronounced after adjusting for baseline differences using propensity score matching (before propensity score: disease-specific survival [DSS], 83%/70%, p < 0.0001; OS, 77%/64%, p < 0.0001; after propensity score: DSS, 87%/68%, p < 0.0001; OS, 81%/63%, p < 0.0001). When patients were stratified according to the operating surgeon and margin status, we found that Surgeon 1 consistently outperformed Surgeons 2 − 3 in terms of clinical outcomes, regardless of surgical margin status (i.e., ≥ 5 mm, < 5 mm, and positive margins).

Conclusions

We conclude that the achievement of clear margins (≥ 5 mm) is a surgeon-dependent variable which is not necessarily related to hospital and/or surgeon volumes, therefore surgeons should strive to achieve adequate margins to optimize survival in OCSCC.

Oral cavity squamous cell carcinoma

Surgeons’ volumes

Cancer registry

Propensity score matching

Survival outcomes

Centralization of activities in the field of surgical oncology is recommended whenever a volume-outcome relationship exists [1, 2]. Previous studies have reported statistically significant associations between surgical volumes and survival outcomes for both head and neck cancer [3] and oral cavity squamous cell carcinoma (OCSCC) [4–6]. Currently, surgical excision remains the mainstay of treatment for OCSCC – eventually followed by adjuvant therapy in patients harboring unfavorable pathological risk factors (RFs) [7, 8]. Surgical margins and neck nodal yield are both surgeon-dependent RFs, and surgical volume has been positively correlated with wider surgical margins and higher neck nodal yields [9, 10]. However, the prognostic impact of surgical margins in patients with OCSCC who had been treated by high-volume surgeons is less well defined [4–6, 9].

Oral malignancies are significant public health concerns in Asian countries. In an effort to maintain high-quality standards in the field of OCSCC surgery, the Taiwan Health Promotion Administration (THPA) has recently issued recommendations for the achievement of both negative margin status (≥ 5 mm) and an adequate neck node yield (≥ 15 lymph nodes) [11]. However, the question as to whether these surgeon-dependent quality endpoints are invariably associated with surgeons’ volume remains unanswered.

In this retrospective nationwide study, we assessed whether Taiwanese high-volume surgeons performing OCSCC removal may differ in terms of margin status, and examined how this variable – as a quality standard – could have an impact on clinical outcomes after adjustment for clinicopathological RFs and treatment modalities [8].

Data sources

Patient-level data were collected from the Taiwanese Cancer Registry Database (TCRD) “long-form”, a nationwide computerized database that contains longitudinal clinical information on cancer stage, disease relapses, and treatment modalities. This dataset also includes pathological data on tumor depth of invasion (DOI), extra-nodal extension (ENE) and margin status recorded since 2011. Clinical and pathological data are systematically recorded from all patients with malignancies admitted to major Taiwanese hospitals. At the present writing, the TCRD collects and organizes data of the vast majority (> 99%) of Taiwanese patients with OCSCC. As we were interested in survival outcomes, we also obtained follow-up data from the Taiwanese National Health Insurance Research Dataset (TNHIRD). The current study follows the reporting recommendations for tumor marker prognostic studies (REMARK) [12, 13]. Ethical approval was granted by the Chang Gung Medical Foundation Institutional Review Board (reference number: 201801398B0A3). The need for patient consent was waived due to the retrospective study design.

Treatment protocol and follow-up protocol

As part of its continued effort to improve the quality of cancer care, the THPA has taken initiative to promote multidisciplinary team care (MDTC) and multidisciplinary case management as of April 2003. All of the Taiwanese hospital specialized in treating OCSCC began implementing an MDTC approach as of January 2004. In general, the follow-up protocols are in accordance with the National Comprehensive Cancer Network (NCCN) treatment guidelines [7].

Patient selection

Eligible medical records were drawn from adults who were diagnosed with first primary oral cavity squamous cell carcinoma between 2011 and 2018, based on the International Classification of Diseases for Oncology, Third Edition (ICD-O-3) codes. Exclusion criteria were as follows: previous history of cancer (n = 11980), initial treatment with non-surgical modalities (n = 4616), unknown pathological stage (n = 602), and unavailable information on tumor depth, surgical margins, and ENE (n = 3988). Patients with missing data on pathological nodal metastases (n = 3655), tumor differentiation (n = 122), and treating surgeons (n = 553) were also excluded. The final study cohort consisted of 13984 patients treated by 702 different surgeons. Figure 1 depicts the flow of patients through the study. The follow-up duration was defined as the time elapsed between surgery and the date of death or date of study termination (December 2019).

Surgeon-based analysis of surgical margin status

Figure 2 shows the frequencies of patients with surgical margins < 5 mm plotted against the number of cases operated by each surgeon. Margins of less than 5 mm (including positive margins) were identified in 49.5% (6927/13984) of the study participants. On analyzing the graph, we identified as an outlier the surgeon with the highest absolute volume (number of operated patients = 560). This surgeon was termed Surgeon 1 and was located below the dashed line denoting the mean value (49.5%). Above the same line, we identified the two surgeons with the highest volumes (termed as Surgeon 2 and Surgeon 3). Notably, the prevalence of margins < 5 mm was markedly lower in Surgeon 1 than in Surgeon 2 and 3, prompting the need for a comprehensive assessment of the prognostic impact of surgical margins in OCSCC between high volume surgeons. The number of operated patients by Surgeon 2 and Surgeon 3 was very similar (229 and 221, respectively, totaling 450 patients). Since an analysis of 5-year disease-specific survival (DSS) and overall survival (OS) rates did not show significant intergroup differences (5-year DSS: 72%/69%, p = 0.3474; 5-year OS: 64%/65%, p = 0.7681, respectively; Supplementary Fig. 1), Surgeon 2 and Surgeon 3 were grouped together for the purpose of analysis and denoted as Surgeons 2 − 3. The prevalence rates of both positive margins and margins < 5 mm were markedly lower for Surgeon 1 than for Surgeons 2 − 3 (2.5%/22.1% versus 8.2%/69.3%, respectively), we subsequently focused on data obtained from these three high volume surgeons.

Data collection

Patients were originally staged according to the AJCC Staging Manual, seventh edition (2010). By integrating information on DOI and ENE [14], staging was subsequently updated using the AJCC Staging Manual, eight edition (2018). Data for this study were extracted from the 2018 release of TCRD and the 2019 release of TNHIRD. Analyses were conducted in August 2022. The TCRD is compliant with the tenets set forth in the Standards for Oncology Registry Entry (STORE) manual [15]. Information on OCSCC-related morbidity and mortality obtained from the TNHIRD was used for the calculation of DSS and OS.

Statistical analysis

The 5-year DSS and OS rates were the main outcomes of interest. Because of concerns of confounding by baseline differences across different surgeons, the propensity score (PS) matching method was used to balance patient characteristics using logistic regression. With the exception of surgical margins, all covariates had a standardized mean difference (SMD) < 0.1 after PS matching, suggesting an adequate balance for all of the remaining variables. The clinical outcomes of patients operated by Surgeon 1 versus Surgeons 2 − 3 were compared before and after PS matching. Survival probabilities were plotted with the Kaplan-Meier method and compared with the log-rank test. Univariable and multivariable Cox proportional hazards regression analyses were employed to examine the associations between the study variables and survival endpoints. We calculated hazard ratios (HRs) and 95% confidence intervals (CIs) using a multivariable stepwise selection procedure, adjusting for all of the variables included in univariable analysis. Data management and analyses were carried out using R, version 4.0.2 (R Foundation for Statistical Computing, Vienna, Austria) and SAS, version 9.4 (SAS Institute Inc., Cary, NC, USA). Values of p < .05 (two-tailed) were taken as significant.

Patient characteristics

Table 1 shows the general characteristics of the study participants according to the operating surgeon (Surgeon 1 versus Surgeons 2 − 3). Before PS matching, the following RFs showed significant intergroup differences: margin status, tumor subsite, sex, pathological T status, pathological stage, tumor differentiation, neck nodal yield, and treatment modality. Patients treated by Surgeon 1 had more severe disease (mainly in terms of a higher frequency of pT4a status, p-Stage IV, and poorly differentiated tumor) compared with those operated by Surgeons 2 − 3. After PS matching (n = 326 in each group), an adequate balance (SMD < 10%) was achieved for all variables, the only exception being margin status.

Table 1

General characteristics of patients with oral cavity squamous cell carcinoma operated by surgeon 1 and surgeons 2–3 before (n = 1010) and after (n = 652) propensity score matching
Characteristics (n, % [patients before PS matching])		Before propensity score matching				After propensity score matching
Characteristics (n, % [patients before PS matching])		Surgeon 1 (n = 560)	Surgeons 2–3 (n = 450)	SMD (%)	p ^a	Surgeon 1 (n = 326)	Surgeons 2–3 (n = 326)	SMD (%)	p ^b
Margin status					< 0.0001				< 0.0001
	Positive (51, 5.0)	14 (2.5)	37 (8.2)	-25.61		8 (2.5)	32 (9.8)	-31.05
	< 5 mm (436, 43.2)	124 (22.1)	312 (69.3)	-107.55		60 (18.4)	220 (67.5)	-114.17
	≥ 5 mm (523, 51.8)	422 (75.4)	101 (22.5)	124.76		258 (79.1)	74 (22.7)	136.78
Tumor subsite					< 0.0001				0.9702
	Lip (34, 3.4)	20 (3.6)	14 (3.1)	2.56		7 (2.1)	9 (2.8)	-3.97
	Tongue (478, 47.3)	226 (40.4)	252 (56.0)	-31.7		169 (51.8)	170 (52.1)	-0.61
	Gum (107, 10.6)	74 (13.2)	33 (7.3)	19.46		25 (7.7)	27 (8.3)	-2.26
	Mouth floor (34, 3.4)	16 (2.8)	18 (4.0)	-6.28		10 (3.1)	11 (3.4)	-1.74
	Hard palate (13, 1.3)	2 (0.4)	11 (2.4)	-17.83		2 (0.6)	3 (0.9)	-3.52
	Buccal (264, 26.1)	167 (29.8)	97 (21.6)	19		88 (27.0)	86 (26.4)	1.39
	Retromolar (59, 5.8)	44 (7.9)	15 (3.3)	19.78		16 (4.9)	13 (4.0)	4.46
	Other sites (21, 2.1)	11 (1.9)	10 (2.3)	-1.8		9 (2.8)	7 (2.1)	3.97
Sex					0.0023				0.6394
	Men (896, 88.7)	512 (91.4)	384 (85.3)	19.11		289 (88.7)	286 (87.7)	2.85
	Women (114, 11.3)	48 (8.6)	66 (14.7)	-19.11		37 (11.3)	40 (12.3)	-2.85
Age (years) (Mean ± SD)		53.98 ± 10.58	55.38 ± 11.15	-12.89	0.0415	53.76 ± 10.25	54.58 ± 10.50	-7.90	0.2685
Pathologic T status					0.0017				0.6082
	T1 (183, 18.1)	83 (14.8)	100 (22.2)	-19.14		63 (19.3)	58 (17.8)	3.95
	T2 (347, 34.4)	191 (34.1)	156 (34.7)	-1.18		120 (36.8)	118 (36.2)	1.27
	T3 (176, 17.4)	94 (16.8)	82 (18.2)	-3.78		52 (16.0)	59 (18.1)	-5.72
	T4 (304, 30.1)	192 (34.3)	112 (24.9)	20.7		91 (27.9)	91 (27.9)	0
Pathologic N status					0.1486				0.1855
	pN0 (664, 65.7)	381 (68.0)	283 (62.9)	10.84		220 (67.5)	215 (66.0)	3.26
	pN1 (102, 10.1)	47 (8.4)	55 (12.2)	-12.62		35 (10.7)	26 (8.0)	9.49
	pN2 (105, 10.4)	54 (9.6)	51 (11.3)	-5.52		26 (8.0)	35 (10.7)	-9.49
	pN3 (139, 13.8)	78 (14.0)	61 (13.6)	1.08		45 (13.8)	50 (15.3)	-4.35
Pathologic stage					0.0473				0.8861
	I (159, 15.7)	75 (13.4)	84 (18.7)	-14.41		57 (17.5)	54 (16.6)	2.45
	II (253, 25.0)	145 (25.9)	108 (24.0)	4.38		84 (25.8)	82 (25.2)	1.41
	III (167, 16.5)	86 (15.4)	81 (18.0)	-7.09		52 (16.0)	53 (16.3)	-0.83
	IV (431, 42.8)	254 (45.3)	177 (39.3)	12.21		133 (40.7)	137 (41.9)	-2.49
Tumor differentiation					< 0.0001				0.7692
	Well (224, 22.2)	104 (18.6)	120 (26.7)	-19.44		74 (22.7)	78 (23.9)	-2.9
	Moderately (687, 68.0)	379 (67.7)	308 (68.4)	-1.64		227 (69.6)	226 (69.3)	0.67
	Poorly (99, 9.8)	77 (13.7)	22 (4.9)	30.84		25 (7.7)	22 (6.8)	3.56
Depth of invasion (mm)					0.3329				0.4795
	< 10 (535, 53.0)	289 (51.6)	246 (54.7)	-6.13		182 (55.8)	176 (54.0)	3.7
	≥ 10 (475, 47.0)	271 (48.4)	204 (45.3)	6.13		144 (44.2)	150 (46.0)	-3.7
Extra-nodal extension					0.1632				0.1495
	No (860, 85.1)	469 (83.8)	391 (86.9)	-8.88		285 (87.4)	276 (84.7)	7.97
	Yes (150, 14.9)	91 (16.2)	59 (13.1)	8.88		41 (12.6)	50 (15.3)	-7.97
Neck nodal yield					< 0.0001				1.000
	< 15 (85, 8.4)	12 (2.1)	73 (16.2)	-50.27		10 (3.1)	10 (3.1)	0.00
	≥ 15 (925, 91.6)	548 (97.9)	377 (83.8)	50.27		316 (96.9)	316 (96.9)	0.00
Treatment modality					< 0.0001				0.3248
	S alone (562, 55.6)	277 (49.5)	285 (63.3)	-28.25		196 (60.1)	188 (57.7)	4.99
	S plus CT S plus RT S plus CT and RT (448, 44.4)	283 (50.5)	165 (36.7)	28.25		130 (39.9)	138 (42.3)	-4.99
^a Chi-square test, T test, ^b McNemar’s test, pair T test. Abbreviations: SD, standard deviation; PS, propensity score; SMD, standardized mean difference; S, surgery; CT, chemotherapy; RT, radiotherapy.

Univariable and multivariable Cox regression analysis

The results of univariable analyses are shown in Table 2. Multivariable analyses pinpointed several factors independently linked to a decreased 5-year DSS rate: Surgeons 2 − 3 (versus Surgeon 1; HR = 2.61, p < 0.0001), pT2-4, pN1-3, and moderately-to-poorly differentiated tumors; and a decreased 5-year OS: Surgeons 2 − 3 (versus Surgeon 1; HR = 2.26, p < 0.0001), pT2-4, pN1-3, and moderately-to-poorly differentiated tumors. Of note, factor of margin status was a significant risk factor in univariable analysis, but not in multivariable analysis.

Table 2

Univariable and multivariable analyses of risk factors for 5-year disease-specific and overall survival in patients with oral cavity squamous cell carcinoma (n = 1010) operated by surgeon 1 and surgeons 2–3 Abbreviations: HR, hazard ratio; CI, confidence interval; S, surgery; CT, chemotherapy; RT, radiotherapy.
Risk factor		Disease-specific survival				Overall survival
		Univariable analysis		Stepwise multivariable analysis		Univariable analysis		Stepwise multivariable analysis
		HR (95% CI)	p	HR (95% CI)	p	HR (95% CI)	p	HR (95% CI)	p
Surgeon
	Surgeon 1	1		1		1		1
	Surgeons 2–3	1.87 (1.43–2.45)	< 0.0001	2.61 (1.97–3.46)	< 0.0001	1.68 (1.33–2.13)	< 0.0001	2.26 (1.76–2.90)	< 0.0001
Margin status (mm)
	Positive margin	3.30 (2.04–5.32)	< 0.0001		ns	2.73 (1.74–4.27)	< 0.0001		ns
	< 5	1.73 (1.31–2.29)	0.0001	-	ns	1.54 (1.20–1.96)	0.0006	-	ns
	≥ 5	1		-		1		-
Tumor subsite
	Lip	1		-		1		-
	Tongue	2.32 (0.86–6.30)	0.0976	-	ns	1.80 (0.79–4.08)	0.1592	-	ns
	Gum	1.93 (0.66–5.63)	0.2273	-	ns	1.94 (0.81–4.65)	0.1378	-	ns
	Mouth floor	1.62 (0.46–5.76)	0.4523	-	ns	1.42 (0.49–4.10)	0.515	-	ns
	Hard palate	2.16 (0.48–9.65)	0.3134	-	ns	2.59 (0.08–8.50)	0.1159	-	ns
	Buccal	1.81 (0.65–4.99)	0.2552	-	ns	1.50 (0.65–3.46)	0.3423	-	ns
	Retromolar	2.29 (0.76–6.91)	0.1404	-	ns	1.93 (0.77–4.84)	0.1589	-	ns
	Other sites	2.58 (0.73–9.14)	0.1422	-	ns	2.70 (0.96–7.60)	0.0593	-	ns
Sex
	Men	0.95 (0.63–1.44)	0.7992	-	ns	1.12 (0.75–1.65)	0.5852	-	ns
	Women	1		-		1
Age (years)		1.006 (0.994–1.019)	0.3203			1.016 (1.005–1.027)	0.0044
Pathologic T status
	T1	1		1		1		1
	T2	2.35 (1.31–4.20)	0.0042	1.90 (1.05–3.42)	0.0335	1.95 (1.23–3.11)	0.0048	1.67 (1.04–2.68)	0.0326
	T3	4.52 (2.51–8.14)	< 0.0001	2.96 (1.62–5.43)	0.0004	3.26 (2.02–5.26)	< 0.0001	2.36 (1.44–3.87)	0.0007
	T4	5.24 (2.99–9.19)	< 0.0001	3.40 (1.89–6.12)	< 0.0001	4.24 (2.71–6.63)	< 0.0001	3.12 (1.95-5.00)	< 0.0001
Pathologic N status
	pN0	1		1		1		1
	pN1	3.00 (1.97–4.58)	< 0.0001	2.18 (1.42–3.34)	0.0004	2.26 (1.54–3.31)	< 0.0001	1.76 (1.20–2.60)	0.0041
	pN2	3.70 (2.49–5.50)	< 0.0001	2.18 (1.42–3.34)	< 0.0001	2.95 (2.07–4.20)	< 0.0001	2.26 (1.57–3.25)	< 0.0001
	pN3	7.84 (5.70-10.77)	< 0.0001	5.22 (3.70–7.37)	< 0.0001	6.38 (4.81–8.45)	< 0.0001	4.67 (3.44–6.34)	< 0.0001
Pathologic stage
	I	1		-		1		-
	II	1.90 (0.89–4.03)	0.0964	-	ns	1.55 (0.88–2.73)	0.1299	-	ns
	III	3.77 (1.81–7.88)	0.0004	-	ns	2.53 (1.44–4.46)	0.0013	-	ns
	IV	8.18 (4.17–16.02)	< 0.0001	-	ns	5.56 (3.38–9.14)	< 0.0001	-	ns
Tumor differentiation
	Well	1		1		1		1
	Moderately	2.63 (1.68–4.10)	< 0.0001	2.18 (1.38–3.44)	0.0008	1.89 (1.34–2.67)	0.0003	1.61 (1.13–2.29)	0.0083
	Poorly	5.27 (3.12–8.88)	< 0.0001	3.93 (2.27–6.82)	< 0.0001	3.17 (2.04–4.91)	< 0.0001	2.36 (1.48–3.76)	0.0003
Depth of invasion (mm)
	< 10	1		-		1		-
	≥ 10	2.79 (2.10–3.70)	< 0.0001	-	ns	2.42 (1.90–3.10)	< 0.0001	-	ns
Extra-nodal extension
	No	1		-		1		-
	Yes	4.04 (3.06–5.33)	< 0.0001	-	ns	3.85 (2.99–4.96)	< 0.0001	-	ns
Neck nodal yield
	< 15 nodes	1		-		1		-
	≥ 15 nodes	1.33 (0.77–2.28)	0.3025	-	ns	1.08 (0.69–1.68)	0.7433	-	ns
Treatment modality
	S alone	1		-		1		-
	S plus CT S plus RT S plus CT and RT	2.92 (2.21–3.87)	< 0.0001	-	ns	2.42 (1.90–3.08)	< 0.0001	-	ns

Outcome analysis

Before PS matching, the 5-year DSS and OS rates for patients operated by Surgeon 1 were higher than those observed for patients operated by Surgeons 2 − 3 (83%/70%, p < 0.0001; and 77%/64%, p < 0.0001, respectively; Fig. 3A-B). In addition, the 5-year DSS and OS rates for patients with clear margins (≥ 5 mm) were higher than those observed for those with close or positive margins (83%/73%/54%, p < 0.0001; and 77%/67%/50%, p < 0.0001, respectively; Fig. 3C-D). These results did not appreciably change in the PS-matched cohort. Specifically, the 5-year DSS and OS rates for patients operated by Surgeon 1 were higher than those observed for patients operated by Surgeons 2 − 3 (87%/68%, p < 0.0001; and 81%/63%, p < 0.0001, respectively; Fig. 4A-B). Similarly, the 5-year DSS and OS rates for patients with clear margins (≥ 5 mm) were higher than those observed for those with close or positive margins (84%/73%/51%, p < 0.0001; and 79%/67%/46%, p < 0.0001, respectively; Fig. 4C-D).

Outcome analysis following stratification according to the operating surgeon and margin status

In this study, we were interested in margin status among surgeons. Patients were subsequently stratified in six groups according to the operating surgeon and margin status, as follows: 1) Surgeon 1/margins ≥ 5 mm, 2) Surgeon 1/margins < 5 mm, 3) Surgeon 1/positive margins, 4) Surgeons 2 − 3/margins ≥ 5 mm, 5) Surgeons 2 − 3/margins < 5 mm, and 6) Surgeons 2 − 3/positive margins. Significant intergroup differences were observed in terms of 5-year DSS and OS both before (Fig. 3E-F) and after (Fig. 4E-F) PS matching (all p < 0.0001). In the PS-matched cohort, the 5-year DSS and OS rates observed in patients operated by Surgeon 1 were higher than those of patients operated by Surgeons 2 − 3 regardless of margin status; however, the difference did not reach the significance threshold for the subgroups with positive margins (Table 3), which should be due to the small number of patients in this subgroup.

Table 3

Kaplan-Meier and Cox regression analyses of survival in patients operated by surgeon 1 versus surgeons 2–3 after stratification for margin status and propensity score matching
Method used for survival analysis	Margins ≥ 5 mm		Margins < 5 mm		Positive margins
Method used for survival analysis	Survival %/ HR (95% CI)	p	Survival %/ HR (95% CI)	p	Survival %/ HR (95% CI)	p
Kaplan-Meier
5-year DSS	87%/73%	0.0008	87%/69%	0.0213	75%/47%	0.3717
5-year OS	82%/69%	0.0217	80%/64%	0.0429	75%/41%	0.2388
Cox Regression
DSS	2.53 (1.44–4.44)	0.0012	2.32 (1.11–4.84)	0.0253	1.93 (0.44–8.42)	0.3803
OS	1.79 (1.08–2.97)	0.0237	1.86 (1.01–3.43)	0.0463	2.35 (0.54–10.12)	0.2531
Abbreviations: HR, hazard ratio; CI, confidence interval; DSS, disease-specific survival; OS, overall survival.

Univariable and multivariable Cox regression analysis according to the operating surgeon and margin status

When operating surgeon and margin status were combined as a risk factor, being operated by Surgeons 2 − 3 was a significant RF in univariable analysis (regardless of margin condition, versus Surgeon 1/margin ≥ 5mm) and was retained as an independent RF in multivariable analysis (Supplementary Table 1).

Additional stratification of margin status

In an attempt to shed further light on the survival differences observed between Surgeon 1 and Surgeons 2 − 3 according to margin status, patients with margins < 5 mm underwent an additional stratification using 2 mm as the cut-off (i.e., margins ≤ 2 mm versus > 2 mm). On analyzing patients with margins < 5 mm, we found that Surgeons 2 − 3 had a higher rate of patients with margins ≤ 2 mm compared with Surgeon 1, both before and after PS matching (62% versus 44%, p = 0.0005 and 66% versus 40%, respectively, p = 0.0003) (Supplementary Table 2). Patients with margins ≥ 5 mm underwent an additional stratification using 7 mm as the cut-off (i.e., margins ≤ 7 mm versus > 7 mm). On analyzing patients with margins ≥ 5 mm, we found that Surgeons 1 had a higher rate of patients with margins > 7 mm compared with Surgeon 2 − 3, both before and after PS matching (49% versus 25%, p < 0.0001 and 48% versus 27%, respectively, p = 0.0013) (Supplementary Table 3)

Previous studies have shown that high-volume surgeons can achieve wider surgical margins and higher cervical nodal yields; however, these findings were obtained by taking into account only a single operator (Table 4) [9, 10]. On analyzing data from three Taiwanese high-volume surgeons (high-volume was defined as ≥ 22 cases treated per year in Taiwan [5] and ≥ 15, 20, or 24 cases treated per year in Australia and New Zealand [4, 9, 10]) (Table 4), we demonstrate that significant survival disparities still exist for patients with OCSCC. Collectively, these findings point to other RFs as significant contributors to survival outcomes. Although some of these variables are patient-related and not modifiable (e.g., pathological T status, pathological N status, pathological stage, tumor differentiation, DOI, and ENE), others are at least partly physician-dependent (e.g., margin status and treatment modality). When patients were stratified according to the operating surgeon and margin status, we found that Surgeon 1 consistently outperformed Surgeons 2 − 3 in terms of clinical outcomes, regardless of surgical margin status (i.e., ≥ 5 mm, < 5 mm, and positive margins). Of note, the achievement of negative margins (≥ 5 mm) is a surgeon-dependent variable which is not necessarily related to hospital and/or surgeon volumes.

Table 4

Published literature focusing on the relationship between surgeons’ volume and clinical outcomes in patients with oral cavity squamous cell carcinoma
Authors (years of recruitment)/ Country	Patients included/ AJCC staging manual (edition)	Number of patients (high volume surgeons/total number of surgeons)	Number of high volume surgeons /total number of surgeons	Number of patients per surgeon per year	Pathological variables	Margin Status /Use of free flaps /Neck nodal yield	Independent risk factors in multivariable analyses
Authors (years of recruitment)/ Country	Patients included/ AJCC staging manual (edition)		Number of high volume surgeons /total number of surgeons	Number of patients per surgeon per year	Pathological variables	Margin Status /Use of free flaps /Neck nodal yield	DFS	DSS	OS
Kang CJ (2011–2018), current study, Taiwan (Nationwide)	First primary OCSCC /eight edition	1010/1010	3/3	Surgeon 1: 70; Surgeons 2–3: 28–29	Margins, depth, ENE, differentiation, nodal yield	Positive: 5% < 5 mm: 43% ≥ 5 mm: 52% Free flaps: 56%	-	High positive/close margins in high volume surgeons, pT, pN, differentiation	High positive/close margins in high volume surgeons, pT, pN, differentiation
⁴Liu T (2008–2013) Australia	New primary OCSCC /seventh edition	284/534	2/20	High volume: ≥ 20	Margin, ENE, perineural invasion, lymphovascular invasion, nodal yield	Positive: 11% < 5 mm: 63% ≥ 5 mm: 26% Free flaps: 44%%	Low volume surgeon, pN1-2	Low volume surgeon, ENE, PNI	Low volume surgeon, age, PNI, pN1-2, pT2-4
⁵Lee CC (2005) Taiwan (Nationwide)	OCSCC/ No staging data	419/1256 (criterion to define high-volume: one third of total patients)	12/215	High volume: ≥ 22	None	No data	-	-	Low volume surgeon, age, Charlson comorbidity index
⁶Lin CC (1997–1999) Taiwan (Nationwide)	OCSCC/ No staging data	2223/6666 (criterion to define high-volume: one third of total patients)	9/427	High volume: ≥ 47	None	No data	-	-	Low volume surgeon, age, Charlson comorbidity index
⁹Ellis OG (2008–2012) Australia	New primary OCSCC	122/250	1/13	High volume: ≥ 24	Margins	High volume: larger surgical margins	-	-	-
¹⁰Morton RT (1989–1998) New Zealand	-	148/317	1/10	High volume: 15	Nodal yield	High volume: higher nodal yield	-	-	-
Abbreviations: AJCC, American Joint Committee on Cancer; DFS, disease-free survival; DSS, disease-specific survival; OS, overall survival; OCSCC, oral cavity squamous cell carcinoma; ENE, extranodal extension; PNI, perineural invasion.

The use of a large nationwide database (TCRD) to select the three surgeons included in the study was motivated by 1) markedly differences in margins < 5 mm status between high volume surgeon, 2) A comprehensive assessment of the prognostic impact of surgical margins on OCSCC is required, and 3) the necessity to identify a large number of patients with OCSCC treated by different surgeons. In the original cohort, patients treated by Surgeon 1 had a higher disease burden at baseline and more frequently received adjuvant therapy. There was also a significant difference in lymph node yield during neck dissection, a well-known quality control factor. Differences in disease burden, frequency of adjuvant therapy, and lymph node yield certainly contribute to differences in survival. While this potential confounder was abrogated by PS matching, the survival differences between patients operated by Surgeon 1 and Surgeons 2 − 3 were even more pronounced in the PS-matched cohort (Fig. 3A-B and Fig. 4A-B). In this study, the most notable margin difference between Surgeon 1 and Surgeons 2 − 3 was related to close margins < 5 mm (SMD: -114.17) followed by positive margins (SMD: -31.05).

The prevalence of positive and close margins in previous large studies focusing on OCSCC has been reported to range from 30–60% [16–25]. These high rates may be generally explained by the adoption of a less radical surgical approach motivated by functional and cosmetic concerns. In the current study, Surgeons 2 − 3 were clearly less aggressive compared with Surgeon 1 (positive margin rates in the PS-matched cohort: 9.8% and 2.5%, respectively), and this may offer an explanation for the less favorable survival figures observed for patients treated by the former. Interestingly, on analyzing margin status in the subgroup of patients with margins < 5 mm using a cut-off of 2 mm, we found that, after PS matching, Surgeons 2 − 3 had a higher rate of margins ≤ 2 mm compared with Surgeon 1 (66% versus 40%, respectively). The adverse prognostic significance of margins ≤ 2 mm is well described in the published literature [8, 24, 26–28]. Another interesting observation is that Surgeon 1 had a higher rate of patients with margins > 7 mm compared with Surgeons 2 − 3 in the subgroup of patients with margins ≥ 5 mm (49% versus 25%, respectively). This finding is consistent with the prognostic significance of a 7-mm cut-off with respect to 5-year local control, DSS, and OS rate [29, 30].

Herein, we took advantage of a large nationwide dataset to screen a high number of surgeons and identify those characterized by high volumes and marked interindividual variability in surgical margin status. The difference in the proportion of margins ≥ 5 mm in patients treated by Surgeon 1 versus Surgeons 2 − 3 (75.4% versus 22.5%, respectively) may explain the clear survival differences observed in our study. One valuable strategy to improve survival of patients with OCSCC is the achievement of a surgical margin of at least 1 cm, which would in turn ensure a pathological margin ≥ 5 mm [31]. The use of free flap reconstruction is also recommended over local flaps, with the former approach being associated with a higher frequency of clear margins and a significant survival advantage [32]. As expected, the use of free-flap reconstructions in Surgeon 1 was higher than that observed in Surgeons 2 − 3 (85.5% versus 18.9%, respectively, p < 0.0001; detailed data not shown). It is clinically important to consider that a bigger resection might have an impact on postoperative functional outcome, especially in terms of swallowing and speech. However, in several studies [33–37], it has been mentioned that reconstruction using free flaps (as opposed to local flaps) has been shown to have better functional outcomes and thus a better quality of life.

Whether being operated by a high-volume surgeon confers a survival advantage in patients with OCSCC remains unclear. The rarity of this malignancy in Western countries has also hindered proper addressing of this research question in the published literature [4, 9, 10]. On examining a nationwide cohort, we were able to conduct a detailed analysis of RFs in relation to surgeons’ volumes. This is, to our knowledge, the second attempt to analyze this relationship in detail, after the seminal findings by Liu et al [4] (Table 4). However, only two high-volume surgeons were included in the previous paper [4]. In their study, 44% of patients underwent free-flap reconstruction and the frequency of clear margins was 26% (low- versus high-volume surgeons: 23% versus 29%, respectively). Therefore, the association between surgeons’ volumes and clear margins was not thoroughly dissected. In general, a higher surgical experience is associated with the ability to achieve more radical excisions characterized by an elevated number of dissected lymph nodes and high rates of negative margins. However, nodal yield was not identified in our study as an independent predictor of clinical outcomes. One potential explanation may lie in the confounding effect of adjuvant therapy – which lowers the risk of regional recurrences related to a limited number of dissected nodes. Further research is also necessary to identify the optimal cut-off in the definition of nodal yield.

There are limitations to this study. The retrospective nature of our research is not conducive to determining a causal relationship between a surgeon’s volume and clinical outcomes in patients with OCSCC; hence, this relationship needs to be thoroughly investigated in future longitudinal investigations – with the margin status included as a covariate in multivariable Cox regression analyses. As the TCRD did not include specific information on local control, we were unable to analyze this endpoint which is directly related to margin status. Additionally, certain pathological RFs (e.g., perineural invasion, lymphatic invasion, and vascular invasion) were included in our nationwide database as of 2018 only; for that reason, their prognostic impact deserves further scrutiny. This study was conducted in a betel quid chewing endemic area, and this factor may have limited the external validity of the results. Finally, we acknowledge that registry studies are prone to information bias.

The main conclusion to be drawn from the study was that surgeon 1 is delivering better care than surgeons 2 − 3. Notably, our data did suggest that the clinical outcomes of patients treated by Surgeon 1 were more favorable than those achieved by cases operated by Surgeons 2 − 3, and this was chiefly driven by the higher proportion of margins ≥ 5 mm obtained by the former. These results persisted after adjustment for baseline differences using propensity score matching. We conclude that high-volume surgeons should strive to achieve adequate margins in an effort to improve survival outcomes in patients with OCSCC.

OCSCC oral cavity squamous cell carcinoma

RFs risk factors

THPA Taiwan Health Promotion Administration

TCRD Taiwanese Cancer Registry Database

DOI depth of invasion

ENE extra-nodal extension

TNHIRD Taiwanese National Health Insurance Research Dataset

REMARK reporting recommendations for tumor marker

MDTC multidisciplinary team care

NCCN National Comprehensive Cancer Network

ICD International Classification of Diseases

DSS disease-specific survival

OS overall survival

STORE Standards for Oncology Registry Entry

PS propensity score

SMD standardized mean difference

HRs hazard ratios

CIs confidence intervals

Acknowledgements

The authors express their gratitude to the Research Service Center for Health Information at Chang Gung University, Taiwan, for their invaluable assistance in designing the study, managing data, and conducting statistical analysis.

Authors’ contributions

Conception and design: Chung-Jan Kang, Yu-Wen Wen, Chih-Yen Chien, Shu-Ru Lee, and Chun-Ta Liao

Analysis and interpretation of data: All authors

Drafting the article or revising it critically for important intellectual content:All authors

Final approval of manuscript:All authors

Agreement to be accountable for all aspects of the work: All authors

Funding

This research received financial support through grants (CMRPD1H0521 and BMRPC55) provided by the Chang Gung Medical Research Program.

Data availability

Data availability in this study is subject to third-party restrictions. As per the Health and Welfare Data Center of the Taiwanese Ministry of Health and Welfare (http://dep.mohw.gov.tw/DOS/), the “Personal Information Protection Act” imposes legal limitations that prevent data sharing. However, the authors were granted a data usage license to conduct this research. The datasets generated during and/or analyzed during this study are available from the corresponding author upon reasonable request, provided that formal permission is secured from the Taiwanese Ministry of Health and Welfare. Data are however available from the authors upon reasonable request and with permission of HWDC, MOHW. Contact the co-first author Yu-Wen Wen (Telephone: +886-3211-8800 [ext. 3759], E-mail: [email protected]) if someone wants to request the data from this study.

Conflict of interest

The authors declare no conflicts of interest.

Ethics approval and consent to participate

The study was conducted according to the guidelines of the Declaration of Helsinki. All procedures were approved by the Chang Gung Medical Foundation Institutional Review Board (reference number: 201801398B0A3). The requirement for written patient informed consent was waived due to the study design.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interest.

Author details

¹Department of Otorhinolaryngology, Head and Neck Surgery, Chang Gung Memorial Hospital and Chang Gung University, Taoyuan, Taiwan, ROC

²Clinical Informatics and Medical Statistics Research Center, Chang Gung University, Taoyuan, Taiwan, ROC

³Division of Thoracic Surgery, Chang Gung Memorial Hospital, Taoyuan, Taiwan, ROC

⁴Department of Otolaryngology, Chang Gung Memorial Hospital Kaohsiung Medical Center, Chang Gung University College of Medicine, Taiwan, ROC

⁵Research Service Center for Health Information, Chang Gung University, Taoyuan Taiwan, ROC

⁶Department of Diagnostic Radiology, Chang Gung Memorial Hospital and Chang Gung University, Taoyuan, Taiwan, ROC

⁷Department of Pathology, Chang Gung Memorial Hospital and Chang Gung University, Taoyuan, Taiwan, ROC

⁸Department of Radiation Oncology, Changhua Christian Hospital, Changhua, Taiwan, ROC

⁹Department of Otolaryngology, National Taiwan University Hospital and College of Medicine, Taipei, Taiwan, ROC

¹⁰Department of Head and Neck Surgery, Koo Foundation Sun Yat-Sen Cancer Center, Taipei, Taiwan, ROC

¹¹Department of Otorhinolaryngology, China Medical University Hospital, Taichung, Taiwan, ROC

¹²Department of Otolaryngology, Shuang Ho Hospital, Taipei Medical University, New Taipei City, Taiwan, ROC

¹³Department of Radiation Oncology, Chang Gung Memorial Hospital, Chiayi, Taiwan, ROC

¹⁴Department of Otorhinolaryngology-Head and Neck Surgery, Chang Gung Memorial Hospital, Chiayi, Taiwan, ROC

¹⁵Department of Oral and Maxillofacial Surgery, Chang Gung Memorial Hospital, Chang Gung University, Taoyuan, Taiwan, ROC

¹⁶Department of Medical Oncology, Chang Gung Memorial Hospital and Chang Gung University, Taoyuan, Taiwan, ROC

¹⁷Department of Plastic and Reconstructive Surgery, Chang Gung Memorial Hospital and Chang Gung University, Taoyuan, Taiwan, ROC

¹⁸Department of Nuclear Medicine and Molecular Imaging Center, Chang Gung Memorial Hospital and Chang Gung University, Taoyuan, Taiwan, ROC

Liu JH, Zingmond DS, McGory ML, et al. Disparities in the utilization of high-volume hospitals for complex surgery. JAMA. 2006;296:1973–80. doi: 10.1001/jama.296.16.1973.
Chang CM, Huang KY, Hsu TW, et al. Multivariate analyses to assess the effects of surgeon and hospital volume on cancer survival rates: a nationwide population-based study in Taiwan. PLoS ONE. 2012;7:e40590. doi: 10.1371/journal.pone.0040590.
Eskander A, Merdad M, Irish JC, et al. Volume-outcome associations in head and neck cancer treatment: a systematic review and meta-analysis. Head Neck. 2014;36:1820–1834. doi: 10.1002/hed.23498.
Liu T, David M, Ellis O, et al. Treatment for oral squamous cell carcinoma: Impact of surgeon volume on survival. Oral Oncol. 2019;96:60–5. doi:10.1016/j.oraloncology.2019.06.030.
Lee CC, Ho HC, Chou P. Multivariate analyses to assess the effect of surgeon volume on survival rate in oral cancer: a nationwide population-based study in Taiwan. Oral Oncol. 2010;46:271–5. doi: 10.1016/j.oraloncology.2010.01.006.
Lin CC, Lin HC. Effects of surgeon and hospital volume on 5-year survival rates following oral cancer resections: the experience of an Asian country. Surgery. 2008;143:343–51. doi: 10.1016/j.surg.2007.09.033.
National Comprehensive Cancer Network Clinical Practice Guidelines in Oncology, Head and Neck Cancers, Version I 2024 Available at: http://www.nccn.org/. (Accessed on November 21, 2023.)
Lin CY, Fan KH, Lee LY, et al. Precision adjuvant therapy based on detailed pathologic risk factors for resected oral cavity squamous cell carcinoma: long-term outcome comparison of CGMH and NCCN guidelines. Int J Radiat Oncol Biol Phys 2020;106:916–25. doi: 10.1016/j.ijrobp.2019.08.058.
Ellis OG, David MC, Park DJ, Batstone MD. High-volume surgeons deliver larger surgical margins in oral cavity cancer. J Oral Maxillofac Surg. 2016;74:1466–72. doi: 10.1016/j.joms.2016.01.026.
Morton RP, Gray L, Tandon DA, Izzard M, McIvor NP. Efficacy of neck dissection: are surgical volumes important? Laryngoscope. 2009;119:1147–52. doi: 10.1002/lary.20167.
Liao CT, Kang CJ, Lee LY, et al. Association between multidisciplinary team care approach and survival rates in patients with oral cavity squamous cell carcinoma. Head Neck 2016;38:Suppl 1, E1544–53. doi: 10.1002/hed.24276.
McShane LM, Altman DG, Sauerbrei W, Taube SE, Gion M, Clark GM. REporting recommendations for tumour MARKer prognostic studies (REMARK). Br J Cancer. 2005;93:387–91. doi: 10.1038/sj.bjc.6602678.
Sauerbrei W, Taube SE, McShane LM, Cavenagh MM, Altman DG. Reporting Recommendations for Tumor Marker Prognostic Studies (REMARK): An Abridged Explanation and Elaboration. J Natl Cancer Inst. 2018;110:803–11. doi: 10.1093/jnci/djy088.
Amin MB, Edge S, Greene F, Byrd DR, Brookland RK, Washington MK, editors. AJCC cancer staging manual, 8th ed. New York: Springer-Verlag; 2017.
American College of Surgeons. Standards for Oncology Registry Entry. STORE 2023. Effective for Cases Diagnosed. January 1, 2023.
Loree TR, Strong EW. Significance of positive margins in oral cavity squamous carcinoma. Am J Surg. 1990;160:410–4. doi: 10.1016/s0002-9610(05)80555-0.
Parsons JT, Mendenhall WM, Stringer SP, Cassisi NJ, Million RR. An analysis of factors influencing the outcome of postoperative irradiation for squamous cell carcinoma of the oral cavity. Int J Radiat Oncol Biol Phys. 1997;39:137–48. doi: 10.1016/s0360-3016(97)00152-1.
Sutton DN, Brown JS, Rogers SN, Vaughan ED, Woolgar JA. The prognostic implications of the surgical margin in oral squamous cell carcinoma. Int J Oral Maxillofac Surg. 2003;32:30–4. doi: 10.1054/ijom.2002.0313.
McMahon J, O'Brien CJ, Pathak I, et al. Influence of condition of surgical margins on local recurrence and disease-specific survival in oral and oropharyngeal cancer. Br J Oral Maxillofac Surg. 2003;41:224–31. doi: 10.1016/s0266-4356(03)00119-0.
Brandwein-Gensler M, Teixeira MS, Lewis CM, et al. Oral squamous cell carcinoma: histologic risk assessment, but not margin status, is strongly predictive of local disease-free and overall survival. Am J Surg Pathol. 2005;29:167–78. doi: 10.1097/01.pas.0000149687.90710.21.
Woolgar JA, Rogers S, West CR, Errington RD, Brown JS, Vaughan ED. Survival and patterns of recurrence in 200 oral cancer patients treated by radical surgery and neck dissection. Oral Oncol. 1999;35:257–65. doi: 10.1016/s1368-8375(98)00113-4.
Woolgar JA, Triantafyllou A. A histopathological appraisal of surgical margins in oral and oropharyngeal cancer resection specimens. Oral Oncol. 2005;41:1034–43. doi: 10.1016/j.oraloncology.2005.06.008.
Cheng A, Cox D, Schmidt BL. Oral squamous cell carcinoma margin discrepancy after resection and pathologic processing. J Oral Maxillofac Surg. 2008;66:523–9. doi: 10.1016/j.joms.2007.08.040.
Nason RW, Binahmed A, Pathak KA, Abdoh AA, Sándor GK. What is the adequate margin of surgical resection in oral cancer? Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2009;107:625–9. doi: 10.1016/j.tripleo.2008.11.013.
Ebrahimi A, Murali R, Gao K, Elliott MS, Clark JR. The prognostic and staging implications of bone invasion in oral squamous cell carcinoma. Cancer. 2011;117:4460–7. doi: 10.1002/cncr.26032.
Lin MC, Leu YS, Chiang CJ, et al. Adequate surgical margins for oral cancer: A Taiwan cancer registry national database analysis. Oral Oncol. 2021;119:105358. doi: 10.1016/j.oraloncology.2021.105358.
Binahmed, A.; Nason, R.W.; Abdoh, A.A. The clinical significanc of the positive surgical margin in oral cancer. Oral Oncol. 2007;43:780–4. doi: 10.1016/j.oraloncology.2006.10.001.
Mitchell DA, Kanatas A, Murphy C, Chengot P, Smith AB, Ong TK. Margins and survival in oral cancer. Br J Oral Maxillofac Surg. 2018;56:820–9. doi: 10.1016/j.bjoms.2018.06.021.
Liao CT, Chang JT, Wang HM, et al. Analysis of risk factors predictive of local tumor control in oral cavity cancer. Ann Surg Oncol 2008;71:371–6. doi: 10.1245/s10434-007-9761-5.
Mishra A, Malik A, Datta S, Mair M, Bal M, Nair D, Nair S, Chaturvedi P. Defining optimum surgical margins in buccoalveolar squamous cell carcinoma. Eur J Surg Oncol. 2019 Jun;45(6):1033-8. doi: 10.1016/j.ejso.2019.01.224.
Sarode SC, Sarode GS, Karmarkar S. Shrinkage based assessment of surgical margins: a need of the hour in oral cancer. Int J Oral Maxillofac Surg. 2011;40:1336–7. doi: 10.1016/j.ijom.2011.05.005.
Liao CT, Wen YW, Lee SR, et al. Clinical outcomes of Taiwanese patients with resected oral cavity squamous cell carcinoma who underwent reconstruction with free versus local flaps. Ann Surg Oncol 2022;29:1130–40. doi: 10.1245/s10434-021-10524-x.
Hsing CY, Wong YK, Wang CP, et al: Comparison between free flap and pectoralis major pedicled flap for reconstruction in oral cavity cancer patients—A quality of life analysis. Oral Oncol. 2011;47:522–7. doi: 10.1016/j.oraloncology.2011.03.024.
Hanasono MM, Friel MT, Klem C, et al. Impact of reconstructive microsurgery in patients with advanced oral cavity cancers. Head Neck. 2009;31:1289–96. doi: 10.1002/hed.21100.
Fang QG, Shi S, Li M, Zhang X, Liu FY, Sun CF. Free flap reconstruction versus non-free flap reconstruction in treating elderly patients with advanced oral cancer. J Oral Maxillofac Surg. 2014;72:1420–4. doi: 10.1016/j.joms.2014.01.010.
Tsue TT, Desyatnikova SS, Deleyiannis FW, et al. Comparison of cost and function in reconstruction of the posterior oral cavity and oropharynx: Free vs pedicled soft tissue transfer. Arch Otolaryngol Head Neck Surg. 1997;123:731–7. doi: 10.1001/archotol.1997.01900070075012.
Chepeha DB, Annich G, Pynnonen MA, et al: Pectoralis major myocutaneous flap vs revascularized free tissue transfer. Arch Otolaryngol Head Neck Surg. 2004;130:181–6. doi: 10.1001/archotol.130.2.181.

No competing interests reported.

SupplementaryFigure1.tif
Supplementary figure 1. Kaplan-Meier plots of 5-year disease-specific survival (panel A) and overall survival (panel B) for patients with oral cavity squamous cell carcinoma operated by Surgeon 2 and Surgeon 3.
SupplementaryTable1.docx
SupplementaryTable2.doc
SupplementaryTable3.doc

Are all high-volume surgeons equally good outcome? A survival analysis of patients with oral cavity squamous cell carcinoma in relation to surgical margin status based on propensity score matching analysis

Status:

Version 1

Abstract

Background

Methods

Results

Conclusions

Figures

Introduction

Methods

Data sources

Treatment protocol and follow-up protocol

Patient selection

Surgeon-based analysis of surgical margin status

Data collection

Statistical analysis

Results

Patient characteristics

Univariable and multivariable Cox regression analysis

Outcome analysis

Outcome analysis following stratification according to the operating surgeon and margin status

Univariable and multivariable Cox regression analysis according to the operating surgeon and margin status

Additional stratification of margin status

Discussion

Conclusions

Abbreviations

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1