Customize your JAMA Network experience by selecting one or more topics from the list below.
Kato MG, Erkul E, Nguyen SA, et al. Extracapsular Dissection vs Superficial Parotidectomy of Benign Parotid Lesions: Surgical Outcomes and Cost-effectiveness Analysis. JAMA Otolaryngol Head Neck Surg. 2017;143(11):1092–1097. doi:10.1001/jamaoto.2017.1618
How do health services and clinical outcomes of extracapsular dissection (ECD) compare with superficial parotidectomy when treating benign parotid neoplasms?
In this medical record review comparing 26 cases of ECD and 20 cases of superficial parotidectomy, operating room, anesthesia, and total hospital charges were significantly lower in the ECD group, and procedure time, anesthesia time, and length of stay for ECD cases were significantly shorter. Postoperative complication rates showed no difference between groups.
Extracapsular dissection is a more favorable approach when considering health services outcomes in the treatment of select benign parotid lesions.
The clinical implications of extracapsular dissection over superficial parotidectomy are controversial and limited in data on cost-effectiveness.
To compare extracapsular dissection with superficial parotidectomy for benign parotid tumors with respect to surgical outcomes and cost-effectiveness.
Design, Setting, and Participants
This was a retrospective medical record review and cost-effectiveness analysis performed from August 2012 to November 2015 at a tertiary care institution. Adult patients (age ≥18 years) who underwent parotidectomy for benign parotid lesions were included.
Treatment by extracapsular dissection or superficial parotidectomy.
Main Outcomes and Measures
Differences in postoperative complication rates and health services outcomes, including procedure time, anesthesia time, length of stay, and charges for surgeon, anesthesia, operating room, and hospital.
A total of 46 parotidectomies consisting of 26 extracapsular dissections and 20 superficial parotidectomies met criteria. Of the 46 patients, 33 were women. Patient ages ranged from 18 to 83 years. Lesion features were similar between groups with most being pleomorphic adenoma. Procedure time (effect size, −1.31; 95% CI, −1.93 to −0.65), anesthesia time (effect size, −1.37; 95% CI, −1.99 to −0.70), and length of stay (effect size, −0.66; 95% CI, −1.25 to −0.05) were significantly shorter for extracapsular dissection compared with superficial parotidectomy. Moreover, anesthesia (effect size, −1.55; 95% CI, −2.19 to −0.86), operating room (effect size, −1.09; 95% CI, −1.69 to −0.45), and total hospital charges (effect size, −1.13; 95% CI, −1.74 to −0.49) were significantly less for extracapsular dissection while remaining charges, including surgeon, showed no difference. Finally, facial nerve weakness, great auricular nerve dysesthesia, and other relevant postoperative complications were comparable between groups.
Conclusions and Relevance
In the hands of an experienced surgeon, extracapsular dissection is a shorter, less costly, and equally safe alternative to traditional superficial parotidectomy when treating benign parotid lesions. Further follow-up is needed to ensure these perceived advantages are maintained over time.
Neoplasms of the parotid are an uncommon entity in the realm of head and neck cancers. Of all primary head and neck neoplasms, only 1% to 3% originate in the parotid, 70% to 90% of which are of benign histopathologic characteristics.1-6 Among these benign lesions, most are pleomorphic adenoma (PA), followed by Warthin tumor (WT), and less frequently, oncocytoma, and basal cell adenoma, among others.4
The preferred surgical treatment of parotid tumors has evolved over the past century. In the early 20th century, intracapsular enucleation primarily sought to avoid damage to the facial nerve (FN) while performing a subtotal removal of the tumor capsule.7 Owing to recurrence rates as high as 45%, a push for more radical techniques took place.8-11 By the mid-20th century, superficial parotidectomy (SP) became widely accepted as the gold standard owing to a reduction in tumor recurrence rates as low as 2%.8 By increasing the volume of gland resected, patients were at a higher risk of developing FN palsy, Frey syndrome (FS), and loss of facial contour. In the later 20th century, partial SP, which involves dissection of only the nerve branches in closest proximity to the tumor, became the favored approach to allow for complete tumor resection with fewer complications. In the past 25 years, experienced salivary surgeons have taken this approach 1 step further by advocating extracapsular dissection (ECD), a technique of removing the tumor and its capsule along with a thin rim of normal glandular tissue without formal identification and dissection of the FN. A recent meta-analysis by Albergotti et al12 and other studies13,14 consistently report similar rates of recurrence between ECD and SP, but with reduced rates of FN paresis and FS with ECD. An updated meta-analysis by Xie et al15 advocated ECD as a safer alternative for selected smaller, superficial, mobile benign lesions without FN involvement. Nonetheless, the debate over which technique should be favored continues.
Although the clinical outcomes of ECD and SP have been compared, less is known about the health services outcomes of the 2 techniques. Therefore, the present study sought to determine differences with regard to the associated costs, procedure time (PT), anesthesia time (AT), and length of stay (LOS) in conjunction with clinical outcomes between each procedure to further delineate the role of ECD in the management of benign parotid tumors.
The present study was approved by the institutional review board of the Medical University of South Carolina (MUSC) in Charleston, and the requirement for informed consent was waived. The study retrospectively reviewed all parotidectomies for benign parotid neoplasms conducted by head and neck surgeons at MUSC between August 2012 and November 2015. The start date of the search was based on the start date of a new electronic medical record (EPIC) at MUSC.
Parotidectomies performed on adult patients (≥18 years) were identified through Current Procedural Terminology codes, 42415 and 42410, encoding for excision procedures on the salivary gland and ducts. Operative notes were then evaluated to classify procedures as either ECD or SP, for which descriptions are provided herein.
Lesions included in this study were limited to primary parotid tumors with pathologic classifications recognized by the World Health Organization as benign epithelial neoplasms of the parotid, namely PA, WT, oncocytoma, basal cell adenoma, oncocytic papillary cystadenoma, among others.16 All tumors had to be benign by fine-needle aspiration (FNA) and located in the superficial lobe of the parotid gland. During the period of study, one staff surgeon (M.B.G.) performed only ECD for these tumors, whereas the remaining faculty surgeons (T.A.D., J.D.H., E.J.L.) performed a standard superficial parotidectomy technique. Only lesions definitively confirmed by surgical pathology reports were included, without restrictions to size, number of foci, and preoperative or intraoperative mobility characteristic. Lesions with multiple foci were designated a size measurement of the largest focus. Patients presenting with recurrent lesions or history of surgery on the affected parotid gland were excluded.
Procedures met inclusion criteria of SP if the surgical technique described formal identification and dissection of the main trunk of the FN, followed by dissection of 1 or more FN branches, with removal of the tumor with a cuff of normal parotid parenchyma. Procedures were classified as ECD if there was no formal identification of the main FN trunk and the tumor was removed along its capsule with limited dissection of exposed nerve branches. All procedures were conducted with the patient under general anesthesia with variable use of FN monitoring technology, dependent on the operating surgeon.
The following patient data were extracted through the MUSC electronic patient record: age; sex; follow-up duration; operating surgeon; nerves identified intraoperatively; final margin status; use of FN monitoring technology; prognostic intraoperative events (eg, sacrificing of nerves); and lesion features, including laterality, pathology, size, and foci count. Extracted surgical outcome data included procedure time (PT), anesthesia time (AT), LOS, and postoperative complications. Any documented sign of FN dysfunction in the immediate postoperative period was considered FN weakness. Facial nerve weakness and great auricular nerve dysesthesia were further characterized using a threshold of more than 6 months of symptom presence to distinguish persistent conditions from transient. Considered a late postoperative complication, FS rates included only patients with a minimum follow-up threshold of 1 month. Owing to inadequate follow-up, the following patients were excluded from calculations related to complication rates: 5 patients from each group to evaluate for FN weakness, 6 and 2 patients to evaluate great auricular nerve dysesthesia for ECD and SP, respectively, and 13 and 5 to evaluate for FS, respectively.
Charge data were used as a proxy for cost owing to differences in actual payment amounts with varying insurance contracts. Charge data (US$) for both groups were obtained through the MUSC billing database. Data consisted of surgeon charge, anesthesia (including certified nurse anesthetist, if used) charge, combined other professional charges, operating room (OR) charge, and total hospital charge. Anesthesia, surgeon, and combined other professional charges considered only professional fees. The OR charge was based on occupancy in minutes. Total hospital charge included medical and surgical supplies and devices, OR charge, pharmaceutical supplies, recovery room occupancy, pathology laboratory expenses, and anesthesia time charge. Cases with supplemental or concomitant procedures in addition to parotidectomy were excluded from analyses comparing costs, PT, AT, and LOS.
All data analyses were performed with SigmaPlot software (version 12.5; SPSS Inc) and MedCalc software (version 16.8; MedCalc Software bvba). Disease information and demographic variables were summarized by means of summary statistics. All continuous variables were tested for normal distribution as determined by the Kolmogorov-Smirnov test. Categorical variables were summarized by frequency, percentage, and/or range. Continuous variables were summarized by mean (SD) or interquartile range where appropriate. Comparisons of categorical variables were performed using a Fisher exact test or χ2 test. For continuous variables, comparisons between groups (ECD vs SP) were made with an independent t test (normal distribution) or a Mann-Whitney (nonnormal distribution). Finally, effect size (ES) estimates at 0.2 with 95% CIs between ECD vs SP, regarding PT, AT, LOS, surgeon charges, anesthesia charges, other professional charges, OR charges, and total hospital charges were calculated.
There were a total of 46 parotidectomies performed for benign tumors (in 45 total patients) by head and neck surgeons at MUSC between August 2012 and November 2015 that met inclusion criteria. Of these, 26 (57%) were ECD and 20 (43%) SP. Within the SP group, 15 of 20 cases dissected all 5 branches of the FN (true SP), while the remaining dissected fewer (partial SP). Cases were performed by a total of 6 operating surgeons. Facial nerve monitoring was used in 39 (85%) cases, equally frequently between the ECD (21 of 26 [81%]) and SP (18 of 20 [90%]) groups (ES, 0.02; 95% CI, −0.22 to 0.34). None of the procedures encompassed any aspect of reconstruction at the surgical site.
Patient demographics and lesion features did not significantly differ between ECD and SP groups (Table 1). Notably, mean (SD) lesion size for ECD (2.2 [0.9] cm) vs SP (2.3 [1.1] cm) was nonsignificant (ES, −0.10; 95% CI, −0.68 to 0.48), and lesion types were similarly distributed within each group, the most common being PA, followed by WT. Less common lesions included 1 case of oncocytoma, 1 case of oncocytic papillary cystadenoma, and 2 cases of basal cell adenoma. Per the pathology report, final margin status showed nonsignificant differences between groups (ES, 0.02; 95% CI, 0.26 to −0.30). Median follow-up time for ECD was shorter than for SP (29 vs 132 days; ES, −0.30; 95% CI, −0.88 to 0.29).
The rate of postoperative complications was comparable between the ECD and SP groups (Table 2). No incidence of permanent FN weakness was observed for either group, while the incidence of transient FN weakness were found to be nonsignificant between groups (ES, 0.02; 95% CI, −0.15 to 0.42). Likewise, rates of transient (ES, 0.03; 95% CI, −0.25 to 0.36) and permanent great auricular nerve dysesthesia (ES, 0.03; 95% CI, −0.30 to 0.31), FS (ES, 0.04; 95% CI, −0.25 to 0.45), and combined sialoceles and seromas (ES, 0.02; 95% CI, −0.26 to 0.30) were found to be nonsignificant between each group. No other relevant postoperative complications were observed in this study population.
Statistically significant differences were observed in analyses comparing charges, PT, AT, and LOS between ECD and SP groups. ECD procedures were found to be substantially shorter in duration compared with SP procedures (ES, −1.31; 95% CI, −1.93 to −0.65) with decreased AT (ES, −1.37; 95% CI, −1.99 to −0.70) and LOS (ES, −0.66; 95% CI, −1.25 to −0.05) (Table 3). In terms of charges (US$), anesthesia (ES, −1.55; 95% CI, −2.19 to −0.86), OR (ES, −1.13; 95% CI, −1.74 to −0.49), and total hospital charges (ES, −1.13; 95% CI, −1.74 to −0.49) were considerably less for ECD while surgeon (ES, −0.31; 95% CI, −0.89 to 0.28) and other professional (ES, −0.07; 95% CI, −0.65 to 0.52) charges showed no difference (Table 4). Three patients were not included in this analysis: 1 patient with concurrent bilateral lesions, each lesion treated by separate techniques in the same operative encounter, and 2 patients, 1 in each group, who underwent additional operative treatments.
The clinical benefits of ECD over SP in the treatment of benign parotid neoplasms continue to be debated. In past decades, a number of studies have contributed data supporting ECD as an alternative to SP with principal attention to known critical postoperative complications, while other health services outcomes, such as cost-effectiveness, have not yet been examined.12-15,17 The present study is therefore novel by investigating both the clinical and cost-effectiveness outcomes of ECD compared with SP.
Comparing health services outcomes and cost-effectiveness is critical, particularly in an era of a growing population, and, thus, an increased burden of parotid tumors. Their assessments will provide insight into ways to optimally allocate clinician and hospital resources to effectively manage benign parotid tumors. In the present study, statistically significant differences related to charges, PT, AT, and LOS were found. To start, the mean (SD) ECD PT was profoundly reduced compared with that of SP (83.5 [36.8] minutes vs 139.0 [48.8] minutes). This was anticipated owing to the time required to initially identify the FN trunk and subsequently dissect its respective branches during SP. This difference translated to a decreased AT in ECD procedures (148.2 [48.1] minutes vs 213.3 [46.8] minutes). Moreover, the LOS (0.5 [0.8] days vs 1.3 [1.6] days) was shorter, owing to the greater use of closed suction drains and overnight admission in the SP group compared with ECD. The various charges dependent on the factor of time, whether in the form of AT, PT, or LOS, were accordingly reflected in the decrease in anesthesia, OR, and total hospital charges, and favored ECD with the remaining non–time-related charges being equal. In this limited sample, the main effect of ECD was not clinical differences, but an overall decrease in charges, which was achieved without compromising patient well-being. Perhaps in the hands of an experienced surgeon, the primary indication to conduct this procedure is its shorter duration and reduced financial burden.
From the clinical standpoint, the present study observed no significant differences in complication rates when comparing the 2 surgical approaches. Permanent FN weakness was not observed in either group, while transient FN weakness (14% vs 47%) and FS (0% vs 13%) were less commonly observed in ECD compared with SP, although this failed to reach significance owing to the limited sample size. This nonsignificant difference, however, is consistent with current literature associating ECD with similar or decreased rates of permanent weakness, and more frequently, diminished rates of transient weakness and FS.12-15 A study by McGurk et al13 observed indistinguishable differences in the rate of permanent weakness, but a substantial decrease in transient weakness (21% vs 71%; P < .001) and FS (20% vs 71%; P < .001) in ECD cases compared with SP cases. In a later study by Dell’Aversana Orabona et al,14 significant reductions in permanent (0% vs 8.9%; P < .001) and transient (3.9% vs 26.8%; P = .001) facial weakness, and FS (0% vs 5.3%; P < .001) were observed with the ECD technique. Ultimately, these studies were collectively presented in a meta-analysis by Xie et al15 illustrating overall improved rates in transient (relative risk [RR], 0.38; 95% CI, 0.16-0.88) and permanent FN weakness (RR, 0.38; 95% CI, 0.16-0.88), and FS (RR, 0.15; 95% CI, 0.07-0.30) for ECD compared with SP. Most recently, Mantsopoulos et al17 reported an increase in the rate of ECD performed at a single institution over a period of nearly 2 decades, attributing it to surgeon familiarity and perceived clinical advantages. When considering recurrence rates, settings of unexpected malignancy, and FN injury, the group acknowledged favorable outcomes and strongly advocated ECD for properly selected patients in the care of experienced surgeons.
The current evidence, however, should be considered with caution owing to the potential for substantial selection bias based on tumor size, location, or proximity to the FN when determining the best surgical approach.14,18,19 For example, Dell’Aversana Orabona et al,14 who reported a dramatic improvement in postoperative comorbidities for ECD, exercised a size criteria that dictated that tumors larger than 3 cm underwent SP while smaller lesions underwent ECD. In the study by McGurk et al,13 lesion sizes were simply described as smaller than 4 cm, and therefore, cannot be remarked on. Furthermore, varying definitions of ECD among authors may further obscure the data.19,20 The absence of statistically significant differences in our study may be primarily explained by the lesion sizes, which were nearly identical between the ECD and SP groups (mean [SD], 2.2 [0.9] cm vs 2.3 [1.1] cm; ES, −0.10; 95% CI, −0.68 to 0.48). Another point to consider is that the 2 aforementioned studies13,14 encompassed larger sample sizes of 630 and 232 parotidectomies for McGurk et al13 and Dell’Aversana Orabona et al,14 respectively, compared with 46 in this study.
Interestingly, great auricular nerve dysesthesia rate comparisons are sparse in the literature, for which our study shows slightly higher rates for transient conditions in the ECD group (15% vs 11%) and lower rates for permanent conditions (0% vs 6%), both revealed to be nonsignificant. A separate study by George and McGurk21 examined transient great auricular nerve dysesthesia rates in patients treated with ECD (5%), however, without comparison with SP.
Finally, owing to inadequate follow-up duration, recurrence was not an outcome for comparison in this study, as the literature indicates these events to occur beyond 5 years postoperatively.22-25 While the overall clinical picture of ECD is promising, the importance of appropriate lesion selection by experienced surgeons at higher higher-volume institutions is stressed. For less experienced surgeons, ECD may be a technique limited to smaller, mobile, superficially located tumors that have FNA evidence of a benign process.15,17 Inadvertant removal of a malignant tumor by ECD is always a risk, and the incidence could not be captured by this study, which was limited to benign histopathology; however, this occurrence should be rare if proceeded by FNA. Previous studies have shown that if this does occur, ultimate patient outcome is not compromised if the patient undergoes subsequent standard parotidectomy approaches.13,17,26
While comparing clinical outcomes did not favor one procedure over the other in the present study, their nonsignificance in differences add greater weight to the results related to health services outcomes and cost-effectiveness. In other words, ECD seems to be superior to SP in the context of these outcomes, although primarily in the earlier postoperative course.
There are a few limitations to consider when evaluating the results of this study. First, the follow-up time for each group was not substantial given the date range of analysis. Frequently, patients were referred from practices external to this institution, and many patients received ongoing follow-up at home. Consequently, obtaining complete data on postoperative complications, particularly recurrence, outside of this timeframe posed a challenge. As a result, the sample size of this study may have been insufficient to observe the differences in early and late postoperative complications reported in previous studies. These factors did not affect the analysis for cost-effectiveness, PT, AT, and LOS owing to the relatively large ES of the outcomes. However, this cost-effective analysis pertains only to the initial procedure of these lesions, without consideration of costs related to long-term postoperative complications, namely, recurrence. In addition, the study did not examine the role of lesion location and its effect on these outcomes. For instance, procedures on deeper lesions, whether approached by ECD or SP, are expected to last longer than for superficial tumors. Finally, it is imperative to note that these data were sourced from a single institution and may not apply to others. Therefore, there is a need for further studies at additional sites to further characterize the capital demands of each procedure.
In the hands of experienced parotid surgeons, ECD is a viable alternative to the traditional SP with regard to clinical outcomes, and may be superior with regard to cost. It is recognized that successful outcomes from ECD require proper selection of patients who are likely to benefit and not be harmed by the procedure. Studies with greater follow-up time are needed to ensure that these perceived advantages are maintained over time.
Corresponding Author: Masanari G. Kato, BS, Department of Otolaryngology–Head and Neck Surgery, Medical University of South Carolina, 135 Rutledge Ave, MSC 550, Charleston, SC 29425 (firstname.lastname@example.org).
Accepted for Publication: June 17, 2017.
Published Online: September 21, 2017. doi:10.1001/jamaoto.2017.1618
Author Contributions: Mr Kato had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Kato, Gillespie.
Acquisition, analysis, or interpretation of data: Kato, Erkul, Nguyen, Day, Lentsch, Gillespie.
Drafting of the manuscript: Kato, Erkul, Gillespie.
Critical revision of the manuscript for important intellectual content: Kato, Erkul, Nguyen, Day, Lentsch, Gillespie.
Statistical analysis: Kato, Nguyen.
Administrative, technical, or material support: Kato, Erkul, Day, Gillespie.
Study supervision: Day, Lentsch, Gillespie.
Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.