Balancing hemostasis and function: the impact of hemostatic materials on voice and swallowing outcomes in nerve-preserved thyroidectomy

Abstract

Objective

Post-thyroidectomy dysphonia may occur even when recurrent laryngeal nerve (RLN) and external branch of the superior laryngeal nerve (EBSLN) integrity are preserved. Hemostatic materials (HMs) are widely used to enhance intraoperative hemostasis and reduce postoperative drainage; however, their use may introduce a functional trade-off between bleeding control and postoperative voice and swallowing outcomes. This study aimed to evaluate how different HMs balance hemostatic efficacy and functional preservation in nerve-preserved total thyroidectomy.

Methods

This retrospective study included adult patients who underwent primary nerve-preserved total thyroidectomy with standardized intraoperative neuromonitoring (IONM). Patients were categorized into three groups according to HM use: no hemostatic material (NHM), fibrin sealant (Tisseel, TIS), and polysaccharide-based agent (4DryField PH, 4DF). Objective voice parameters and subjective voice and swallowing assessments were obtained preoperatively and at six weeks postoperatively.

Results

175 patients were analyzed (NHM:74/TIS:30/4DF:71), with comparable demographic, surgical, and pathological characteristics among groups. TIS provided the most robust hemostatic effect, with 0% drain placement, but was associated with significantly greater deterioration in pitch-related voice parameters and patient-reported voice outcomes. In contrast, 4DF showed postoperative voice outcomes comparable to NHM and demonstrated a significantly lower incidence of swallowing impairment (p = 0.007).

Conclusion

In nerve-preserved total thyroidectomy, hemostatic efficacy and functional preservation exist along a balance rather than a binary choice. Tisseel provides superior bleeding control through robust tissue sealing, however, their strong mechanical restriction on key extralaryngeal muscles may compromise postoperative voice function. 4DF achieves adequate hemostasis while preserving voice and swallowing function. These findings support a personalized, anatomy-aware selection of hemostatic materials, emphasizing functional priorities and precise avoidance of critical voice-related structures, rather than routine or uniform application.

Introduction

Total thyroidectomy is a common surgical procedure for benign and malignant thyroid diseases [1]. Although modern surgical techniques and perioperative care have markedly improved safety, several complications, including postoperative hematoma, hypoparathyroidism, infection, and injuries to the recurrent laryngeal nerve (RLN) or the external branch of the superior laryngeal nerve (EBSLN), may still occur [2]. RLN and EBSLN injuries are the major neural factors contributing to postoperative voice changes, ranging from hoarseness and impaired pitch elevation to airway obstruction in severe bilateral injuries [3]. Intraoperative neuromonitoring (IONM) provides electromyographic (EMG) feedback and facilitates identification and functional assessment of the RLN and EBSLN during surgery, thereby reducing neural complications [4, 5].

Because the thyroid gland is highly vascularized, postoperative hematoma remains a clinically significant concern [6]. Although the overall incidence is low (0.43–6.54%), an expanding cervical hematoma can rapidly cause airway compression and respiratory distress, potentially leading to life-threatening suffocation [7, 8]. Consequently, the use of hemostatic materials (HMs) has become a common adjunct after thyroid surgery to reduce bleeding, decrease drainage volume, and mitigate complications such as hematoma, seroma, and prolonged hospitalization[9].

Two widely used HMs, Tisseel (TIS; Baxter AG, Vienna, Austria) and 4DryField PH (4DF; PlantTec Medical GmbH, Lüneburg, Germany), operate via distinct mechanisms. TIS is a fibrin-based sealant, mimics the terminal coagulation cascade to create a stable fibrin matrix, thereby reducing postoperative dead space and enhancing hemostasis[10, 11]. Its benefits in the Head and Neck surgery, including lower drainage volume and shorter hospital stay, have been well documented[12]. Lower drainage volume with the TIS application in the thyroid surgery was also reported[11]. In contrast, 4DF is a polysaccharide-based powder that absorbs fluid, forms a gel-like matrix, and exhibits anti-adhesive properties[13]. Studies in orthopedic and gynecologic have shown that 4DF effectively reduces blood loss [14, 15]. Animal peritoneal studies have also showed the capability of postoperative adhesion prevention. [16, 17] Recent evidence demonstrates that 4DF can be effectively used in thyroidectomy to guide selective drain placement and control postoperative drainage, even in higher-risk patients, without increasing bleeding-related complications [18].

Despite the hemostatic benefits of both TIS and 4DF, their potential influence on postoperative voice and swallowing outcomes remains unclear. When applied in thyroid surgery, the robust fibrin matrix formed by TIS providing strong tissue sealing may inadvertently limit the gliding mechanism of the extralaryngeal muscles, resulting in mechanical restriction [19]. The anti-adhesive characteristics attributed to 4DF may help preserve tissue mobility. These differences raise the possibility of non-neural post-thyroidectomy dysphonia (PTD). In this study, IONM was used to confirm the preserved RLN and EBSLN, thereby minimizing neural factors to postoperative functional outcome. On this basis, the study aimed to investigate the effects of different HM applications on postoperative voice and swallowing outcomes in nerve-preserved total thyroidectomy, and to provide evidence supporting a selective HM application strategy.

Methods

Patients

This retrospective study enrolled 214 adult patients (age > 18 years) who underwent primary nerve-preserved thyroid surgery without lateral neck dissection between June 2021 and December 2023 at a tertiary academic and referral center. All procedures were performed by the IONM surgical team using intraoperative neuromonitoring (IONM; Medtronic NIM system, Jacksonville, FL, USA) and the Ligasure Exact Dissector (LED; Medtronic, Minneapolis, MN, USA). Nerve preservation was confirmed when all of the following criteria were met: (1) the RLN demonstrated less than a 20% reduction in electromyographic (EMG) amplitude after dissection compared with its baseline value; (2) proximal stimulation of the external branch of the superior laryngeal nerve (EBSLN) elicited a visible cricothyroid muscle (CTM) twitch; and (3) postoperative fiber laryngoscopy showed intact and symmetric vocal fold motion.

Exclusion criteria were as follows: (1) use of HM other than TIS or 4DF (n = 11); (2) history of previous thyroid or neck surgery or prior neck irradiation (n = 9); (3) incomplete medical records or missing subjective/objective voice assessments (n = 17); and (4) creation of a U-shaped muscle flap or transection of at least one strap muscle (n = 2).

Patients were further classified according to their preoperative, informed decision regarding the use of HM into three groups: non-hemostatic material (NHM), 4DF, and TIS. The study flowchart is presented in Fig. 1. Ethical approval for this study was obtained from the Institutional Review Board (****IRB-E(I)−20,250,105).

Fig. 1.

Flowchart of patient enrollment and group allocation

HM application, and drain placement

All procedures involved standard exposure and removal of the thyroid lesion. Bleeding check was assessed following lesion removal, and wound closure was initiated once adequate hemostasis had been achieved. In the NHM group, after hemostasis was completed, a drain was routinely placed, followed by layered wound closure. For the TIS and 4DF groups, the same general protocol was applied after completion of hemostasis. A uniform layer of HM was applied to all surgical planes deep to the strap muscles, including the operative surface of thyroid lobes and the pretracheal plane, covering the cartilages and the surface of the CTM, using 3–3.5 mL of TIS (sprayed evenly with an air-powered applicator) in the TIS group (Fig. 2A), and 1–3 g of 4DryField PH manually spread in the 4DF group (Fig. 2B). After application, the operative field was observed for approximately 30 s. If persistent oozing was noted on reassessment, the hemostatic status of the affected area was re-evaluated, and an additional amount of TIS or 4DF was applied accordingly. If hemostasis was achieved, the wound was closed without drain placement; if bleeding continued, a drain was inserted. Following confirmation of hemostasis, the midline-divided strap muscles were re-approximated. Before skin closure, an additional thin layer of TIS or 4DF was applied over the strap muscles and subplatysmal plane.

Fig. 2.

Application of hemostatic materials to all surgical planes deep to the strap muscles after total thyroidectomy. (A) Tisseel. (B) 4DryField PH

Due to an increasing number of patient-reported subjective symptoms associated with TIS, and following discussions among patient groups and with the surgical team, our practice pattern regarding TIS was adjusted after October 2022. Thereafter, TIS was no longer routinely offered as an option for patients undergoing primary thyroid surgery without lateral neck dissection and was reserved mainly for selected cases with a higher anticipated bleeding risk, particularly those undergoing lateral neck dissection involving level IV, which were not included in the present study.

Except for the reduced availability of TIS as a preoperative HM option after October 2022, no other changes were made during the study period in patient selection criteria, surgical technique, IONM protocol, NHM and 4DF management, drain-placement strategy, or the schedule of voice and swallowing assessments.

Data collection

Demographic and operative variables were collected, including sex, age, surgical extent (unilateral vs. bilateral thyroidectomy), whether central neck dissection (CND) was performed, and whether a drain was placed. Pathologic information comprised the final diagnosis (benign or malignant thyroid disease), main tumor volume (mL), and specimen weight (g). Preoperative thyroid function tests and antibody profiles were reviewed. Serum free thyroxine (free T4; reference range 0.7–1.48 ng/dL), thyroid-stimulating hormone (TSH; 0.35–4.94 μIU/mL), and thyroglobulin (Tg; < 50 ng/mL) were measured. Thyroid antibody levels—including thyroglobulin antibody (TgAb; < 40 IU/mL) and thyroid microsomal antibody (TMAb; < 35 IU/mL)—were assessed to determine whether either marker was elevated. Postoperative outcomes were also documented, including the presence of complications such as hematoma or seroma formation and local wound infection.

Objective and subjective voice analysis were conducted in a standardized manner at two scheduled time points: preoperatively (within two months before surgery) and at six weeks after surgery. Objective voice analyses were obtained using the Multidimensional Voice Program (MDVP; Model 5105, Version 3.1.7, KayPENTAX, USA), which generated data on maximum phonation time (MPT), mean fundamental frequency (Mean F₀), jitter, shimmer, and the noise-to-harmonic ratio (NHR). Voice frequency characteristics were further evaluated with the Voice Range Profile (VRP; Model 4326, Version 3.3.0, KayPENTAX, USA), providing maximum pitch frequency (F_max), minimum pitch frequency (F_min), and pitch range (PR), calculated as the number of semitones between F_max and F_min.

Subjective vocal analyses were assessed using several validated instruments: the Voice Handicap Index-10 (VHI-10; score range 0–40), the VHI-30 (0–120), the Voice-Related Quality of Life (VRQoL; 0–100), and the Index of Voice and Swallowing Handicap of Thyroidectomy (IVST), a tool specifically developed for patients undergoing thyroid surgery. The IVST comprises a total score (IVST-T; 0–20) and two subscores addressing voice symptoms (IVST-V; 0–14) and swallowing symptoms (IVST-S; 0–6).

For analysis, changes in objective measures were expressed as Δ = (postoperative – preoperative)/preoperative, whereas subjective score changes were defined as Δ = postoperative – preoperative.

Statistical analysis

All statistical analyses were conducted using R software (version 4.4.3; R Foundation for Statistical Computing, Vienna, Austria). Categorical variables were summarized as frequencies and percentages, whereas continuous variables were reported as mean ± standard deviation (SD). Comparisons between groups were performed using independent t-tests for continuous data and chi-square tests for categorical data, with Fisher’s exact test applied when expected cell counts were below five. Multivariable logistic regression was performed with NHM as the reference group. All tests were two-sided, and statistical significance was defined as a p-value < 0.05.

Results

Patient clinical and surgical characteristics

A total of 175 patients were included in the study across three groups, 74 in the NHM group, 30 in the TIS group, and 71 in the 4DF group (Table 1). Sex, age, surgical extent (bilateral vs. unilateral), CND (with vs. without), thyroid pathology (benign vs. malignant) showed no significant difference among the groups. Main tumor size and specimen weight also demonstrated no statistical difference among the three groups. Thyroid profiles, including free T4, TSH, Tg, and the proportion of patients with elevated thyroid antibodies, were likewise comparable (p > 0.05).

Table 1.

Patient clinical and surgical characteristics of the non-hemostatic material (NHM), Tisseel (TIS), and 4DryField PH (4DF) groups

	NHM	TIS	4DF	p value
Case number	74	30	71
Sex				0.130
Female	51 (68.9%)	18 (60.0%)	56 (78.9%)
Male	23 (31.1%)	12 (40.0%)	15 (21.1%)
Age				0.302
mean ± SD (year)	49.6 ± 15.9	49.7 ± 8.3	52.8 ± 11.6
Surgical extent				0.428
Bilateral (%)	50 (67.6%)	23 (76.7%)	45 (63.4%)
Unilateral (%)	24 (32.4%)	7 (23.3%)	26 (36.6%)
Central neck dissection				0.646
With (%)	26 (35.1%)	13 (43.3%)	24 (33.8%)
Without (%)	48 (64.9%)	17 (56.7%)	47 (66.2%)
Pathology				0.570
Benign (%)	46 (62.2%)	17 (56.7%)	38 (53.5%)
Malignant (%)	28 (37.8%)	13 (43.3%)	33 (46.5%)
Main tumor size				0.764
mean ± SD (mL)	61.5 ± 42.8	56.0 ± 35.6	64.3 ± 64.7
Specimen weight				0.666
mean ± SD (g)	31.1 ± 21.1	29.1 ± 16.2	34.1 ± 34.1
Free T4				0.867
mean ± SD (ng/dL)	1.13 ± 0.35	1.13 ± 0.27	1.16 ± 0.18
TSH				0.413
mean ± SD (µIU/mL)	2.61 ± 7.88	1.47 ± 0.69	1.58 ± 1.37
Tg				0.286
mean ± SD (ng/mL)	96.8 ± 209.4	43.9 ± 34.1	74.5 ± 118.1
Elevated thyroid antibody level	28 (37.8%)	7 (23.3%)	14 (19.7%)	0.307
Drain placement (%)	74 (100.0%)	0 (0.0%)	14 (19.7%)	< 0.001*
Complications
Hematoma/Seroma	0 (0%)	0 (0%)	0 (0%)
Local infection	0 (0%)	0 (0%)	0 (0%)

In contrast, drain placement differed significantly (p < 0.001). Owing to routine practice, all NHM patients had drains placed. In comparison, the use of TIS completely eliminated the need for drain placement (0% vs. 100% in NHM), whereas drain placement was required in 19.7% of patients in the 4DF group following a standardized 30-s observation after 4DF application. Importantly, no postoperative hematoma, seroma, or local infection was observed in any group.

Voice parameter changes and intergroup comparisons

Changes (Δ) in objective and subjective voice parameters for the three groups were summarized in Table 2. For each parameter, changes were compared pairwise across NHM, TIS, and 4DF. For objective outcomes, most parameters showed no significant intergroup difference; however, the TIS group exhibited a significantly greater decline in ΔPR compared with NHM (p = 0.016). A similar pattern was observed for ΔF_Max, the TIS group showed a larger reduction than both NHM and 4DF, although the differences did not reach statistical significance (p ≈ 0.06–0.07). Notably, the comparison between TIS and 4DF for ΔPR was limited by the markedly larger standard deviation in the 4DF group; otherwise, the mean difference between these two groups would have been more pronounced.

Table 2.

Changes in Voice Parameters among the NHM, TIS, and 4DF groups and Intergroup Comparisons

	NHM	TIS	4DF	p value
	Mean ± SD	Mean ± SD	Mean ± SD	NHM vs. TIS	TIS vs. 4DF	NHM vs. 4DF
ΔMPT	−3.1 ± 22.6%	−11.4 ± 17.5%	−4.9 ± 24.1%	0.051	0.136	0.654
ΔMean F0	4.7 ± 85.1%	−5.2 ± 13.1%	0.0 ± 25.8%	0.339	0.193	0.652
ΔJitter	47.3 ± 191.3%	27.7 ± 74.8%	45.6 ± 127.3%	0.460	0.389	0.949
ΔShimmer	26.2 ± 76.1%	17.2 ± 55.8%	8.7 ± 51.4%	0.509	0.485	0.108
ΔNHR	12.5 ± 39.0%	3.8 ± 16.6%	5.1 ± 27.6%	0.118	0.774	0.192
ΔFMax	−14.9 ± 30.1%	−25.4 ± 23.0%	−15.8 ± 23.8%	0.062	0.067	0.839
ΔFmin	7.1 ± 55.2%	4.4 ± 27.6%	−2.5 ± 20.3%	0.739	0.231	0.166
ΔPR	−7.1 ± 26.5%	−19.7 ± 21.9%	−4.3 ± 73.3%	0.016*	0.114	0.762
ΔVHI-10	1.7 ± 5.7	3.4 ± 6.4	0.9 ± 4.9	0.221	0.063	0.345
ΔVHI-30	4.4 ± 15.4	12.4 ± 21.3	4.3 ± 13.2	0.074	0.064	0.954
ΔVRQoL	1.3 ± 4.0	4.3 ± 5.9	1.0 ± 3.2	0.019*	0.008*	0.562
ΔIVST-T	1.5 ± 4.3	3.4 ± 3.2	0.9 ± 2.5	0.022*	0.001*	0.290
ΔIVST-V	1.0 ± 3.0	2.4 ± 2.9	0.8 ± 1.9	0.040*	0.010*	0.575
ΔIVST-S	0.4 ± 1.9	1.0 ± 2.2	0.1 ± 1.0	0.244	0.057	0.251

For subjective outcomes, the TIS group demonstrated the greatest postoperative deterioration, showing significantly greater changes in VRQoL, IVST-T, and IVST-V compared with both the NHM and 4DF groups. These findings corresponded with the increasing number of patient-reported symptoms associated with TIS, which contributed to an adjusted practice pattern after October 2022. In addition, no significant differences in voice parameters were observed between the NHM and 4DF groups.

In multivariable logistic regression analysis, TIS use was associated with ΔPR reduction ≥ 20% compared with NHM (OR 2.91, 95% CI 1.06–8.01, p = 0.039). Specimen weight was also associated with ΔPR reduction ≥ 20% (OR 1.06, 95% CI 1.01–1.10, p = 0.020). No other covariates showed statistically significant associations with ΔPR reduction or ΔIVST-T elevation. For ΔIVST-T elevation ≥ 4, TIS showed a similar direction of effect compared with NHM, although this did not reach statistical significance (OR 2.70, 95% CI 0.89–8.19, p = 0.079) (Table 3).

Table 3.

Multivariable logistic regression analysis of factors associated with ΔPR and ΔIVST-T

	ΔPR ↓ ≥ 20%		ΔIVST-T ↑ ≥ 4
	OR (95% CI)	p value	OR (95% CI)	p value
4DF vs NHM	1.14 (0.50–2.62)	0.759	0.38 (0.13–1.06)	0.064
TIS vs NHM	2.91 (1.06–8.01)	0.039*	2.70 (0.89–8.19)	0.079
Sex	1.95 (0.85–4.43)	0.113	0.49 (0.16–1.48)	0.208
Age	1.00 (0.97–1.03)	0.949	1.03 (0.99–1.06)	0.146
Surgical extent	0.66 (0.26–1.65)	0.371	0.81 (0.26–2.52)	0.721
Central neck dissection	0.98 (0.43–2.23)	0.954	1.08 (0.42–2.80)	0.877
Pathology	0.77 (0.33–1.82)	0.555	1.57 (0.57–4.28)	0.381
Main tumor size	0.98 (0.96–1.00)	0.057	1.01 (0.98–1.04)	0.605
Specimen weight	1.06 (1.01–1.10)	0.020*	0.98 (0.93–1.04)	0.566
TSH	1.12 (0.83–1.51)	0.474	0.76 (0.50–1.17)	0.212
Tg	1.00 (1.00–1.00)	0.372	1.00 (0.99–1.00)	0.303
Elevated thyroid antibody level	0.71 (0.31–1.63)	0.419	0.47 (0.19–1.18)	0.107

Comparison of abnormal voice-related outcomes between NHM and 4DF

To further evaluate whether the proportion of patients showing postoperative changes differed between NHM and 4DF, we analyzed the distribution of patients exceeding predefined thresholds of abnormal voice or swallowing outcomes (Table 4). The results showed that the NHM group had a significantly higher proportion of patients with increased IVST-S compared with the 4DF group (41.9% vs. 21.1%, p = 0.007). For all other objective and subjective voice parameters, no significant differences were observed between the two groups.

Table 4.

Comparison of abnormal voice-related outcomes between the NHM and 4DF groups

	Definition of Abnormal	NHM	4DF	p value
ΔMPT	↓≧20%	12 (16.2%)	18 (25.4%)	0.175
ΔMean F0	↓or↑≧20%	16 (21.6%)	11 (15.5%)	0.343
ΔJitter	↑≧20%	29 (39.2%)	32 (45.1%)	0.473
ΔShimmer	↑≧20%	30 (40.5%)	21 (29.6%)	0.167
ΔNHR	↑≧20%	21 (28.4%)	13 (18.3%)	0.153
ΔFMax	↓≧20%	32 (43.2%)	29 (40.8%)	0.770
ΔFmin	↓or↑≧20%	20 (27.0%)	19 (26.8%)	0.971
ΔPR	↓≧20%	22 (29.7%)	21 (29.6%)	0.984
ΔVHI-10	↑≧5	18 (24.3%)	10 (14.1%)	0.118
ΔVHI-30	↑≧10	18 (24.3%)	16 (22.5%)	0.799
ΔVRQoL	↑≧5	9 (12.2%)	8 (11.3%)	0.867
ΔIVST-T	↑≧4	17 (23.0%)	9 (12.7%)	0.106
ΔIVST-V	↑≧3	17 (23.0%)	12 (16.9%)	0.361
ΔIVST-S	↑≧1	31 (41.9%)	15 (21.1%)	0.007*

Discussion

This study highlights several clinically relevant findings regarding the influence of different HMs on postoperative voice and swallowing outcomes in nerve-preserved total thyroidectomy. First, although the use of TIS provides a relative advantage over NHM and 4DF by effectively eliminating the need for drain placement, its application is associated with consistently less favorable functional postoperative voice outcomes. Despite the absence of preoperative demographic or surgical differences among the groups (Table 1), the TIS cohort demonstrated significantly greater changes across multiple objective (ΔPR, trend in ΔFMax) and subjective (VRQoL, IVST-T, IVST-V) voice parameters, even within a relatively small sample of 30 patients (Table 2). Using NHM as the reference, multivariable logistic regression further showed that TIS remained associated with clinically relevant PR reduction after adjustment, with a similar but non-significant trend for IVST-T elevation (Table 3). These findings may be associated with the mechanical restriction related to fibrin sealants, which may limit the mobility of the strap muscles and CTM, thereby affecting high-frequency voice and overall voice quality. This pattern aligns with the increasing frequency of patient-reported symptoms, which prompted an adjustment in our practice pattern. In contrast, the 4DF group showed postoperative voice outcomes comparable to the NHM group across all parameters. However, the proportion of patients with abnormal swallowing scores (IVST-S) was significantly higher in the NHM group than in the 4DF group (Table 4), suggesting that any mobility-preserving effect of 4DF may be more evident in swallowing function rather than phonation. These findings also underscore the importance of incorporating swallowing assessment into postoperative evaluation after thyroidectomy.

The most common complaints of PTD include hoarseness, transient vocal fatigue, and pitch changes [3]. Standardized IONM in thyroid surgery serves as a cornerstone for the evaluation of postoperative voice outcomes. According to the recommendations of the International Neural Monitoring Study Group (INMSG), informed consent for IONM is essential and should be incorporated into the preoperative planning for all patients undergoing thyroid and parathyroid surgery, ensuring effective shared decision-making and patient understanding of the role and limitations of this technology. [20] A voice-centered thyroidectomy protocol should include both preoperative and postoperative voice assessments as well as laryngoscopic examinations. [3] Intraoperatively, a standardized IONM steps following the V1–R1–S1–S2–R2–V2 protocol should be performed, with particular emphasis on the assessment of proximal and distal R2 signals (R2p and R2d), which is critical for evaluating RLN integrity and guiding intraoperative decision-making. [21, 22] In addition, monitoring of EBSLN with confirmation of appropriate CTM contraction is recommended to verify neural integrity and to assess the risk of high-pitched voice impairment [23, 24].

Non-neural factors for PTD have been widely recognized as important contributors to postoperative voice impairment after thyroid surgery. These factors include intubation- and anesthesia-related laryngeal irritation, postoperative local edema or confined hematoma, as well as intraoperative traction or blunt injury to the strap muscles and other extralaryngeal musculature [3]. Previous studies have shown that intubation- and anesthesia-related factors, together with postoperative edema or limited hematoma formation, are present in nearly all patients undergoing thyroid surgery, and their impact on voice function is predominantly confined to the early postoperative period, particularly within the first two weeks, during which these effects are generally transient and reversible [25–27]. In contrast, intraoperative over-traction or muscular blunt trauma, although capable of affecting extralaryngeal muscle function and voice stability [24], was observed across all study groups in the present cohort and did not demonstrate a meaningful association with postoperative voice outcomes, suggesting that it is not a primary determinant of voice impairment in this study. In addition, postoperative adhesive fibrosis has been increasingly recognized as an important non-neural contributor to functional impairment after thyroidectomy. Cho et al. demonstrated, using ultrasound, that impaired laryngotracheal elevation after thyroidectomy is associated with postoperative swallowing difficulty, suggesting adhesion between the laryngotracheal unit and superficial cervical soft tissues as a potential mechanism. [28] Similarly, Hyun et al. reported that conventional open thyroidectomy is associated with increased adhesion between the subplatysmal flap and the underlying strap muscles compared with gasless transaxillary endoscopic approaches, supporting the concept that postoperative adhesion at this interface may interfere with normal strap muscle mobility [29]. Therefore, based on both the current findings and existing literature, after excluding overt neural injury, the principal non-neural factors influencing postoperative voice function in this study may be associated with the mechanical restriction resulting from a limited gliding mechanism between adjunct extralaryngeal structures. These structural changes, particularly involving the strap muscles, the CTM, and the laryngotracheal unit, may limit coordinated motion, thereby impairing phonatory coordination and voice stability, rather than reflecting transient postoperative changes alone.

The impact of HMs on postoperative voice outcomes may arise from their mechanical interactions with adjacent muscles, particularly the CTM and the strap muscles [30, 31]. CTM gives a significant contribution to the production of high-frequency sounds in humans[32]. Infrahyoid strap muscles help stabilizing the larynx and assist in the control of vocal fold tension and thereby indirectly influencing pitch and resonance [33]. Notably, previous studies suggest that postoperative functional impairment is not solely determined by structural muscle disruption. Aygun et al. demonstrated that strap muscle transection did not significantly worsen postoperative voice outcomes compared with muscle preservation [33], and Ko et al. further reported comparable voice outcomes even with extensive strap muscle mobilization using a U-shaped strap muscle flap technique [34]. Based on the needle-EMG study by Roubeau et al. (1997), the strap muscles show a distinct re-activation pattern at the upper limit of pitch elevation, although the CTM provides the primary mechanism of vocal fold elongation, the strap muscles become increasingly recruited during high-frequency phonation to stabilize the laryngeal framework, modulate vertical laryngeal position, and assist in maintaining the tension required for extreme F₀ production[35]. Therefore, as these muscles decreased contraction, it may be associated with limiting fine pitch modulation and reduced Fmax[32, 36]. HMs with strong tissue-sealing characteristics, such as TIS, may inadvertently impair postoperative voice by limiting the normal gliding mechanism of adjacent muscle. Mechanical restriction involving the CTM may limit effective vocal fold elongation and tension control, selectively reducing high-frequency phonation and maximum pitch. Conversely, mechanical restriction affecting the strap muscles may compromise vertical laryngeal mobility and stabilization, may be associated with more complex and heterogeneous vocal changes that are not readily captured by a single objective voice parameter. The differential functional effects on specific muscle groups help explain the discrepancy observed in this study between objective and subjective voice analyses. Although the TIS group did not demonstrate uniform changes across all objective voice parameters, a downward trend was observed in pitch-related measures, accompanied by a significant increase in patient-reported voice impairments. Subjective voice parameters, including the VRQoL, IVST-T, and IVST-V, showed significant declines, reflecting the overall functional burden perceived by patients during everyday voice use rather than an isolated voice deficit. These findings suggest that mechanical restriction related to a limited gliding mechanism may disrupt the fine coordination required for stable and comfortable phonation, even when certain basic vocal tasks remain achievable. In contrast, HMs with anti-adhesive properties, such as 4DF, appear to better preserve postoperative muscle mobility and tissue compliance, thereby minimizing interference with the function of the CTM and strap muscles. Overall, our results underscore that in thyroid surgery where voice outcomes are a critical consideration, hemostatic efficacy alone is insufficient as a surgical endpoint. Mechanical restriction induced by HMs, and its differential impact on extralaryngeal muscle groups, may exert a substantial and disproportionate effect on patient-perceived voice quality, even when objective acoustic changes are relatively subtle.

In the comparison between the 4DF and NHM groups, the 4DF group demonstrated a significantly lower incidence of abnormal IVST-S findings (p = 0.007). PTD can be broadly categorized into neurogenic and non-neurogenic factors [37]. Previous studies have reported that resection of the inferior pharyngeal constrictor muscles (IPCMs) during thyroid surgery is significantly associated with prolonged dependence on tube feeding [38]. In the present study, after minimizing RLN and EBSLN injuries through the use of IONM, dysfunction of the IPCMs and insufficient laryngotracheal elevation due to mechanical restriction involving the laryngotracheal unit may be emerged as the primary considerations [28, 39]. Although swallowing-related muscle function may be influenced by other factors, such as surgical traction or inflammatory responses, reducing postoperative mechanical restriction is likely to have a beneficial effect in preventing postoperative swallowing dysfunction after thyroid surgery. Further studies specifically designed to investigate this finding are warranted.

From a clinical perspective, the findings of this study suggest that the selection and application of HMs in thyroid surgery should be guided not only by hemostatic efficacy but also by anticipated functional impact. Preoperative or intraoperative estimation of bleeding risk and the likelihood of drain placement may assist surgeons in selecting an appropriate HM. When the use of fibrin sealants such as Tisseel is deemed necessary, particularly in cases with diffuse oozing, selective application should be emphasized. Excessive or indiscriminate spraying over the pretracheal surface and bilateral upper pole regions should be avoided to minimize interference with the gliding mechanism of the CTM and strap muscles. In addition, given the strong tissue-sealing effect provided by TIS, postoperative management may require adjustment, including reduction or avoidance of compressive neck dressings to prevent additional external constraint. Finally, for patients receiving TIS, early initiation of neck exercise and, when clinical indicated, voice or swallowing therapy may be beneficial to reduce mechanical restriction.

This study has several limitations that should be acknowledged. First, long-term follow-up of postoperative voice and swallowing function in patients undergoing nerve-preserved total thyroidectomy is inherently challenging in routine clinical practice. As a result, the overall sample size in this study was limited. In addition, due to increasing patient-reported voice-related discomfort associated with the use of TIS, enrollment in the TIS group was intentionally reduced during the later phase of the study following the adjustment of its availability after October 2022. However, aside from the reduced availability of TIS as a preoperative option, no other changes were made in patient selection criteria or perioperative management protocols. Therefore, the potential impact of temporal selection bias is considered limited, although it cannot be completely excluded. Furthermore, postoperative analyses were primarily conducted at six weeks after surgery. Therefore, the present findings should be interpreted specifically as early postoperative outcomes and should not be extrapolated to long-term voice or swallowing function. The six-week postoperative visit captures a clinically important phase in which overall voice and swallowing symptoms are often most apparent after the immediate postoperative inflammatory effects have subsided. Long-term follow-up remains meaningful and should be addressed in future prospective studies using parameters specifically designed to evaluate sustained functional recovery, late mechanical restriction or fibrosis, and symptom resolution over time. Extended follow-up is also susceptible to attrition bias, as patients with symptom resolution are less likely to seek continued evaluation, whereas those with persistent complaints may be overrepresented. Second, the mechanisms underlying postoperative voice and swallowing changes after thyroidectomy are multifactorial and complex. Despite careful surgical technique, numerous neural and non-neural factors may contribute to functional outcomes. In the present study, the use of standardized IONM allowed us to minimize and effectively exclude overt neural injury to the RLN and EBSLN, thereby reducing a major source of confounding. This approach enabled a more focused evaluation of non-neural factors, particularly the role of HMs, under relatively controlled conditions. By limiting intergroup differences primarily to the type of HM used, the study was able to delineate distinct functional profiles associated with different HMs. However, direct measurement of mechanical restriction (e.g., imaging or functional muscle assessment) was not performed in the present study. While causality cannot be definitively established, these findings not only highlight clinically relevant differences among HMs but also provide an important framework and necessary control conditions for future investigations into postoperative mechanical restriction, fibrosis, and their functional consequences following thyroid surgery.

Conclusion

In this study of nerve-preserved total thyroidectomy, we demonstrated that the choice of HMs is associated with clinically meaningful differences in postoperative voice and swallowing function. Although fibrin sealants such as Tisseel offer reliable hemostasis and may eliminate the need for drain placement, their strong tissue-sealing effect may be associated with mechanical restriction and a limited gliding mechanism involving the CTM and strap muscles, thereby predisposing patients to impaired high-pitch voice and deterioration in overall voice-related quality of life.

In contrast, the polysaccharide-based agent 4DryField PH showed non-inferior voice preservation and a lower incidence of swallowing dysfunction compared with the NHM group. These findings highlight that, even in the absence of neural injury, non-neural factors, particularly mechanical restriction affecting extralaryngeal muscle coordination, play a significant role in postoperative functional outcomes.

Therefore, functional considerations, especially those related to high-pitch voice and swallowing, should be incorporated into the selection and application of HMs in thyroid surgery. A personalized, anatomy-aware approach, taking into account anticipated bleeding risk, the likelihood of drain placement, and the need for selective application to avoid critical voice-related structures, may help balance hemostatic efficacy with functional preservation. Future prospective, randomized studies with longer follow-up are warranted to validate these findings and to further elucidate how different HMs influence postoperative laryngeal function.

Funding

The study was supported by grants from the National Science and Technology Council, Taiwan (NSTC 113-2314‐B‐037 − 033 ‐ MY2, NSTC 114-2314-B-037-074); Kaohsiung Medical University Gangshan Hospital (KMGH-113R004, KMGH-114-R011), Kaohsiung Medical University, Kaohsiung, Taiwan.

Data availability

The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding authors.

Declarations

Conflicts of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.