Abstract
Background
Compensatory hyperhidrosis (CH) remains the most prevalent postoperative adverse event following endoscopic thoracic sympathectomy (ETS) for primary hyperhidrosis (PH). Current predictive models lack reliability in estimating CH severity. This study introduces a novel predictive framework utilizing rough set theory to establish decision rules for CH stratification.
Methods
In this single‑center retrospective cohort study, clinical data from 225 PH patients undergoing ETS were analyzed, including 37 predictive indicators. These variables were subjected to correlation analysis, regression analysis, and rough set analysis with CH severity.
Results
There were 93.3% (210/225) of patients exhibiting CH following ETS, with 33.3% classified as grade III CH, and no grade IV CH was noted. Body mass index (BMI), the level of sympathectomy, and the temperature difference of the right hand after surgery and before anaesthesia were shown to be significantly correlated with CH on correlation analysis. However, no valid regression model was established with significant correlations involving indicators for further regression analysis. By switching to rough set analysis, four predictive rules for grade III CH were derived: (I) BMI >22 kg/m2 + initial onset age of PH >11 years, 84% accuracy; (II) BMI 19.5–22 kg/m2 + surgical age >28.5 years, 82% accuracy; (III) BMI 18.5–19.4 kg/m2 + postoperative right-hand temperature >36.6 ℃, 77% accuracy; (IV) BMI <18.5 kg/m2 + postoperative right-hand temperature <37.0 ℃ + initial PH onset age <10 years, 71% accuracy.
Conclusions
Rough set analysis provides a promising approach for exploring the patterns of CH severity following ETS in patients with PH, and thus which merits further investigation through multicenter, large-sample studies. The four preliminary decision rules for predicting grade III CH derived from rough set analysis show potential clinical relevance but remain tentative, as their utility requires validation in prospective cohorts prior to widespread clinical application.
Keywords: Compensatory hyperhidrosis (CH), primary hyperhidrosis (PH), prediction, rough set analysis, endoscopic thoracic sympathectomy (ETS)
Highlight box.
Key findings
• Following endoscopic thoracic sympathectomy (ETS) for primary hyperhidrosis (PH), compensatory hyperhidrosis (CH) occurred in 93.3% of patients, with 33.3% classified as grade III severity.
• Utilizing rough set theory, four decision rules for predicting grade III CH were derived, incorporating body mass index (BMI), age at PH onset, surgical age, and postoperative hand temperature, with predictive accuracies ranging from 71% to 84%.
What is known and what is new?
• CH is the most prevalent adverse event after ETS, but existing models lack reliability in predicting its severity.
• This study introduces a novel predictive framework based on rough set analysis, which successfully generated interpretable decision rules for CH stratification when conventional correlation and regression analyses failed to establish a valid model.
What is the implication, and what should change now?
• Rough set analysis offers a promising alternative approach for understanding the patterns of CH severity and could enhance preoperative risk assessment.
• The derived predictive rules show clinical potential but are preliminary; their utility must be validated in prospective, multicenter, large-sample studies prior to any clinical application.
Introduction
Endoscopic thoracic sympathectomy (ETS) has an effective rate of nearly 100% (1,2) and is currently the most recognised therapeutic option for primary hyperhidrosis (PH). Owing to the progress of minimally invasive surgery technology, such as single-port thoracoscopic surgery and surgery using the laryngeal mask airway without endotracheal intubation (3-6), there has been a gradual increase in the number of patients undergoing ETS recently. ETS is commonly a highly safe operation, and the perioperative-related complications are only about 0.1%. However, postoperatively, compensatory hyperhidrosis (CH) of all severitiescan be as high as 14–90%, with grade IV CH at 3–5%, which causes long-term and serious difficulties to patients (7-9). Reducing the incidence or severity of CH has remained a key challenge for thoracic surgeons.
The mechanism underlying CH remains poorly understood. Some studies suggest that CH may result from compensatory feedback enhancement of the hypothalamic thermoregulatory center following sympathetic chain disruption, leading to hyperactivation of sweat glands in non-surgical areas (1,10). However, whether this mechanism directly correlates with specific clinical indicators (e.g., postoperative temperature changes) requires further validation. While certain studies propose that isolated R4 (fourth thoracic sympathetic ganglion) sympathectomy during initial surgery achieves satisfactory outcomes with reduced CH rates (approximately 41.99%), moderate-to-severe CH still occurs in 7.88% of cases (11-14).
In the absence of definitive techniques to substantially mitigate CH, identifying preoperative predictors of CH severity could guide surgical decision-making—such as avoiding ETS in high-risk patients or implementing intraoperative measures to alleviate CH. To address this, we retrospectively analyzed clinical data from PH patients who underwent ETS at our institution between 2018 and 2024, supplemented by postoperative CH assessments via WeChat-based questionnaires. We aimed to identify clinical features predictive of CH severity and construct a predictive model. Notably, this study is the first to apply rough set analysis, a statistical method, to CH prediction, yielding promising results. The core reasons for selecting rough set analysis lie in its three advantages that fit clinical prediction scenarios: (I) it can generate clear and human-readable decision rules, which can be directly applied without professional algorithmic knowledge; (II) it does not require strong distributional assumptions and can flexibly handle discrete attributes and mixed-type variables; (III) it focuses on attribute reduction, enabling the screening of core factors from a large number of candidate indicators to reduce model complexity. Compared with other nonlinear methods such as decision trees (with redundant rules) and random forests/gradient boosting (with unexplainable “black-box” models), it is more aligned with the core objective of this study—“constructing clinically operable prediction rules”. We present this article in accordance with the TRIPOD reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-aw-2074/rc).
Methods
Surgical patient selection
This is a single-center retrospective cohort study. The study subjects were recruited from patients with PH treated from October 2018 to October 2024 in our hospital. The diagnostic PH criteria mainly referred to the macroscopic hypersecretion of sweat glands for >6 months without obvious inducement. This was formulated by Hornberge et al. (15) of the American Dermatology Association in 2004, an expert collaboration group. Meanwhile, it can be diagnosed if it meets two of the following criteria at the same time: (I) symmetrical sites of hyperhidrosis bilaterally; (II) at least once a week; (III) age of onset <25 years old; (IV) positive family history; (V) no hyperhidrosis during sleep; and (VI) with impact on daily work and life. The PH symptoms were divided into three degrees according to the Lai classification (16): (I) mild: moist palm skin without obvious change in skin temperature; (II) moderate: sweating palms that could wet a handkerchief; (III) severe: sweating palms with drops of sweat.
Surgical patients selection criteria: (I) age ≥16 and ≤50 years; (II) moderate or severe PH diagnosis, without secondary hyperhidrosis by preoperative examination, such as hyperthyroidism; (III) no obvious psychological instability, such as nervousness, depression, irritability, etc.; (IV) no hyperhidrosis in multiple parts of the whole body except axillary hyperhidrosis, foot hyperhidrosis and craniofacial hyperhidrosis (17).
Surgical procedure
Both hands and bilateral armpit temperature of patients were measured by an infrared thermometer before anaesthesia. General anaesthesia was given with a laryngeal mask airway. Both hands and bilateral armpit temperature were again measured using an infrared thermometer after anaesthesia. Patients were adjusted to maintain a semi-recumbent position (45°) with both hands abducted and back raised to fully expose the bilateral subaxillary region. A small 5-mm in length incision was made under both armpits (the 3rd intercostal midaxillary line space). ETS was performed first on the right and then on the left. A 5.5-mm trocar was placed into the chest after the skin was incised with a sharp surgical knife. A 5-mm 30-degree thoracoscope was introduced to confirm the R3/R4 (third/fourth thoracic sympathetic ganglion) sympathetic trunk location after artificial pneumothorax formation. The trocar was withdrawn from the incision (sleeved on the thoracoscope at this time) with the maintenance of the position of the thoracoscope. The electrocoagulation hook was delivered from the incision. The sympathetic trunk was transected on the surface of R3/R4, with the distance between the upper and lower broken ends of >8 mm. Furthermore, the area within 1–2 cm around the sympathetic trunk was ablated along the surface of the rib. The electrocoagulation hook was withdrawn from the incision after satisfactory resection, and the trocar was pushed back into the incision. A 16 G suction catheter was delivered from the trocar and kept without movement with the withdrawal of the thoracoscope. The end of the suction catheter was placed in water to deflate the lungs and expel the accumulated air in the chest after the withdrawal of the trocar. The suction catheter was then pulled out. The incision was bonded with medical glue. The temperature of both hands and bilateral armpits was measured with an infrared thermometer for the third time after operation. Patients were observed in the inpatient ward after operation and were discharged the next morning.
Postoperative follow-up
Daily consultation utilised WeChat and telephone contacts. Questionnaires were sent to each patient through WeChat on the 3rd, 9th and 12th months after operation. Preferentially, data from the 12th-month questionnaire were used; if the patient was less than 12 months postoperatively, data from the 9th-month questionnaire were adopted. Since all patients had passed the 9th postoperative month, 3rd-month questionnaire data were not utilised. The questionnaire contents included: height, weight, native place, permanent residence, occupation, working and living environment temperature, psychological character type, surgical effect, CH condition, surgical incision satisfaction, PH recurrence, life and work improvement after surgery and overall surgery satisfaction (Appendix 1).
Patient eligibility and study inclusion criteria
(I) Patients who underwent bilateral ETS; (II) patients with complete clinical records; (III) patients who had accepted the questionnaire and finished the questionnaire completely and effectively.
Indicators for retrospective analysis
Indicators for this retrospective analysis include gender, age at operation, height, weight, body mass index (BMI), climate zone of origin, climatic zone of the native place, occupation type (pure manual workers, civil workers, professional technicians with more hand operations, students), working and living environment temperatures (low temperature, normal temperature or high temperature), psychological character type (patients were categorized into two types based on self-assessment: (I) relatively stable psychological state with mild character; (II) unstable psychological state with irritable character), preoperative PH location, PH degree, initial age of PH, duration of PH, family history, level of sympathectomy (R3/R4), changes of hand/axillary temperature before/after anaesthesia and after surgery (all measurements were conducted in the operating room at a constant ambient temperature of 25 ℃. Specifically, hand temperature was measured at the center of the palm, and axillary temperature at the center of the axilla. All measurements were performed by the same team of physicians in accordance with the instrument operating instructions, with uniform training provided prior to data collection, surgical effect (remission of PH), occurrence of CH, degree of CH, satisfaction with surgical incision, recurrence of PH, improvement of life and work after surgery and overall satisfaction with surgery (Appendix 2).
The CH classification was based on Tu’s 4-level method (1), as shown in Table 1.
Table 1. Classification of compensatory hyperhidrosis by Tu’s 4-level method.
| Degree of CH | Symptoms |
|---|---|
| Grade I (mild) | The skin is moist, without sweating or any discomfort |
| Grade II (moderate) | In obvious sweating and discomfort, but tolerable |
| Grade III (severe) | Excessive sweating, perspiration can flow, change clothes due to excessive sweating within a day, but tolerable, do not regret the operation |
| Grade IV (very severe) | Excessive sweating, perspiration can flow, seriously affect the quality of life, unbearable, regret surgery |
CH, compensatory hyperhidrosis.
Statistical analysis
The measurement data were expressed in mean ± standard deviation, and the categorical data were presented in numbers (percentage). The observation indicators described above were divided into three categories and represented by a1–a44 and b (Table 2): (I) preoperative condition: gender-family history (a1–a15); (II) intraoperative condition: sympathectomy level (left)-temperature difference after operation and after anaesthesia (right axilla) (a16–a37); (III) postoperative condition: surgical effect (left hand)-overall satisfaction with surgery (a38–a44); and b indicated CH.
Table 2. General data of patients.
| Situation | Observation indicators | Values |
|---|---|---|
| Preoperative situation | Gender | |
| Male | 49.3 (111/225) | |
| Female | 50.7 (114/225) | |
| Age at operation, years | 23.75±6.92 | |
| Height, cm | 166.127±7.201 | |
| Weight, kg | 55.733±8.524 | |
| BMI, kg/m2 | 20.131±2.288 | |
| Climatic zone of native place | ||
| Temperate monsoon climate | 9.3 (21/225) | |
| Subtropical monsoon climate | 90.7 (204/225) | |
| Climatic zone of the long-lived area | ||
| Temperate monsoon climate | 4.0 (9/225) | |
| Subtropical monsoon climate | 96.0 (216/225) | |
| Occupational type | ||
| Office worker | 36.0 (81/225) | |
| Professional and technical personnel with more hand operation | 22.7 (51/225) | |
| Students | 41.3 (93/225) | |
| Working and living environment temperature | ||
| Low temperature (<18 ℃) | 5.3 (12/225) | |
| Normal temperature (18–29 ℃) | 89.3 (201/225) | |
| High temperature (≥30 ℃) | 5.3 (12/225) | |
| Psychological state | ||
| Stable | 81.3 (183/225) | |
| Instable | 18.7 (42/225) | |
| PH location | ||
| Hand | 6.7 (15/225) | |
| Hand + foot | 50.7 (114/225) | |
| Hand + foot + axils | 41.3 (93/225) | |
| Hand + face | 1.3 (3/225) | |
| PH degree | ||
| Mild | 2.7 (6/225) | |
| Moderate | 42.7 (96/225) | |
| Severe | 54.7 (123/225) | |
| Initial age of PH, years | 9.053±3.605 | |
| PH duration, years | 14.693±7.192 | |
| Family history | ||
| Yes | 36.0 (81/225) | |
| No | 64.0 (144/225) | |
| Intraoperative situation | Level of sympathicotomy (left side) | |
| T3 | 97.3 (219/225) | |
| T4 | 2.7 (6/225) | |
| Level of sympathicotomy (right side) | ||
| T3 | 97.3 (219/225) | |
| T4 | 2.7 (6/225) | |
| Left-hand temperature before anesthesia, ℃ | 35.597±0.786 | |
| Right-hand temperature before anesthesia, ℃ | 36.721±0.930 | |
| Left axillary temperature before anesthesia, ℃ | 35.553±0.789 | |
| Right axillary temperature before anesthesia, ℃ | 36.641±0.688 | |
| Left hand temperature after anesthesia | 35.962±3.821 | |
| Right-hand temperature after anesthesia, ℃ | 36.419±0.470 | |
| Left axillary temperature after anesthesia, ℃ | 36.423±0.347 | |
| Right axillary temperature after anesthesia, ℃ | 36.491±0.390 | |
| Postoperative left-hand temperature, ℃ | 37.028±0.321 | |
| Postoperative right-hand temperature, ℃ | 36.488±0.284 | |
| Left hand temperature after anesthesia, ℃ | 37.143±0.322 | |
| Right-hand temperature after anesthesia, ℃ | 36.555±0.262 | |
| Left-hand temperature (postoperative − before anesthesia), ℃ | 1.431±0.744 | |
| Right hand temperature (postoperative − before anesthesia), ℃ | −0.233±1.017 | |
| Left axillary temperature (postoperative − before anesthesia), ℃ | 1.589±0.870 | |
| Right axillary temperature before anesthesia (postoperative − before anesthesia), ℃ | −0.087±0.756 | |
| Left hand temperature (postoperative − after anesthesia), ℃ | 1.065±3.790 | |
| Right hand temperature (postoperative − after anesthesia), ℃ | 0.069±0.646 | |
| Left axillary temperature (postoperative − after anesthesia), ℃ | 0.720±0.417 | |
| Right axillary temperature (postoperative − after anesthesia), ℃ | 0.064±0.507 | |
| Postoperative situation | Effect of surgery (left hand) | |
| Completely relieved | 86.7 (195/225) | |
| Partially relieved | 13.3 (30/225) | |
| Not relieved | 0 (0/225) | |
| Effect of surgery (right hand) | ||
| Completely relieved | 86.7 (195/225) | |
| Partially relieved | 13.3 (30/225) | |
| Not relieved | 0 (0/225) | |
| Effect of surgery (foot) | ||
| Completely relieved | 4.4 (9/204) | |
| Partially relieved | 82.4 (168/204) | |
| Not relieved | 13.2 (27/204) | |
| Effect of surgery (axilla) | ||
| Completely relieved | 22.6 (21/93) | |
| Partially relieved | 64.5 (60/93) | |
| Not relieved | 12.9 (12/93) | |
| Effect of surgery (face) | ||
| Partially relieved | 100 (3/3) | |
| Satisfaction with surgical incision (total score: 10) | 7.760±2.117 | |
| Recurrence of PH | 4.0 (9/225) | |
| Improve living and work | ||
| Significantly improved | 80.0 (180/225) | |
| Partially improved | 18.7 (42/225) | |
| Helpless | 1.3 (3/225) | |
| Overall satisfaction with surgery | 88.013±12.907 | |
| CH | ||
| No CH | 6.7 (15/225) | |
| Grade I | 4.0 (9/225) | |
| Grade II | 56.0 (126/225) | |
| Grade III | 33.3 (75/225) | |
| Grade IV | 0 (0/225) |
Data are presented as mean ± standard deviation or % (n/N). BMI, body mass index; CH, compensatory hyperhidrosis; PH, primary hyperhidrosis.
Spearman correlation analysis was used to analyse the correlation and significance and provided a visualisation of three types of indicators and CH. For the Spearman correlation tests between multiple predictors and CH severity, the false discovery rate (FDR) correction (Benjamini-Hochberg method) was applied to control for multiple comparison bias, with a significance level of α=0.05. The variables with significant correlation were included in the regression analysis with CH, and the variables were screened by stepwise regression. In the multivariate regression analysis, the adjusted R2 was used to evaluate the model’s explanatory power, and the model’s practicality was comprehensively judged by combining coefficient stability and prediction accuracy (calculated based on the confusion matrix). Statistical analysis was performed using the R programming language (R Core Team, 2000). Correlation heatmaps were generated using the ggplot2 package in R, with statistically significant variables annotated in red to emphasize their clinical relevance.
Rough set analysis was performed using Rosetta 2.0 software (University of Warsaw, Poland). A decision table was first constructed, in which the severity of CH (grades I–IV) was defined as the decision attribute, while other parameters served as conditional attributes. For continuous variables, discretization was conducted via Entropy Minimization Discretization to generate optimal cut-off points. Attribute reduction was then performed using the entropy-based method to screen core attributes. Finally, decision rules were induced by the Learning from Examples based on Rough Sets (LERS) algorithm, with the minimum support set to 0.1 and minimum confidence to 0.7 to ensure the generalizability of the rules. Notably, all included data were complete without missing values, so no additional missing data handling strategies were applied. The cut-off points for continuous variables in this study (e.g., BMI stratification, age >28.5 years, initial onset age >11 years, temperature >36.6 ℃) were automatically generated by the Entropy Minimization Discretization algorithm, which are data-driven thresholds aimed at maximizing the discrimination of categories with different CH severity levels.
The inspection level was α=0.05.
Ethical review
The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the Xiamen Humanity Hospital, Fujian Medical University (No. HAXM-MEC-20220519-X002-01) and individual consent for this retrospective analysis was waived.
Results
From 308 eligible patients with complete clinical records, 225 (73.1%) provided fully verified questionnaire data and were included in the final analysis; the main reason for exclusion of the remaining patients was incomplete questionnaires, with only 7 cases (2.3%) being non-respondents. Mean follow-up duration was 358.90±176.12 days. Postoperatively, 93.3% (210/225) of patients developed CH. The incidence rates of grade I, II, and III CH were 4.0% (9/225), 56.0% (126/225), and 33.3% (75/225), respectively, with no cases of grade IV CH observed. Only 6.7% (15/225) of patients did not develop postoperative CH. General characteristics of the included patients are summarized in Table 2.
Correlation analysis of preoperative parameters with CH severity: The analysis revealed a correlation between a5 BMI (r=0.227, P=0.03) and CH severity, while there was no obvious correlation between the other indicators and the CH degree, as shown in Tables 3,4 and Figure 1.
Table 3. Correlation analysis between preoperative general data and the CH degree.
| Variables | a1 | a2 | a3 | a4 | a5 | a6 | a7 | a8 | a9 | a10 | a11 | a12 | a13 | a14 | a15 | b |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| a1 | 1.000 | 0.096 | −0.836 | −0.655 | −0.251 | 0.142 | 0.207 | −0.185 | −0.077 | 0.131 | 0.094 | 0.012 | −0.123 | 0.204 | −0.073 | 0.061 |
| a2 | 0.096 | 1.000 | −0.156 | 0.126 | 0.293 | −0.222 | −0.246 | −0.718 | 0.191 | 0.145 | 0.114 | −0.040 | 0.144 | 0.842 | −0.206 | 0.174 |
| a3 | −0.836 | −0.156 | 1.000 | 0.658 | 0.142 | −0.167 | −0.227 | 0.191 | 0.008 | −0.178 | −0.177 | −0.034 | 0.100 | −0.252 | 0.036 | −0.121 |
| a4 | −0.655 | 0.126 | 0.658 | 1.000 | 0.820 | −0.135 | −0.213 | 0.008 | 0.062 | 0.025 | −0.104 | −0.093 | 0.130 | 0.021 | 0.109 | 0.122 |
| a5 | −0.251 | 0.293 | 0.142 | 0.820 | 1.000 | −0.078 | −0.115 | −0.148 | 0.057 | 0.156 | −0.034 | −0.087 | 0.091 | 0.228 | 0.131 | 0.227 |
| a6 | 0.142 | −0.222 | −0.167 | −0.135 | −0.078 | 1.000 | 0.636 | 0.023 | −0.111 | −0.082 | −0.021 | −0.023 | −0.130 | −0.071 | 0.141 | 0.045 |
| a7 | 0.207 | −0.246 | −0.227 | −0.213 | −0.115 | 0.636 | 1.000 | 0.094 | −0.070 | 0.098 | 0.021 | 0.079 | 0.086 | −0.150 | 0.272 | 0.090 |
| a8 | −0.185 | −0.718 | 0.191 | 0.008 | −0.148 | 0.023 | 0.094 | 1.000 | −0.186 | −0.030 | 0.088 | 0.139 | −0.023 | −0.630 | 0.146 | −0.221 |
| a9 | −0.077 | 0.191 | 0.008 | 0.062 | 0.057 | −0.111 | −0.070 | −0.186 | 1.000 | 0.061 | 0.019 | −0.024 | 0.140 | 0.117 | −0.089 | −0.077 |
| a10 | 0.131 | 0.145 | −0.178 | 0.025 | 0.156 | −0.082 | 0.098 | −0.030 | 0.061 | 1.000 | −0.006 | 0.167 | 0.061 | 0.108 | −0.211 | 0.052 |
| a11 | 0.094 | 0.114 | −0.177 | −0.104 | −0.034 | −0.021 | 0.021 | 0.088 | 0.019 | −0.006 | 1.000 | 0.314 | −0.189 | 0.242 | −0.090 | −0.005 |
| a12 | 0.012 | −0.040 | −0.034 | −0.093 | −0.087 | −0.023 | 0.079 | 0.139 | −0.024 | 0.167 | 0.314 | 1.000 | −0.207 | 0.078 | −0.138 | 0.040 |
| a13 | −0.123 | 0.144 | 0.100 | 0.130 | 0.091 | −0.130 | 0.086 | −0.023 | 0.140 | 0.061 | −0.189 | −0.207 | 1.000 | −0.357 | 0.058 | 0.201 |
| a14 | 0.204 | 0.842 | −0.252 | 0.021 | 0.228 | −0.071 | −0.150 | −0.630 | 0.117 | 0.108 | 0.242 | 0.078 | −0.357 | 1.000 | −0.233 | 0.094 |
| a15 | −0.073 | −0.206 | 0.036 | 0.109 | 0.131 | 0.141 | 0.272 | 0.146 | −0.089 | −0.211 | −0.090 | −0.138 | 0.058 | −0.233 | 1.000 | 0.059 |
| b | 0.061 | 0.174 | −0.121 | 0.122 | 0.227 | 0.045 | 0.090 | −0.221 | −0.077 | 0.052 | −0.005 | 0.040 | 0.201 | 0.094 | 0.059 | 1.000 |
Definitions for all parameter codes are provided in Appendix 2. CH, compensatory hyperhidrosis.
Table 4. Significance analysis of correlation coefficient between preoperative general data and the CH degree.
| Variables | a1 | a2 | a3 | a4 | a5 | a6 | a7 | a8 | a9 | a10 | a11 | a12 | a13 | a14 | a15 | b |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| a1 | <0.001 | 0.65 | <0.001 | <0.001 | 0.02 | 0.23 | 0.08 | 0.12 | 0.67 | 0.26 | 0.31 | 0.92 | 0.23 | 0.30 | 0.53 | 0.79 |
| a2 | 0.65 | <0.001 | 0.12 | 0.21 | 0.01 | 0.08 | 0.04 | <0.001 | 0.052 | 0.26 | 0.21 | 0.81 | 0.11 | <0.001 | 0.08 | 0.06 |
| a3 | <0.001 | 0.12 | <0.001 | <0.001 | 0.24 | 0.27 | 0.11 | 0.05 | 0.94 | 0.14 | 0.22 | 0.76 | 0.49 | 0.06 | 0.66 | 0.45 |
| a4 | <0.001 | 0.21 | <0.001 | <0.001 | <0.001 | 0.18 | 0.06 | 0.99 | 0.67 | 0.84 | 0.38 | 0.55 | 0.43 | 0.41 | 0.32 | 0.26 |
| a5 | 0.02 | 0.01 | 0.24 | <0.001 | <0.001 | 0.38 | 0.22 | 0.16 | 0.63 | 0.16 | 0.77 | 0.54 | 0.53 | 0.02 | 0.32 | 0.03 |
| a6 | 0.23 | 0.08 | 0.27 | 0.18 | 0.38 | <0.001 | <0.001 | 0.87 | 0.37 | 0.49 | 0.81 | 0.80 | 0.07 | 0.44 | 0.23 | 0.58 |
| a7 | 0.08 | 0.04 | 0.11 | 0.06 | 0.22 | <0.001 | <0.001 | 0.44 | 0.57 | 0.40 | 0.91 | 0.56 | 0.65 | 0.09 | 0.02 | 0.72 |
| a8 | 0.12 | <0.001 | 0.05 | 0.99 | 0.16 | 0.87 | 0.44 | <0.001 | 0.05 | 0.80 | 0.61 | 0.25 | 0.66 | <0.001 | 0.23 | 0.11 |
| a9 | 0.67 | 0.052 | 0.94 | 0.67 | 0.63 | 0.37 | 0.57 | 0.05 | <0.001 | 0.46 | 0.85 | 0.92 | 0.38 | 0.16 | 0.26 | 0.36 |
| a10 | 0.26 | 0.26 | 0.14 | 0.84 | 0.16 | 0.49 | 0.40 | 0.80 | 0.46 | <0.001 | 0.92 | 0.15 | 0.67 | 0.38 | 0.07 | 0.93 |
| a11 | 0.31 | 0.21 | 0.22 | 0.38 | 0.77 | 0.81 | 0.91 | 0.61 | 0.85 | 0.92 | <0.001 | <0.001 | 0.11 | 0.04 | 0.47 | 0.73 |
| a12 | 0.92 | 0.81 | 0.76 | 0.55 | 0.54 | 0.80 | 0.56 | 0.25 | 0.92 | 0.15 | <0.001 | <0.001 | 0.15 | 0.62 | 0.20 | 0.74 |
| a13 | 0.23 | 0.11 | 0.49 | 0.43 | 0.53 | 0.07 | 0.65 | 0.66 | 0.38 | 0.67 | 0.11 | 0.15 | <0.001 | 0.01 | 0.72 | 0.13 |
| a14 | 0.30 | <0.001 | 0.06 | 0.41 | 0.02 | 0.44 | 0.09 | <0.001 | 0.16 | 0.38 | 0.04 | 0.62 | 0.01 | <0.001 | 0.06 | 0.31 |
| a15 | 0.53 | 0.08 | 0.66 | 0.32 | 0.32 | 0.23 | 0.02 | 0.23 | 0.26 | 0.07 | 0.47 | 0.20 | 0.72 | 0.06 | <0.001 | 0.48 |
Definitions for all parameter codes are provided in Appendix 2. CH, compensatory hyperhidrosis.
Figure 1.
Visualization of correlation between preoperative general data and the CH degree. Definitions for all parameter codes are provided in Appendix 2. CH, compensatory hyperhidrosis.
Correlation analysis of intraoperative parameters with CH severity: Further analysis demonstrated correlations between CH severity and three intraoperative variables: sympathetic resection level (left side; a16; r=−0.187, P=0.03), sympathetic resection level (right side; a17; r=−0.451, P=0.03), and postoperative temperature difference between the right hand and pre-anesthesia baseline (a31; r=0.2149, P=0.04). No significant correlations were observed between other intraoperative parameters and CH severity (Tables 5,6 and Figure 2).
Table 5. Correlation analysis between intraoperative data and the CH degree.
| Variables | a16 | a17 | a18 | a19 | a20 | a21 | a22 | a23 | a24 | a25 | a26 | a27 | a28 | a29 | a30 | a31 | a32 | a33 | a34 | a35 | a36 | a37 | b |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| a16 | 1.000 | 1.000 | −0.141 | −0.100 | −0.081 | −0.179 | 0.029 | −0.046 | −0.094 | −0.085 | −0.035 | −0.079 | 0.058 | 0.012 | 0.119 | 0.057 | 0.086 | 0.136 | −0.027 | −0.042 | 0.115 | 0.092 | −0.187 |
| a17 | 1.000 | 1.000 | −0.141 | −0.100 | −0.081 | −0.179 | 0.029 | −0.046 | −0.094 | −0.085 | −0.035 | −0.079 | 0.058 | 0.012 | 0.119 | 0.057 | 0.086 | 0.136 | −0.027 | −0.042 | 0.115 | 0.092 | −0.187 |
| a18 | −0.141 | −0.141 | 1.000 | 0.634 | 0.855 | 0.612 | 0.308 | 0.052 | 0.501 | 0.187 | 0.307 | 0.069 | −0.088 | −0.135 | −0.855 | −0.560 | −0.748 | −0.584 | −0.069 | −0.020 | −0.403 | −0.249 | 0.045 |
| a19 | −0.100 | −0.100 | 0.634 | 1.000 | 0.642 | 0.911 | 0.163 | 0.151 | 0.363 | 0.240 | 0.150 | −0.097 | 0.151 | −0.008 | −0.579 | −0.916 | −0.477 | −0.781 | −0.070 | −0.160 | −0.128 | −0.256 | −0.197 |
| a20 | −0.081 | −0.081 | 0.855 | 0.642 | 1.000 | 0.616 | 0.117 | 0.233 | 0.236 | 0.330 | 0.133 | −0.105 | −0.045 | −0.178 | −0.794 | −0.622 | −0.878 | −0.645 | 0.004 | −0.202 | −0.149 | −0.347 | −0.032 |
| a21 | −0.179 | −0.179 | 0.612 | 0.911 | 0.616 | 1.000 | 0.083 | 0.135 | 0.345 | 0.265 | 0.023 | −0.179 | 0.180 | −0.005 | −0.607 | −0.853 | −0.450 | −0.848 | −0.056 | −0.180 | −0.099 | −0.280 | −0.175 |
| a22 | 0.029 | 0.029 | 0.308 | 0.163 | 0.117 | 0.083 | 1.000 | −0.386 | 0.752 | −0.324 | 0.452 | 0.380 | 0.092 | 0.188 | −0.063 | −0.035 | −0.046 | 0.083 | −0.669 | 0.403 | −0.515 | 0.384 | 0.167 |
| a23 | −0.046 | −0.046 | 0.052 | 0.151 | 0.233 | 0.135 | −0.386 | 1.000 | −0.174 | 0.868 | 0.039 | −0.315 | 0.229 | −0.095 | 0.059 | −0.237 | −0.073 | −0.215 | 0.597 | −0.808 | 0.406 | −0.701 | −0.159 |
| a24 | −0.094 | −0.094 | 0.501 | 0.363 | 0.236 | 0.345 | 0.752 | −0.174 | 1.000 | −0.017 | 0.447 | 0.188 | 0.137 | 0.018 | −0.233 | −0.290 | −0.077 | −0.275 | −0.437 | 0.169 | −0.619 | 0.016 | 0.033 |
| a25 | −0.085 | −0.085 | 0.187 | 0.240 | 0.330 | 0.265 | −0.324 | 0.868 | −0.017 | 1.000 | 0.064 | −0.286 | 0.328 | −0.189 | −0.068 | −0.307 | −0.104 | −0.355 | 0.540 | −0.698 | 0.341 | −0.867 | −0.100 |
| a26 | −0.035 | −0.035 | 0.307 | 0.150 | 0.133 | 0.023 | 0.452 | 0.039 | 0.447 | 0.064 | 1.000 | 0.347 | 0.279 | 0.066 | 0.117 | −0.008 | 0.037 | 0.092 | 0.244 | 0.106 | −0.122 | −0.008 | 0.034 |
| a27 | −0.079 | −0.079 | 0.069 | −0.097 | −0.105 | −0.179 | 0.380 | −0.315 | 0.188 | −0.286 | 0.347 | 1.000 | 0.035 | 0.644 | −0.030 | 0.439 | 0.096 | 0.507 | −0.253 | 0.771 | −0.183 | 0.542 | 0.192 |
| a28 | 0.058 | 0.058 | −0.088 | 0.151 | −0.045 | 0.180 | 0.092 | 0.229 | 0.137 | 0.328 | 0.279 | 0.035 | 1.000 | 0.228 | 0.244 | −0.101 | 0.435 | −0.071 | 0.303 | −0.069 | 0.610 | −0.182 | −0.058 |
| a29 | 0.012 | 0.012 | −0.135 | −0.008 | −0.178 | −0.005 | 0.188 | −0.095 | 0.018 | −0.189 | 0.066 | 0.644 | 0.228 | 1.000 | 0.051 | 0.307 | 0.184 | 0.461 | −0.135 | 0.458 | 0.030 | 0.605 | 0.032 |
| a30 | 0.119 | 0.119 | −0.855 | −0.579 | −0.794 | −0.607 | −0.063 | 0.059 | −0.233 | −0.068 | 0.117 | −0.030 | 0.244 | 0.051 | 1.000 | 0.511 | 0.807 | 0.577 | 0.197 | −0.063 | 0.352 | 0.137 | −0.010 |
| a31 | 0.057 | 0.057 | −0.560 | −0.916 | −0.622 | −0.853 | −0.035 | −0.237 | −0.290 | −0.307 | −0.008 | 0.439 | −0.101 | 0.307 | 0.511 | 1.000 | 0.480 | 0.900 | 0.009 | 0.406 | 0.088 | 0.446 | 0.215 |
| a32 | 0.086 | 0.086 | −0.748 | −0.477 | −0.878 | −0.450 | −0.046 | −0.073 | −0.077 | −0.104 | 0.037 | 0.096 | 0.435 | 0.184 | 0.807 | 0.480 | 1.000 | 0.486 | 0.127 | 0.111 | 0.366 | 0.160 | 0.019 |
| a33 | 0.136 | 0.136 | −0.584 | −0.781 | −0.645 | −0.848 | 0.083 | −0.215 | −0.275 | −0.355 | 0.092 | 0.507 | −0.071 | 0.461 | 0.577 | 0.900 | 0.486 | 1.000 | −0.031 | 0.419 | 0.073 | 0.548 | 0.176 |
| a34 | −0.027 | −0.027 | −0.069 | −0.070 | 0.004 | −0.056 | −0.669 | 0.597 | −0.437 | 0.540 | 0.244 | −0.253 | 0.303 | −0.135 | 0.197 | 0.009 | 0.127 | −0.031 | 1.000 | −0.499 | 0.587 | −0.520 | −0.145 |
| a35 | −0.042 | −0.042 | −0.020 | −0.160 | −0.202 | −0.180 | 0.403 | −0.808 | 0.169 | −0.698 | 0.106 | 0.771 | −0.069 | 0.458 | −0.063 | 0.406 | 0.111 | 0.419 | −0.499 | 1.000 | −0.279 | 0.750 | 0.183 |
| a36 | 0.115 | 0.115 | −0.403 | −0.128 | −0.149 | −0.099 | −0.515 | 0.406 | −0.619 | 0.341 | −0.122 | −0.183 | 0.610 | 0.030 | 0.352 | 0.088 | 0.366 | 0.073 | 0.587 | −0.279 | 1.000 | −0.257 | −0.109 |
| a37 | 0.092 | 0.092 | −0.249 | −0.256 | −0.347 | −0.280 | 0.384 | −0.701 | 0.016 | −0.867 | −0.008 | 0.542 | −0.182 | 0.605 | 0.137 | 0.446 | 0.160 | 0.548 | −0.520 | 0.750 | −0.257 | 1.000 | 0.083 |
| b | −0.187 | −0.187 | 0.045 | −0.197 | −0.032 | −0.175 | 0.167 | −0.159 | 0.033 | −0.100 | 0.034 | 0.192 | −0.058 | 0.032 | −0.010 | 0.215 | 0.019 | 0.176 | −0.145 | 0.183 | −0.109 | 0.083 | 1.000 |
Definitions for all parameter codes are provided in Appendix 2. CH, compensatory hyperhidrosis.
Table 6. Significance analysis of correlation coefficient between intraoperative data and the CH degree.
| Variables | a16 | a17 | a18 | a19 | a20 | a21 | a22 | a23 | a24 | a25 | a26 | a27 | a28 | a29 | a30 | a31 | a32 | a33 | a34 | a35 | a36 | a37 | b |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| a16 | <0.001 | <0.001 | 0.32 | 0.57 | 0.72 | 0.26 | 0.83 | 0.84 | 0.48 | 0.49 | 0.73 | 0.49 | 0.64 | 0.98 | 0.37 | 0.75 | 0.61 | 0.31 | 0.80 | 0.88 | 0.34 | 0.60 | 0.03 |
| a17 | <0.001 | <0.001 | 0.32 | 0.57 | 0.72 | 0.26 | 0.83 | 0.84 | 0.48 | 0.49 | 0.73 | 0.49 | 0.64 | 0.98 | 0.37 | 0.75 | 0.61 | 0.31 | 0.80 | 0.88 | 0.34 | 0.60 | 0.03 |
| a18 | 0.32 | 0.32 | <0.001 | <0.001 | <0.001 | <0.001 | 0.84 | 0.34 | <0.001 | 0.45 | <0.001 | 0.19 | 0.72 | 0.69 | <0.001 | <0.001 | <0.001 | <0.001 | 0.65 | 0.20 | <0.001 | 0.43 | 0.89 |
| a19 | 0.57 | 0.57 | <0.001 | <0.001 | <0.001 | <0.001 | 0.31 | 0.88 | 0.02 | 0.24 | 0.13 | 0.14 | 0.16 | 0.21 | <0.001 | <0.001 | <0.001 | <0.001 | 0.25 | 0.46 | 0.41 | 0.12 | 0.08 |
| a20 | 0.72 | 0.72 | <0.001 | <0.001 | <0.001 | <0.001 | 0.69 | 0.03 | 0.15 | 0.01 | 0.09 | 0.26 | 0.61 | 0.17 | <0.001 | <0.001 | <0.001 | <0.001 | 0.59 | 0.04 | 0.11 | 0.01 | 0.83 |
| a21 | 0.26 | 0.26 | <0.001 | <0.001 | <0.001 | <0.001 | 0.057 | 0.16 | 0.01 | <0.001 | 0.81 | 0.04 | 0.03 | 0.49 | <0.001 | <0.001 | <0.001 | <0.001 | 0.053 | 0.051 | 0.73 | 0.01 | 0.07 |
| a22 | 0.83 | 0.83 | 0.84 | 0.31 | 0.69 | 0.057 | <0.001 | 0.10 | 0.27 | 0.03 | 0.23 | 0.15 | 0.63 | 0.98 | 0.46 | 0.18 | 0.85 | 0.08 | <0.001 | 0.07 | 0.19 | 0.10 | 0.11 |
| a23 | 0.84 | 0.84 | 0.34 | 0.88 | 0.03 | 0.16 | 0.10 | <0.001 | <0.001 | <0.001 | 0.77 | <0.001 | 0.07 | 0.19 | 0.25 | 0.24 | 0.21 | 0.08 | 0.09 | <0.001 | <0.001 | <0.001 | 0.19 |
| a24 | 0.48 | 0.48 | <0.001 | 0.02 | 0.15 | 0.01 | 0.27 | <0.001 | <0.001 | 0.19 | <0.001 | 0.048 | 0.052 | 0.63 | 0.02 | 0.12 | 0.56 | 0.04 | 0.44 | <0.001 | <0.001 | 0.21 | 0.58 |
| a25 | 0.49 | 0.49 | 0.45 | 0.24 | 0.01 | <0.001 | 0.03 | <0.001 | 0.19 | <0.001 | 0.37 | <0.001 | <0.001 | 0.13 | 0.68 | 0.058 | 0.29 | <0.001 | 0.03 | <0.001 | <0.001 | <0.001 | 0.41 |
| a26 | 0.73 | 0.73 | <0.001 | 0.13 | 0.09 | 0.81 | 0.23 | 0.77 | <0.001 | 0.37 | <0.001 | 0.08 | <0.001 | 0.67 | 0.50 | 0.38 | 0.98 | 0.95 | 0.63 | 0.59 | 0.88 | 0.64 | 0.76 |
| a27 | 0.49 | 0.49 | 0.19 | 0.14 | 0.26 | 0.04 | 0.15 | <0.001 | 0.048 | <0.001 | 0.08 | <0.001 | 0.99 | <0.001 | 0.52 | <0.001 | 0.30 | <0.001 | 0.19 | <0.001 | 0.10 | <0.001 | 0.14 |
| a28 | 0.64 | 0.64 | 0.72 | 0.16 | 0.61 | 0.03 | 0.63 | 0.07 | 0.052 | <0.001 | <0.001 | 0.99 | <0.001 | 0.26 | 0.03 | 0.20 | <0.001 | 0.10 | 0.41 | 0.19 | <0.001 | 0.06 | 0.47 |
| a29 | 0.98 | 0.98 | 0.69 | 0.21 | 0.17 | 0.49 | 0.98 | 0.19 | 0.63 | 0.13 | 0.67 | <0.001 | 0.26 | <0.001 | 0.55 | <0.001 | 0.10 | <0.001 | 0.99 | <0.001 | 0.65 | <0.001 | 0.89 |
| a30 | 0.37 | 0.37 | <0.001 | <0.001 | <0.001 | <0.001 | 0.46 | 0.25 | 0.02 | 0.68 | 0.50 | 0.52 | 0.03 | 0.55 | <0.001 | <0.001 | <0.001 | <0.001 | 0.50 | 0.26 | <0.001 | 0.53 | 0.99 |
| a31 | 0.75 | 0.75 | <0.001 | <0.001 | <0.001 | <0.001 | 0.18 | 0.24 | 0.12 | 0.058 | 0.38 | <0.001 | 0.20 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | 0.15 | 0.01 | 0.76 | <0.001 | 0.04 |
| a32 | 0.61 | 0.61 | <0.001 | <0.001 | <0.001 | <0.001 | 0.85 | 0.21 | 0.56 | 0.29 | 0.98 | 0.30 | <0.001 | 0.10 | <0.001 | <0.001 | <0.001 | <0.001 | 0.85 | 0.17 | <0.001 | 0.09 | 0.94 |
| a33 | 0.31 | 0.31 | <0.001 | <0.001 | <0.001 | <0.001 | 0.08 | 0.08 | 0.04 | <0.001 | 0.95 | <0.001 | 0.10 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | 0.08 | <0.001 | 0.63 | <0.001 | 0.11 |
| a34 | 0.80 | 0.80 | 0.65 | 0.25 | 0.59 | 0.053 | <0.001 | 0.09 | 0.44 | 0.03 | 0.63 | 0.19 | 0.41 | 0.99 | 0.50 | 0.15 | 0.85 | 0.08 | <0.001 | 0.07 | 0.20 | 0.09 | 0.11 |
| a35 | 0.88 | 0.88 | 0.20 | 0.46 | 0.04 | 0.051 | 0.07 | <0.001 | <0.001 | <0.001 | 0.59 | <0.001 | 0.19 | <0.001 | 0.26 | 0.01 | 0.17 | <0.001 | 0.07 | <0.001 | <0.001 | <0.001 | 0.11 |
| a36 | 0.34 | 0.34 | <0.001 | 0.41 | 0.11 | 0.73 | 0.19 | <0.001 | <0.001 | <0.001 | 0.88 | 0.10 | <0.001 | 0.65 | <0.001 | 0.76 | <0.001 | 0.63 | 0.20 | <0.001 | <0.001 | 0.01 | 0.30 |
| a37 | 0.60 | 0.60 | 0.43 | 0.12 | 0.01 | 0.01 | 0.10 | <0.001 | 0.21 | <0.001 | 0.64 | <0.001 | 0.06 | <0.001 | 0.53 | <0.001 | 0.09 | <0.001 | 0.09 | <0.001 | 0.01 | <0.001 | 0.58 |
Definitions for all parameter codes are provided in Appendix 2. CH, compensatory hyperhidrosis.
Figure 2.
Visualization of correlation between postoperative data and the CH degree. Definitions for all parameter codes are provided in Appendix 2. CH, compensatory hyperhidrosis.
Correlation analysis of postoperative outcomes with CH severity: A significant negative correlation was observed between overall surgical satisfaction (a44; r=−0.363, P=0.001) and CH severity. Suggesting higher satisfaction was associated with milder CH. Other postoperative indicators, including surgical efficacy (for hands, feet, axillae, and face), incision satisfaction, palmar hyperhidrosis recurrence, postoperative improvement in quality of life/work, showed negligible correlations with CH severity (Tables 7,8 and Figure 3).
Table 7. Correlation analysis between postoperative data and the CH degree.
| Variables | a38 | a39 | a40 | a41 | a42 | a43 | a44 | b |
|---|---|---|---|---|---|---|---|---|
| a38 | 1.000 | 0.744 | 0.127 | −0.136 | −0.339 | 0.022 | −0.208 | −0.127 |
| a39 | 0.744 | 1.000 | 0.018 | −0.191 | −0.339 | 0.022 | −0.098 | −0.137 |
| a40 | 0.127 | 0.018 | 1.000 | 0.099 | 0.050 | −0.035 | 0.064 | −0.053 |
| a41 | −0.136 | −0.191 | 0.099 | 1.000 | 0.136 | −0.049 | 0.432 | −0.229 |
| a42 | −0.339 | −0.339 | 0.050 | 0.136 | 1.000 | −0.433 | 0.260 | −0.031 |
| a43 | 0.022 | 0.022 | −0.035 | −0.049 | −0.433 | 1.000 | −0.220 | −0.036 |
| a44 | −0.208 | −0.098 | 0.064 | 0.432 | 0.260 | −0.220 | 1.000 | −0.363 |
| b | −0.127 | −0.137 | −0.053 | −0.229 | −0.031 | −0.036 | −0.363 | 1.000 |
Definitions for all parameter codes are provided in Appendix 2. CH, compensatory hyperhidrosis.
Table 8. Significance analysis of correlation coefficient between postoperative data and the CH degree.
| Variables | a38 | a39 | a40 | a41 | a42 | a43 | a44 | b |
|---|---|---|---|---|---|---|---|---|
| a38 | <0.001 | <0.001 | 0.30 | 0.46 | 0.001 | 0.51 | 0.03 | 0.27 |
| a39 | <0.001 | <0.001 | 0.86 | 0.29 | 0.001 | 0.51 | 0.21 | 0.12 |
| a40 | 0.30 | 0.86 | <0.001 | 0.47 | 0.71 | 0.79 | 0.28 | 0.50 |
| a41 | 0.46 | 0.29 | 0.47 | <0.001 | 0.37 | 0.53 | 0.00 | 0.14 |
| a42 | 0.001 | 0.001 | 0.71 | 0.37 | <0.001 | <0.001 | 0.00 | 0.70 |
| a43 | 0.51 | 0.51 | 0.79 | 0.53 | <0.001 | <0.001 | 0.02 | 0.85 |
| a44 | 0.03 | 0.21 | 0.28 | <0.001 | <0.001 | 0.02 | <0.001 | <0.001 |
Definitions for all parameter codes are provided in Appendix 2. CH, compensatory hyperhidrosis.
Figure 3.
Visualization of correlation between intraoperative data and the CH degree. Definitions for all parameter codes are provided in Appendix 2. CH, compensatory hyperhidrosis.
Regression analysis incorporating pre- and intraoperative variables significantly correlated with CH demonstrated that only the sympathetic trunk resection level (left side; a16) reached statistical significance (coefficient =−1.195, P=0.02), indicating that resection of the left R4 sympathetic ganglion reduced CH incidence. The remaining three variables (a5, a17, a31) showed no significant associations (P>0.05), and no valid regression equation could be established (the constructed multivariate regression model has limited explanatory power, adjusted R2=0.2423, and the effect stability of the core variables is insufficient). Detailed results are provided in Table 9.
Table 9. Regression analysis of variables with significant correlation and CH.
| Code | Variables | Estimate | Standard error | t value | Pr(>|t|) | Var inflation |
|---|---|---|---|---|---|---|
| Intercept | 11.925 | 6.967 | 1.712 | 0.091 | ||
| a5 | BMI | 0.050 | 0.036 | 1.377 | 0.173 | 1.07 |
| a16 | Level of sympathicotomy (left side) | −1.195 | 0.503 | −2.375 | 0.020 | 1.04 |
| a17 | Level of sympathicotomy (left side) | −0.205 | 0.181 | −1.132 | 0.262 | 2.44 |
| a31 | Right hand temperature (postoperative − before anesthesia) | 0.083 | 0.120 | 0.690 | 0.492 | 2.33 |
Definitions for all parameter codes are provided in Appendix 2. R2: 0.2935, adjust R2: 0.2423. BMI, body mass index; CH, compensatory hyperhidrosis.
Rough set analysis results: The entropy method was used for attribute reduction, and four core attributes were obtained, including age (a2), BMI (a5), initial onset age (a13) and postoperative right-hand temperature (a27). BMI was categorized into four tiers: >22, 19.5–22, 18.5–19.4, and <18.5 kg/m2. From this stratification, four predictive rules for grade III CH were derived: (I) BMI >22 kg/m2 combined with an initial onset age of PH >11 years; (II) BMI =19.5–22 kg/m2 paired with a surgical age >28.5 years; (III) BMI =18.5–19.5 kg/m2 alongside a postoperative right-hand temperature >36.6 ℃; (IV) BMI <18.5 kg/m2 with a postoperative right-hand temperature <37.0 ℃ and an initial PH onset age <10 years. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy of these four rules in predicting grade III CH are summarized in Table 10. When all four rules were applied in combination, the predictive model demonstrated a sensitivity of 66%, specificity of 87%, PPV of 76%, NPV of 80%, and overall accuracy of 79%.
Table 10. The rules for predicting grade III CH obtained through rough set analysis.
| Rules for predicting grade III CH | Predicted CH level | Real CH level | Sensitivity (95% CI) |
Specificity (95% CI) |
PPV (95% CI) | NPV (95% CI) | Accuracy rate (95% CI) | |
|---|---|---|---|---|---|---|---|---|
| Grade III | < Grade III | |||||||
| Rule (I): BMI >22 kg/m2 + initial onset age of PH >11 years | Grade III | 6 | 1 | 0.86 (0.42–0.99) |
0.83 (0.44–0.98) |
0.75 (0.35–0.95) |
0.91 (0.59–0.99) |
0.84 (0.60–0.96) |
| < Grade III | 2 | 10 | ||||||
| Rule (II): BMI 19.5–22 kg/m2 + surgical age >28.5 years | Grade III | 7 | 1 | 0.88 (0.47–0.99) |
0.79 (0.47–0.95) |
0.70 (0.35–0.91) |
0.92 (0.62–0.99) |
0.82 (0.60–0.94) |
| < Grade III | 3 | 11 | ||||||
| Rule (III): BMI 18.5–19.4 kg/m2 + postoperative right-hand temperature >36.6 ℃ | Grade III | 2 | 3 | 0.40 (0.09–0.81) |
1.00 (0.63–1.00) |
1.00 (0.29–1.00) |
0.73 (0.43–0.91) |
0.77 (0.50–0.93) |
| < Grade III | 0 | 8 | ||||||
| Rule (IV): BMI <18.5 kg/m2 + postoperative right-hand temperature <37.0 ℃ + initial PH onset age <10 years | Grade III | 4 | 5 | 0.44 (0.14–0.79) |
0.92 (0.62–0.99) |
0.80 (0.28–0.99) |
0.69 (0.46–0.86) |
0.71 (0.48–0.88) |
| < Grade III | 1 | 11 | ||||||
| All rules | Grade III | 19 | 10 | 0.66 (0.48–0.81) |
0.87 (0.74–0.95) |
0.76 (0.56–0.89) |
0.80 (0.67–0.89) |
0.79 (0.69–0.87) |
| < Grade III | 6 | 40 | ||||||
< Grade III: grade 0/I/II. BMI, body mass index; CH, compensatory hyperhidrosis; CI, confidence interval; NPV, negative predictive value; PH, primary hyperhidrosis; PPV, positive predictive value.
Recurrence of PH was observed in 0.89% (2/225) of patients, both presenting with bilateral recurrence. The two cases involved preoperative severe PH persisting postoperatively at identical severity levels, with recurrence timing at 1 week and 1 month, respectively. Both patients developed grade II CH. Notably, one patient who underwent initial bilateral R3 ganglion resection via ETS experienced recurrence; complete resolution was achieved following secondary T2 resection.
In this study, 86.7% (195/225) of patients demonstrated complete resolution of PH, while 13.3% (30/225) achieved partial remission. Among concurrent hyperhidrosis conditions, plantar hyperhidrosis was present in 92.0% (207/225), axillary hyperhidrosis in 41.3% (93/225), and facial hyperhidrosis in 1.3% (3/225), with detailed in Table 2. Regarding quality-of-life impact, 80.0% (180/225) reported marked improvement in daily/work activities, 18.7% (42/225) noted partial improvement, and 1.3% (3/225) perceived no benefit (the latter being one of the recurrent PH cases). The mean overall satisfaction score was 88.01%±12.91%, showing significant correlation with CH severity. Comparative satisfaction rates across CH grades are presented in Table 11.
Table 11. Overall satisfaction with surgery for patients with different degrees of CH.
| Indicators | No CH | Grade I | Grade II | Grade III | Grade IV | F value | P value |
|---|---|---|---|---|---|---|---|
| Ratio | 6.7 (15/225) | 4.0 (9/225) | 56.0 (126/225) | 33.3 (75/225) | 0.00 | − | − |
| Overall satisfaction with surgery | 97.40±5.27 | 92.00±3.86 | 90.57±10.42 | 81.36±15.15 | 0.00 | 4.232 | 0.008 |
Data are presented as mean ± standard deviation or % (n/n). CH, compensatory hyperhidrosis.
Discussion
As the most common and severe complication after ETS, severe CH (grade III or above) significantly reduces patients’ satisfaction with ETS (18). In this study, the postoperative satisfaction score of grade III CH patients was only 81.36±15.15, which was significantly lower than the score above 90 points in grade I/II patients. Although current studies suggest that expanding the scope of sympathectomy (T5–8 or T5–12) (19-22) or performing sympathetic nerve reconstruction (23) may improve CH symptoms, these methods increase surgical trauma and may increase the risk of arrhythmia (24). Therefore, preoperative/intraoperative prediction of CH risk remains a core issue for clinical decision-making optimization. If severe CH can be predicted preoperatively, both patients and surgeons can make more cautious choices. If intraoperative prediction of severe CH is feasible immediately after sympathetic chain transection, T5–12 ablation can be simultaneously performed during the initial ETS surgery to avoid secondary surgery.
Existing CH prediction methods have limitations. Lee et al. (25) proposed thoracoscopic local anesthetic blockade of the sympathetic trunk for CH risk assessment, showing 94.4% specificity but only 33.3% sensitivity, with the need for two invasive surgeries limiting clinical application. Kara et al. (26) found that postoperative CH was related to age >21 years, BMI >22 kg/m2, and smoking history, but did not establish a predictive model. Jeong et al. (27) attempted to use heart rate variability to predict CH, while Hyun et al. (28) applied machine learning methods. Although these studies demonstrate preliminary feasibility, their clinical value is limited for CH with 90% incidence. Predicting CH severity (especially grade ≥ III) rather than simple occurrence has greater clinical significance. Current methods still face challenges in prediction accuracy, operational invasiveness, or clinical practicality.
Compared with previous studies, this study systematically evaluated 37 predictive indicators for the first time, including previously overlooked factors: climate zone, occupational characteristics, psychological characteristics, and perioperative temperature changes (pre-anesthesia baseline vs. postoperative palm/axillary temperature). However, traditional correlation analysis and regression analysis failed to obtain satisfactory predictive models. Further analysis showed that the overall incidence of CH in this cohort reached 93.3% (210/225), with grade III CH accounting for 33.3% (75/225). Using a 100-point satisfaction scale (100 = completely satisfied), grade I and II CH patients scored 92.00±3.86 and 90.57±10.42 respectively, while grade III patients scored 81.36±15.15. In the questionnaire survey, the main reason for patient dissatisfaction with the surgery was also grade III CH. Therefore, we believe that the focus of predicting CH is to identify patients with grade III and above CH, that is, to predict whether grade III and above CH will occur after ETS.
With the advice of statistical experts, this study pioneered the application of rough set analysis in hyperhidrosis research. Rough set analysis is usually suitable for exploring fuzzy and uncertain sets and does not require any prior information other than the data set. It can delete redundant information, analyze the roughness of knowledge, the dependence and importance between attributes, and generate classification or decision rules. This method breaks through the linear assumption of traditional statistics and is particularly suitable for handling the highly non-linear and multi-factor interactive clinical data in this study.
In this study, through rough set analysis, four core attributes were obtained: age, BMI, initial onset age of PH, and right-hand temperature after surgery. Then, based on BMI stratification, four rules for predicting grade III CH were obtained (see Table 10). The sensitivity, specificity, and accuracy of Rules (I) and (II) are relatively good, basically above 80%, and they have the potential to be used as a preoperative risk assessment tool. The sensitivity of Rules 3 and 4 is poor, only about 40%, but the PPVs reach 100% and 80% respectively, indicating that if a patient is predicted to have grade III CH after surgery, the probability of actually developing grade III CH after surgery is very high, which has clear early-warning value. In addition, the specificities of Rules 3 and 4 are also good, 100% and 92% respectively, which can effectively screen out patients with a low probability of developing grade III CH after surgery, that is, low-risk patients. In general, the four rules for predicting grade III CH obtained through rough set analysis have certain clinical application value.
In this study, a “strange” phenomenon was also found. In the correlation analysis, the temperature difference between the right hand after surgery and before anesthesia was significantly correlated with CH, while the temperature difference of the left hand was not significantly correlated. In the rough set analysis, a right-hand temperature >36.6 ℃ after surgery was also shown to be one of the core attributes for predicting grade III CH, while the left-hand temperature was not relevant. This phenomenon has not been reported in previous studies. We consider that this may be related to the fact that the vast majority of patients in this study were right-handed (99.2%, 253/255), and the activity level of the right upper limb was higher than that of the left upper limb. However, further verification is still needed.
In addition, due to the “pursuit of dry hands”, our unit is more accustomed to performing R3 sympathetic nerve chain resection. In this study, 97.3% (219/225) of the patients underwent R3 sympathetic nerve chain resection. This may also lead to a higher proportion of grade III CH (33.3%) in this study compared to other reported studies. This is also the reason why the rough set analysis results did not include the level of sympathetic nerve chain resection as a core attribute. Therefore, more precisely, the four rules for predicting grade III CH obtained in this study are more suitable for predicting patients who are to undergo R3 sympathetic nerve chain resection. In our unit, there has been no grade IV CH, so rules for predicting grade IV CH could not be obtained, and a multi-center study is needed to make up for this deficiency.
It should be particularly noted that the principle of rough set analysis determines that the accuracy of the rules that it analyzes increases with the increase in sample size. The sample size of this study is relatively small, which may have a certain impact on the accuracy of the analysis results, such as causing over-fitting of the rules. Additionally, the cut-off points for continuous variables in this study are data-driven, which may pose an over-fitting risk and limit the applicability of the derived rules in external populations; future multi-center large-sample studies are required for external validation to optimize these thresholds and enhance generalizability. Furthermore, the performance metrics of the rules were evaluated solely on the derivation dataset without internal validation (e.g., cross-validation), potentially leading to optimism bias. In addition, the accuracy of our analysis was limited by the relatively small proportion of patients who underwent R4 sympathectomy, the absence of grade IV CH cases, and the inherent subjectivity in assessing CH severity through patient self-reporting. In the future, multi-center and large-sample studies are needed to obtain more reliable results. These rules need to be verified through prospective, multi-center, large-sample cohort studies, and exploration of including predictive factors such as genetic (29) and metabolic markers (30) in the statistical analysis.
Conclusions
This study, for the first time through rough set analysis, develops exploratory prediction rules for grade III CH that integrate BMI stratification with surgical age, initial PH onset age, and postoperative right-hand temperature metrics. These rules offer a preliminary reference for preoperative risk assessment in patients undergoing ETS: Rules (I) and (II) may assist in identifying high-risk populations to enhance shared decision-making through risk visualization, while Rules (III) and (IV) could provide targeted warnings for specific subgroups and screen for low-risk patients for grade III CH. Notably, these rules are exploratory in nature, and their non-invasive and low-cost characteristics only initially indicate potential clinical relevance. Further prospective, multicentre validation is indispensable before their formal clinical implementation.
Supplementary
The article’s supplementary files as
Acknowledgments
None.
Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the Ethics Committee of the Xiamen Humanity Hospital, Fujian Medical University (No. HAXM-MEC-20220519-X002-01) and individual consent for this retrospective analysis was waived.
Footnotes
Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-aw-2074/rc
Funding: None.
Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-aw-2074/coif). The authors have no conflicts of interest to declare.
Data Sharing Statement
Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-aw-2074/dss
References
- 1.Tu Y, Liu Y. Clinical guideline for minimally invasive treatment of palmar hyperhidrosis in China (2021). Chinese Clinical Journal of Thoracic and Cardiovascular Surgery 2021;28:1133-9. [Google Scholar]
- 2.Gregoriou S, Sidiropoulou P, Kontochristopoulos G, et al. Management Strategies Of Palmar Hyperhidrosis: Challenges And Solutions. Clin Cosmet Investig Dermatol 2019;12:733-44. 10.2147/CCID.S210973 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Lin C, Wang D, Yan Y, et al. Transnasal humidified rapid-insufflation ventilator exchange compared with laryngeal mask airway for endoscopic thoracic sympathectomy: a randomized controlled trial. Front Med (Lausanne) 2023;10:1252586. 10.3389/fmed.2023.1252586 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Zhang W, Lin J, Zhao H, et al. Evaluation of the clinical efficacy of ultra-fast track anesthesia for endoscopic thoracic sympathectomy of palmar hyperhidrosis. J Cosmet Dermatol 2024;23:3327-34. 10.1111/jocd.16425 [DOI] [PubMed] [Google Scholar]
- 5.Lin JB, Kang MQ, Chen JF, et al. Transareolar single-port endoscopic thoracic sympathectomy with a flexible endoscope for primary palmar hyperhidrosis: a prospective randomized controlled trial. Ann Transl Med 2020;8:1659. 10.21037/atm-20-7399 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kuijpers M, van Zanden JE, Harms PW, et al. Minimally Invasive Sympathicotomy for Palmar Hyperhidrosis and Facial Blushing: Current Status and the Hyperhidrosis Expert Center Approach. J Clin Med 2022;11:786. 10.3390/jcm11030786 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Wu JY, Xiong JK, Huang CB, et al. Compensatory hyperhidrosis following endoscopic thoracic sympathectomy: a 5-year follow-up study of risk factors and symptom progression. J Cardiothorac Surg 2025;21:22. 10.1186/s13019-025-03720-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Martínez-Hernández NJ, Estors-Guerrero M, Galbis-Caravajal JM, et al. Long-term outcomes and predictors of compensatory sweating after bilateral endoscopic thoracic sympathectomy. Eur J Cardiothorac Surg 2025;67:ezaf108. 10.1093/ejcts/ezaf108 [DOI] [PubMed] [Google Scholar]
- 9.Nachira D, Vita ML, Napolitano AG, et al. Predictors of Compensatory Sweating and Satisfaction Following Endoscopic Thoracic Sympathetic Chain Clipping for Palmar/Axillary Hyperhidrosis. J Clin Med 2025;14:326. 10.3390/jcm14020326 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Cramer MN, Jay O. Compensatory hyperhidrosis following thoracic sympathectomy: a biophysical rationale. Am J Physiol Regul Integr Comp Physiol 2012;302:R352-6. 10.1152/ajpregu.00419.2011 [DOI] [PubMed] [Google Scholar]
- 11.Zhang W, Yu D, Wei Y, et al. A systematic review and meta-analysis of T2, T3 or T4, to evaluate the best denervation level for palmar hyperhidrosis. Sci Rep 2017;7:129. 10.1038/s41598-017-00169-w [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Dogru MV, Sezen CB, Girgin O, et al. Is there any relationship between quality of life and the level of sympathectomy in primary palmar hyperhidrosis? Single-center experience. Gen Thorac Cardiovasc Surg 2020;68:273-9. 10.1007/s11748-019-01210-7 [DOI] [PubMed] [Google Scholar]
- 13.Sang HW, Li GL, Xiong P, et al. Optimal targeting of sympathetic chain levels for treatment of palmar hyperhidrosis: an updated systematic review. Surg Endosc 2017;31:4357-69. 10.1007/s00464-017-5508-y [DOI] [PubMed] [Google Scholar]
- 14.Soares TJ, Dias PG, Sampaio SM. Impact of Video-Assisted Thoracoscopic Sympathectomy and Related Complications on Quality of Life According to the Level of Sympathectomy. Ann Vasc Surg 2020;63:63-67.e1. 10.1016/j.avsg.2019.07.018 [DOI] [PubMed] [Google Scholar]
- 15.Hornberger J, Grimes K, Naumann M, et al. Recognition, diagnosis, and treatment of primary focal hyperhidrosis. J Am Acad Dermatol 2004;51:274-86. 10.1016/j.jaad.2003.12.029 [DOI] [PubMed] [Google Scholar]
- 16.Lai YT, Yang LH, Chio CC, et al. Complications in patients with palmar hyperhidrosis treated with transthoracic endoscopic sympathectomy. Neurosurgery 1997;41:110-3; discussion 113-5. 10.1097/00006123-199707000-00023 [DOI] [PubMed] [Google Scholar]
- 17.Tu Y, Lin M, Chen J. Analysis of 10 Years of Outcomes Among 2206 Cases of Endoscopic Thoracic Sympathotomy for Primary Palmar Hyperhidrosis. Chinese Journal of Minimally Invasive Surgery 2017;17:99-103. [Google Scholar]
- 18.Hemead HM, Etman W, Hemead S, et al. Patients' satisfaction after bilateral thoracoscopic sympathicolysis. J Minim Access Surg 2023;19:478-81. 10.4103/jmas.jmas_179_22 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Vasconcelos CFM, Aguiar WS, Cordeiro GG, et al. Modified R5-R8 Thoracic Sympathectomy for Severe Compensatory Hyperhidrosis. Ann Thorac Surg 2021;111:e57-9. 10.1016/j.athoracsur.2020.05.099 [DOI] [PubMed] [Google Scholar]
- 20.Akil A, Semik M, Fischer S. Efficacy of Miniuniportal Video-Assisted Thoracoscopic Selective Sympathectomy (Ramicotomy) for the Treatment of Severe Palmar and Axillar Hyperhidrosis. Thorac Cardiovasc Surg 2019;67:415-9. 10.1055/s-0038-1642030 [DOI] [PubMed] [Google Scholar]
- 21.Han JW, Kim JJ, Kim YH, et al. New sympathicotomy for prevention of severe compensatory hyperhidrosis in patients with primary hyperhidrosis. J Thorac Dis 2020;12:765-72. 10.21037/jtd.2019.12.91 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Vasconcelos CFM, Aguiar WS, Tavares RM, et al. Bilateral R5-R8 sympathectomy for compensatory hyperhidrosis: complications and patient satisfaction. Rev Col Bras Cir 2020;47:e20202398. 10.1590/0100-6991e-20202398 [DOI] [PubMed] [Google Scholar]
- 23.Rojas D, Duggan SM, Mauduit M, et al. Impact of robotic-assisted and video-assisted sympathetic nerve reconstruction on quality of life for severe compensatory hyperhidrosis after thoracic sympathectomy. Interdiscip Cardiovasc Thorac Surg 2023;36:ivad106. 10.1093/icvts/ivad106 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Raveglia F, Orlandi R, Guttadauro A, et al. How to Prevent, Reduce, and Treat Severe Post Sympathetic Chain Compensatory Hyperhidrosis: 2021 State of the Art. Front Surg 2021;8:814916. 10.3389/fsurg.2021.814916 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Lee J, Jeong JY, Suh JH, et al. Thoracoscopic sympathetic block to predict compensatory hyperhidrosis in primary hyperhidrosis. J Thorac Dis 2021;13:3509-17. 10.21037/jtd-21-229 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Kara M, Kose S, Cayirci CE, et al. Can we predict the compensatory hyperhidrosis following a thoracic sympathectomy? Indian J Thorac Cardiovasc Surg 2019;35:190-5. 10.1007/s12055-018-0769-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Jeong SC, Kim JJ, Kim YH, et al. Heart rate variability as a potential diagnostic tool to predict compensatory hyperhidrosis after sympathectomy in patients with primary focal hyperhidrosis. J Thorac Dis 2020;12:6789-96. 10.21037/jtd-20-2038 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Hyun KY, Kim JJ, Im KS, et al. Machine learning analysis of primary hyperhidrosis for classification of hyperhidrosis type and prediction of compensatory hyperhidrosis. J Thorac Dis 2023;15:4808-17. 10.21037/jtd-23-471 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Simes BC, Moore JP, Brown TC, et al. Genetic polymorphism analysis of patients with primary hyperhidrosis. Clin Cosmet Investig Dermatol 2018;11:477-83. 10.2147/CCID.S176842 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Hashmonai M, Cameron AEP, Connery CP, et al. The Etiology of Primary Hyperhidrosis: A Systematic Review. Clin Auton Res 2017;27:379-83. 10.1007/s10286-017-0456-0 [DOI] [PubMed] [Google Scholar]