The Safety Assessment of Irrigation Fluid Management for Shoulder Arthroscopy and Its Effect on Postoperative Efficacy

Abstract

Objective

Fluid extravasation is a potentially dangerous complication associated with shoulder arthroscopy. Most relevant studies have involved respiratory system, while the primary purpose was to reveal the effects of the fluid extravasation on cardiovascular system and postoperative function.

Methods

The clinical data of 92 patients was retrospective analyzed, in which 84 cases with rotator cuff injury, three cases with shoulder instability, three cases with fractures of the greater tuberosity of the humerus, and two cases with frozen shoulder. All the patients were undergoing shoulder arthroscopy. The relationship between the basic information of the patients and cardiac index (CI) or pulse pressure variation (PPV) were evaluated by linear regression analysis. The change of CI or PPV at different states were evaluated by the one‐way analysis of variance. The liquid retention (TR) and postoperative clinical outcomes was analyzed using linear regression.

Results

The preoperative CI was affected by anesthesia status and body position, while PPV was not affected. Multivariate mixed‐effects model analysis of CI found that there was a statistically significant difference in groups of older than 55 years old and those with obesity (BMI > 24). After the operation, the retention of irrigation fluid significantly influenced the circumference of the deltoid (P < 0.001 (95%CI: [0.30, 1.00])), but not on the circumference of the deltoid, neck, and arm. The multivariate analysis of the American Shoulder and Elbow Surgery (ASES) scores at 3 and 6 months after surgery showed that the fluid retention volume was correlated with the ASES score at 3 months after surgery, especially when the retention volume was greater than 2 L (P = 0.001 (95%). %CI: [−12.49, −3.22]).

Conclusion

The retention of irrigation fluid after shoulder arthroscopic surgery causes swelling of local limbs, and has an effect on peripheral blood vessels, which is mainly reflected in its influence on PPV and the postoperative function. Therefore, surgeons need to improve their surgical technique, shorten the operation time and reduce fluid retention.

The irrigation fluid of shoulder arthroscopy had a significant influence on the cardiac index, and the postoperative liquid retention had an impact on the postoperative function recovery.

Introduction

The shoulder arthroscopic surgery has become increasingly more popular and worldwide. ¹ With the increasing popularity of shoulder arthroscopy, the related complications associated with shoulder arthroscopy have gradually attracted surgeon's attention. ² During the arthroscopy process, it is necessary to maintain the continuous infusion of normal saline and keep enough water pressure for adequate surgical field of view. However, the continuous fluid irrigation could lead to local fluid retention, which could cause potentially dangerous complications, such as airway compromise, and similar complications are more likely to occur with the use of a pressurized pump. ³ , ⁴ , ⁵ Although the similar complications were rare, the excessive fluid extravasation could cause the surrounding tissues edema and produce breathing compression, even severe cases could lead to death. ⁶ , ⁷ , ⁸ Moreover, large amounts of fluid retention, which was gradually and systemically absorbed, would contribute to laboratory abnormalities. The earliest complication of fluid extravasation was reported in the 1990s, including swelling around the neck, shoulders, and chest. ⁹ Chellam et al. followed up 32 patients with shoulder arthroscopy and found that 30 patients had neck circumference changes of about 1.17 ± 1.16 cm and two patients had neck circumference changes of more than 4 cm with severe airway compression. ¹⁰ Furthermore, Memon et al. systematically reviewed a total of 26 related studies including 205 patients and found there were many kinds of complication associated with fluid extravasation. ⁹ Recently, much fluid extravasation complication‐related research has mainly focused on the respiratory or nerve system problems, including severe upper respiratory compression and dysfunction and damage of remote organs, but there was little attention paid to the impact on the other important systems. ² , ⁷ Zhang et al. found a 12.9% incidence of moderate‐to‐severe pulmonary edema through 93 patients who had arthroscopic surgery. ¹¹ Sari et al. found the incidence of hypothermia after postoperative 2 h was 100% and the hypothermia could lead to increased risk of cardiovascular disease. ¹² Therefore, the occurrence of severe post‐arthroscopic complications requires consideration on the impact on other organ systems and more effective evaluation factors. Due to the potential effect of perfusion fluid on the cardiovascular system, we would like to investigate the safety of the arthroscopic operation with regard to the cardiovascular system. In addition, most research has focused on immediate postoperative severe complications or surrounding tissue swelling, but there was a lack of research on the effect of extraversive irrigation fluid on the postoperative function recovery.

Therefore, this research focused on the factors affecting the residual fluid retention during shoulder arthroscopy and postoperative functional recovery. We collected the basic information of arthroscopic patients and changes in indicators related to arthroscopic surgery over a period of time. The aims of the retrospective study were as follows: (i) under the condition of pressure pump perfusion, factors such as perfusion speed, postoperative perfusion volume, and operation time are related to patients, the influence of the systemic circulatory system, including whether there is a correlation between cardiac output index (CI) and pulse pressure variability index (PPV); (ii) whether the retention of perfusate will cause tissue changes around the shoulder joint; and (iii) whether the amount of perfusate retained affects the functional improvement of patients after surgery.

Finally, we aimed to reveal the absence of knowledge on the influence of retained fluid on other systems and figure out more appropriate duration and state of arthroscopic treatment, so as to provide suggestions and reference for further arthroscopic clinical treatment.

Patients and Methods

Inclusion criteria: patients (i) who were over 18 years of old; (ii) who were diagnosed with rotator cuff injury, acromion impingement sign, and adhesion of shoulder capsulitis and all received conservative treatments, with poor efficacy; (iii) who had shoulder joint instability; and (iv) who had shoulder joint dislocation and injury.

Exclusion criteria: patients (i) who were undergoing conservative treatment; (ii) who had unexplained shoulder joint pain and limited mobility; (iii) who were complicated with cervical spondylosis; (iv) who had a history of shoulder surgery on the affected side; (v) whose affected shoulder joint was infected; and (vi) who had neurological or psychiatric diseases.

The IRB/IEC Number: Ethics Approval was granted by the Ruijin Hospital Ethics Committee. (2021‐107).

General Information

A retrospective analysis was made on the clinical data of 92 patients who were actually enrolled in the study from 102 consecutive shoulder arthroscopy patients treated in our hospital from June 2020 to September 2020. Four patients (two patients with cervical spondylosis, one patient with ipsilateral shoulder reoperation, and one patient with Parkinson's disease) were excluded, and six patients who lost follow‐up were excluded. Among them, there were 36 left shoulders and 56 right shoulders, 27 males and 65 females, with an average age of (58.05 ± 10.77) years (19–74 years). There were 84 cases of rotator cuff injury, three cases of shoulder instability, three cases of fracture of the greater tuberosity of the humerus, and two cases of frozen shoulder. A total of 46 patients with comorbidities were followed up after 6 months. (Table 1).

TABLE 1.

Descriptive characteristics of the participants

Total	92
Age	58.05 ± 10.77
BMI	23.72 ± 3.54
Operation duration	88.80 ± 29.70
Sex
Male	27
Female	65
Operative category
Rotator cuff injury	84
Shoulder joint instability	3
Shoulder joint dislocation and injury	3
Frozen shoulder	2
Affected limb
Left	36
Right	56
Comorbidities	46
Hypertension	19
Diabetes	9
Coronary artery disease	7
Stroke	4
Lung disease	3
Tumor	2
Ankylosing spondylitis	1
Ulcerative colitis	1

Data Collection Procedures

For the enrolled patients, CNAP Monitor 500 was used to record the basic vital signs, CI and PPV of the patients before anesthesia; bilateral arm circumference, deltoid circumference, neck circumference, CI and PPV were measured in the supine position after general anesthesia, and then it was changed to lateral decubitus traction of the affected limb, and the CI and PPV values were recorded. All patients were treated with arthroscopic surgery. During the operation, a 3‐L bag of normal saline produced by Shandong Qilu Pharmaceutical was used as the perfusate for the arthroscopy, and the total infusion of the perfusate (TP) was counted after the operation. After the operation starts, a collection bag (Ioban, 3M) was used to collect the intraoperative perfusate, which was guided into a special measuring barrel (in liters) that had been marked on the wall of the cylinder in advance; suction channel and ground suction were also used to introduce the outflowing perfusate into the suction bottle; the moisture of the dressing was squeezed into the measuring barrel, and finally the measuring barrel and suction bottle were connected to the Zimmer waste liquid collector and the total amount of effluent (TO) was measured (suction pressure was maintained at 120 mmHg). Stryker's perfusion pump was used to maintain a water pressure of 50 mmHg in the operation. After the fluid collection was completed, the fluid retention (TR) inside the shoulder joint was calculated. The calculation formula: TR = TP − TO. The patient's CI and PPV values were collected throughout the operation and after the operation. After the operation, the patient's bilateral arm circumference, deltoid circumference and neck circumference were measured in the supine position.

Observation Indicators

Circulation Indicators

After the patient entered the operating room, general information was collected based on the inquiry and medical history data. The patient's CI and PPV were collected under the following conditions: (i) awake and in supine position; (ii) in supine position under general anesthesia; (iii) in lateral decubitus position under general anesthesia; (iv) in lateral decubitus position after perfusion; and (v) in lateral decubitus position after operation.

Body surface index

After the patient entered the operating room, the preoperative arm circumference, deltoid circumference and neck circumference were measured; after the operation, the above data were measured again in the supine position, and the data were analyzed statistically.

Indicators of pain and function

The visual analogue scale (VAS) before and 24 h after the operation was scored; the American Shoulder and Elbow Surgery Score (ASES) was measured before the operation, 3 and 6 months after the operation, and the data were analyzed statistically.

Statistical Analysis

Data are described as counts (%) for categorical variables and as mean (standard deviation) for continuous variables. CI and PPV values at different body positions and different surgical time was compared by One‐Way Repeated Measures analysis of variance, then a paired t‐test was used to perform multiple comparison with P value being adjusted by the Bonferroni method. Potential influencing factors on CI and PPV at different stages of surgery were analyzed using a mixed‐effects model. The relationship between volume of liquid retention (TR) and postoperative clinical outcomes (i.e., ASES functional scores, differences between preoperative and postoperative circumference of arm, deltoid muscle and neck) was analyzed using linear regression. The test level α value is 0.05 on both sides. All analyses were conducted in R‐4.1.2.

Results

The Influence from Anesthesia

A total of 92 patients were enrolled. The cardiac indexes were recorded in different conditions of the patient (awake state supine, anesthesia state supine, anesthesia state side lying). It was found that the CI was lower in the anesthesia state than that in the awake state (P < 0.001). After turning the lateral position under anesthesia, the CI would further decrease (P = 0.026 < 0.05). Both changes were statistically significant. There was no significant correlation between age and CI when awake. Under anesthesia and the subsequent lateral position, age and CI showed a certain linear correlation (anesthesia in the supine position [ACI]: p = 0.001 < 0.05, R ² = 0.115; LCI: P < 0.001, R ² = 0.224) (Fig. 1); There was no difference in PPV whether the patient in the awake or anesthesia station, or even in the state of postural changes (Fig. 2). The CI was analyzed by a multivariate mixed‐effects model, and it was found that age and BMI were influencing factors for CI, especially for elderly patients (older than 55 years old) and obese patients (BMI > 24; Table 2). Compared with CI, the PPV reflecting the peripheral circulatory system had no statistical difference in various states. And the PPV was not associated with the patient's age, gender and BMI.

Fig. 1.

(A) The value of cardiac index in the different period during the surgery. (B) CI value in the supine position (SCI). (C) CI value under anesthesia in the supine position (ACI). (D) CI value in the lateral position (LCI)

Fig. 2.

(A) The pulse pressure variation in the different period during the surgery. (B) PPV value in the supine position (SPPV). (C) PPV value under anesthesia in the supine position (APPV). (D) PPV value in the lateral position (LCI)

TABLE 2.

Analysis of the influencing factors of cardiac index (CI) after anesthesia

	Estimate	Pr(>|t|)	Low.ci	Upper.ci
Age
Age <= 55 (baseline)	0
Age > 55	−0.31	<0.001	−0.47	−0.15
BMI
BMI < 20 (baseline)	0
20 <= BMI < 24	−0.20	0.06	−0.41	0.01
BMI >= 24	−0.23	0.04	−0.45	−0.02
Sex
Male (baseline)	0
Female	0.14	0.11	−0.03	0.30

Influence Factors of Irrigation Fluid on Circulatory System

Based on the multivariate linear mixed effects model for CI and PPV, it was found that the obesity (BMI > 24) was still the factor affecting the CI of patients (P = 0.01 (95%CI: [0.09, 0.68])). The other potential factors, such as the amount of retention fluid, the operation time, has little effect on CI. However, the amount of retention fluid could affect the PPV and showed a statistically significant difference in the effect on PPV, especially greater than 2 L (P = 0.04 (95%CI: [0.13, 3.66]; Tables 3 and 4).

TABLE 3.

Analysis of the influencing factors of cardiac index (CI) during irrigation

	n	Estimate	Pr(>|t|)	Low.ci	Upper.ci
Age
Age <= 55 (baseline)	31	0
Age > 55	61	−0.11	0.33	−0.32	0.11
BMI
BMI < 20 (baseline)	11	0
20 <= BMI < 24	41	0.20	0.19	−0.10	0.49
BMI >= 24	40	0.38	0.01	0.09	0.68
Sex
Male (baseline)	27
Female	65	−0.06	0.58	−0.29	0.16
Retention
Retention < 2 (baseline)	46	0
Retention >= 2	46	−0.05	0.60	−0.24	0.14
Operation time
Time < 90 min	55	0
Time >= 90 min	37	−0.04	0.69	−0.15	0.22

TABLE 4.

Analysis of influencing factors on pulse pressure variation (PPV) during irrigation

	n	Estimate	Pr(>|t|)	Low.ci	Upper.ci
Age
Age <= 55 (baseline)	31	0
Age > 55	61	−1.17	0.26	−3.22	0.88
BMI
BMI < 20 (baseline)	11	0
20 <= BMI < 24	41	1.21	0.38	−1.54	3.96
BMI >= 24	40	−0.78	0.57	−3.55	1.98
Sex
Male (baseline)	27	0
Female	65	−0.36	0.74	−2.48	1.76
Retention
Retention < 2 (baseline)	46	0
Retention >= 2	46	1.90	0.04	0.13	3.66
Operation time
Time < 90 min	55	0
Time >= 90 min	37	0.38	0.66	−1.36	2.13

The Influence of the Retention Fluid Postoperation

The Spearman correlation coefficient results showed that there was a linear relationship between the preoperative and postoperative difference respectively in the circumference of the deltoid muscle (P = 0.26, p = 0.01), neck (P = 0.10, P = 0.35) and arm (P = 0.05, P = 0.68). However, multivariate analysis results showed that there was a significant difference between the amount of retention fluid and the difference of the deltoid muscle circumference before and after surgery (P < 0.001 (95%CI: [0.30, 1.00]); Fig. 3 and Table 5).

Fig. 3.

The relationship between the retention volume and deltoid muscle (A), neck (B) and arm circumference (C) before and after operation

TABLE 5.

Statistics of trtention volume, BMI and the difference of deltoid muscle circumference before and after operation

	Estimate	Pr(>|t|)	Low.ci	Upper.ci
Retention	0.65	<0.001	0.30	1.00
BMI
BMI < 20 (baseline)	0
20 <= BMI < 24	−0.15	0.90	−1.97	1.67
BMI >= 24	−1.51	0.21	−3.35	0.32

Based on the paired t‐test results, the ASES scores showed statistically improvements at postoperative 6 months compared with that at postoperative 3 months ((95%CI: [−22.26, −18.06]) P < 0.001) (Table 6). The multivariate analysis of ASES scores showed that the retention fluid volume was correlated with the ASES score at 3 months after surgery, especially the volume greater than 2 L (P = 0.001 (95%). CI: [−12.49, −3.22]; Fig. 4 and Table 7). The 6‐month ASES score was not affected by these factors. There was no correlation between retention fluid volume and the VAS score (24 h after operation: P = 0.264 > 0.05; 3 months after operation: P = 0.831 > 0.05; 6 months after operation: P = 0.321 > 0.05).

TABLE 6.

ASES score after 3 and 6 months follow‐up

Variable	n	Mean	sd
3 months	92	71.3	10.5
6 months	92	91.5	7.71

Note: Mean difference: −20.16 (95%CI: [−22.26, −18.06]) P < 0.001 (paired t‐test).

Fig. 4.

Multivariate analysis of ASES score with 3‐month follow‐up

TABLE 7.

Multivariate analysis of ASES score with 3‐month follow‐up

	Estimate	Pr(>|t|)	Low.ci	Upper.ci
Age
Age <= 55 (baseline)	0
Age > 55	0.06	0.98	−5.21	5.32
BMI
BMI < 20 (baseline)	0
20 <= BMI < 24 (baseline)	−2.69	0.45	−9.70	4.33
BMI >= 24	−3.38	0.34	−10.40	3.64
Retention
Retention < 2 (baseline)	0
Retention >= 2	−7.85	0.001	−12.49	−3.22
Irrigation
Irrigation < 20 L (baseline)
Irrigation >= 20 L	0.51	0.82	−4.04	5.07

Discussion

This research was one of the series of studies on arthroscopic enhanced recovery after surgery (ERSA), through retrospective study of the related circulatory system indicators about the intraoperative perfusion fluid flow management, in order to clarify the impact of intraoperative perfusion on circulatory system relationship and postoperative efficacy.

Pressure and Total Amount of Irrigation

With the increasing promotion of shoulder arthroscopy, the application of pressure pumps has become increasingly popular. The irrigation pressure not only determines the clarity of the surgical field of view, but also affects the safety of the operation. Sperber and Wredmark applied high‐pressure (100–150 mmHg, and the patient's intraoperative systolic blood pressure was controlled at 90 mmHg) during acromioplasty and measured the pressure in the deltoid and supraspinatus muscles. ¹³ During this period, the average intermuscular pressure is always less than 60 mmHg, when the intermuscular pressure is higher than 60 mmHg, it would affect the local microcirculation. ¹⁴ , ¹⁵ However, there would be a risk of compartment syndrome in a continuous high‐pressure state. At present, the irrigation pressure used in shoulder arthroscopy is generally controlled within the pressure range of 40–60 mmHg, which would not contribute to local tissue damage. Although the level of irrigation pressure determines the clarity of surgical field, surgeons prefer not to increase irrigation pressure for a clear view. Morrison et al. found that a good surgical field of vision could be maintained when the pulse pressure difference was less than 49 mmHg. ¹⁶ So, it was very important to maintain the pulse pressure difference and lowering systolic blood pressure would be much more effective and safer.

Generally, the average range of fluid perfusion in shoulder arthroscopy surgery is about 20–36 L with the corresponding operation time is 90–180 min, and some perfusion volumes up to 50–70 L even over 200 min. ⁹ In this study, the average perfusion volume was 22 L ± 13 L, and the average operation time was 89 min ±29 min, in which the longest operation time was 170 min, and the maximum perfusion volume was 56 L. Obviously, the perfusion volume and the operation duration increased proportionally. This study showed that the fluid retention was 2.3 L ± 1.7 L (0–7.6 L), and the patient with the largest fluid retention was female, with a preoperative chest circumference of 114 cm and a BMI index of 29.9. Our study made it clear that there was a correlation between the operation duration and perfusion volume and fluid retention, but it was hard to define the appropriate safe range for perfusion volume and operation time. Many authors suggested that the operation duration be controlled at 90–120 min, ⁹ , ¹⁶ which could reduce the risk of complications. In most cases of complications, the operation duration exceeded 180 min. ¹⁷ , ¹⁸ , ¹⁹

The Effect of Retention of Perfusion on the Circulatory System

Almost all patients after arthroscopy have swelling in the neck and shoulders, which was mainly caused by fluid extravasation during the operation. ² , ²⁰ , ²¹ Many reports had focused on its impact on the respiratory system, ¹⁷ , ²² , ²³ but would the intraoperative retention of perfusion affect the patient's cardiac function and overall volume status? In this study, the PPV and CI were used as the observation indicators of the intraoperative circulatory system. CI is an index that reflects cardiac output and could indirectly reflect cardiac pump function. However, CI is a static index, which would be easily affected by many factors, such as anesthesia status and body position as well as age and other factors. CI showed a negative correlation with the age increasing during the periods: before and after anesthesia, the change of body position and the end of operation. PPV is an index for evaluating the volume status of patients undergoing mechanical ventilation under general anesthesia and judging their volume responsiveness based on the mechanism of the cardiopulmonary interaction. It is a dynamic index and has recently been viewed for the commonly used indicators of anesthesia volume management. ²⁴ , ²⁵ Renner et al. believed that it was better to use the PPV to predict fluid reactivity with higher accuracy than using static indicators such as CVP and CI. ²¹ Therefore, PPV had great clinical value in predicting volume responsiveness. Our study showed that the change of PPV was only related to the retention of perfusion fluid, especially there was a statistical difference when the retention volume was greater than 2 L. This change did not appear in the change of CI. So, retention of perfusion fluid would affect the peripheral circulatory system, but due to the compensatory function of the heart, a few amounts of peripheral changes would not cause significant changes in cardiac load, which resulted in CI appearing to be different. Therefore, under the conditions of the current pressure pump perfusion pressure (50 mmHg) and flow rate (0.6–1.0 L/min) (rather than the high perfusion pressure used by Sperber and Wredmark), the normal saline was the safe and effective perfusion fluid for the patient's circulatory system. Even if acromioplasty is performed, it would not cause a large amount of extravasated fluid enter into the circulatory system or affect the function of the heart.

Influence of Retention Fluid on Postoperative

Generally, after the shoulder arthroscopy, the affected arm would swell due to fluid extravasation into the tissues. Our study evaluated the relationship between the fluid retention and the swelling ratio of the neck, arm, and deltoid muscle circumference before and after surgery. There was a positive linear correlation between the fluid retention and the swelling ratio of deltoid circumference, but no correlation between the ratio of the neck and arm circumference. The lateral position might cause the fluid flowing to the proximal of the arm. Due to the considerably dense tissue of deltoid muscle tissue, the neck circumference was not affected obviously by fluid extravasation. Therefore, fluid extravasation was less dangerous to the vital organs of the neck. Gupta et al. also believed that although the amount of fluid extravasation was large, it would not affect the airway and respiratory tract severely. ²⁶

Fluid retention did not significantly affect the VAS score after 24 h of post operation. Therefore, there were two inferences: (i) the pain after 24 h of post operation was mainly caused by surgical trauma and local inflammation; and (ii) the local tissue swelling caused by the fluid retention within 24 h could be basically relieved. Therefore, in terms of postoperative analgesia, more consideration should be given to treatments to eliminate local inflammation (ice pack cold compress, non‐steroidal anti‐inflammatory drugs, etc.), rather than relying on the dehydrating agent. ²⁰ , ²⁷ , ²⁸ , ²⁹ , ³⁰

Although the local tissue swelling caused by fluid extravasation always disappear in a short period of post operation, the authors found that the amount of fluid extravasation was negatively correlated with the ASES score after 3 months of post operation and had no correlation with the ASES score after postoperative 6 months. Therefore, the extravasation fluid showed a continuous effect on muscle tissue, which would lead to changes in ADL parameters and affecting the ASES score. Sperber and Wredmark found that the fluid perfusion increases the pressure in the supraspinatus muscle and the deltoid muscle. ¹³ With the prolongation of the operation time, the effect on the supraspinatus would determine the 3‐month functional assessment after the operation. However, functional scores mainly improved due to increased shoulder strength training after 6 months. In the shoulder joint rehabilitation program, whether strength training could enhance the early improvement of function requires need further research.

Strengths and Limitations

This research systematically summarized the impact of shoulder arthroscopy perfusion on surgical safety and postoperative outcomes. Based on dynamic non‐invasive detection of PPV and CI, we verified that the amount of fluid retention after shoulder arthroscopy would not affect the circulatory system During the entire operation, PPV was more reliable than CI in assessing the changes in the circulatory system of the non‐postural environment. Furthermore, the amount of fluid remaining after shoulder arthroscopic perfusion showed a negative correlation with the patient's short‐term postoperative function.

However, there were still some limitations. Regarding study limitations, first, cause the pulse pressure variation rate (PPV) was a dynamic value throughout the operation, we could only take the average value as the valid data in the special phase, so the advantages of the PPV dynamic character could not be well reflected. Second, the non‐invasive advantage was the primary reason for selecting PPV as the monitoring factor, and the non‐invasive monitoring approaches were widely used in clinics, but there is still a certain error with the invasive monitoring data. Finally, unfortunately this is not a prospective cohort study, so there are certain limitations. In the later series of clinical studies, we will make up for this shortcoming. Despite the above limitation, our study was the first to reveal the impact of fluid management on the cardiovascular system during the shoulder arthroscopy.

Conclusion

Extravasation of irrigation fluid in shoulder arthroscopic surgery is often a common complication. We found that the extravasation of irrigation fluid would affect the peripheral circulatory system but would not affect the heart function. The extravasation of irrigation fluid mainly causes changes in the circumference of the deltoid muscle and has little impact on the cervical airway and respiratory tract. However, the extravasation had no correlation with postoperative pain. The amount of extravasation was related to time and affected the postoperative shoulder function recovery. Therefore, the surgeon needs to improve the surgical technique and shorten the operation time to reduce the fluid retention and improve the postoperative function.

Author Contributions

CY Zhuang, RH Yang and Yang X performed surgery and designed experiments and contributed equally to this work. YY Song and Y Zhang previously analyzed the data. JF Liu, F Yang and XH Huang collected the follow‐up data. YY Song and J Liu carried out all analyses. XN Wang and Y Wang wrote the first draft of the submitted manuscript. All authors provided feedback and commented on the manuscript. CY Zhuang and L Wang secured funding for the project. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.