Iatrogenic pneumothorax associated with surgeries at anatomically thoracic-adjacent and non-adjacent sites: case report and scoping review

Abstract

Background

Iatrogenic pneumothorax associated with surgeries at anatomically thoracic-adjacent and non-adjacent sites is characterized by an insidious onset, delayed presentation, and high rates of misdiagnosis, posing significant risks to patients. This study presents a case report and scoping review, and proposes a practical and cost-effective proactive monitoring strategy.

Case presentation

We report a rare case of complete pneumothorax following thoracolumbar spinal fixation surgery. The patient developed hypoxemia and diminished right breath sounds in the post-anesthesia care unit (PACU). Bedside lung ultrasound by the anesthesiologist suggested pneumothorax. Given the unavailability of bedside chest X-ray and the thoracic surgeon’s doubt regarding ultrasound diagnostic qualifications, an emergency CT scan was performed, which confirmed right-sided complete pneumothorax caused by pedicle screws penetrating the vertebral body into the thoracic cavity. Immediate closed chest drainage led to rapid symptom relief.

The scoping review identified 18 sites of pneumothorax resulting from surgeries at anatomically thoracic-adjacent and non-adjacent sites, including the thyroid, clavicle, breast, liver, stomach, spleen, spine, shoulder arthroscopy, back acupuncture, liposuction, pacemaker placement (involving the hypoglossal nerve and heart), subdural-peritoneal shunt, costal cartilage, kidney, gallbladder, inguinal hernia, and the frontal facial region. Anatomical adjacency is the primary cause of pneumothorax resulting from surgeries at thoracic-adjacent sites. Pneumothorax can also occur in cases where the surgery is distant from chest structures, with carbon dioxide pneumothorax resulting from laparoscopic surgery being the most common. Clinical symptoms remain the primary means of initial detection; however, there is a lack of predictability and proactivity.

Conclusion

We suggest the implementation of a low-threshold, symptom-driven lung ultrasound screening protocol in the PACU for patients who present with unexplained hypoxemia, respiratory symptoms, or suspected intraoperative pleural injury, particularly after undergoing surgeries at anatomically thoracic-adjacent or non-adjacent sites. This targeted strategy effectively balances the need for early diagnosis with clinical feasibility.

Introduction

Postoperative chest complications vary considerably, primarily involving pulmonary, pleural, and cardiac issues. These pose significant challenges in perioperative management and can lead to increased mortality and healthcare burdens [1, 2]. Among these, iatrogenic pneumothorax associated with surgeries at anatomically thoracic-adjacent and non-adjacent sites is characterized by insidious onset, delayed manifestation, and high rates of misdiagnosis. This may lead to it being underestimated and overlooked in clinical practice, and its clinical significance is often not fully recognized [3, 4]. Current postoperative monitoring protocols often rely on the appearance of overt symptoms, such as sudden hypoxemia and dyspnea, which typically only manifest when the pneumothorax has progressed to a severe stage [5]. This delay in detection can lead to catastrophic consequences, including respiratory distress, tension pneumothorax, longer hospital stays, reintubation, and even death [6].

Early recognition and intervention are crucial for improving outcomes. Lung ultrasound (LUS) is a mature, effective, portable, and non-invasive tool for diagnosing chest complications. Although its advantages in early diagnosis are evident, its application during the perioperative period remains inadequate, particularly in the post-anesthesia care unit (PACU) [7, 8]. Here, we describe a case of pneumothorax, provide a scoping review summarizing the clinical features of the condition, and propose an early proactive monitoring strategy for ultrasound screening in the PACU. Additionally, we explore the clinical value, feasibility, and legal implications of this strategy.

Case report

Written informed consent was obtained from the patient prior to the writing of this report.

Case 1

A 73-year-old female presented for treatment after a fall that resulted in a compression fracture of the T12 thoracic vertebra and thoracic spinal stenosis. She was scheduled for posterior spinal fusion with bone grafting at T10-L2 and internal fixation with pedicle screws. Her medical history included hypertension and dyslipidemia, with regular long-term administration of nimodipine and atorvastatin medication that controlled her blood pressure effectively. She had previously undergone general anesthesia without complications. The routine general anesthesia protocol was followed preoperatively and intraoperatively, and both the surgical and anesthetic procedures proceeded smoothly. Anesthesia was induced using propofol, sufentanil, and cisatracurium, and maintained with remifentanil at a rate of 0.2 µg/kg/min and propofol at a rate of 4 mg/kg/h, with additional doses of cisatracurium administered every 45 min. No inhaled anesthetic agents were employed. During the operation, the airway pressure was maintained at 20–25 cmH₂O. After the placement of the screw, there was a gradual and slight increase in peak pressure; however, it did not exceed 25 cmH₂O. Post-surgery, she was transferred to the PACU for recovery. After ensuring that she met the standard criteria for routine recovery in the PACU, the endotracheal tube was removed for continued observation. Prior to extubation, the patient exhibited a respiratory rate of 20–25 breaths/min, a tidal volume of 350–400 milliliters, and a heart rate ranging from 100 to 110 beats/min. Additionally, no nebulization therapy was administered after extubation.

However, while inhaling pure oxygen, her oxygen saturation gradually decreased, reaching a minimum of 87% and fluctuating between 97, 87, and 87–92%. At this time, her modified aldrete score was 8 (with scores of 2 for activity, 1 for respiration, 2 for circulation, 2 for consciousness, and 1 for oxygen saturation). At this moment, the patient’s heart rate is 105 beats per minute, blood pressure is 135 over 70 millimeters of mercury, and auscultation of both lungs indicated a markedly reduced breath sound on the right side. The anesthesiologist in the PACU immediately performed a bedside chest ultrasound, revealing the disappearance of the right pleural sliding sign and an M-mode pattern consistent with a “stratosphere sign,” indicative of pneumothorax. An urgent request was made to the thoracic surgeon to perform a right chest tube placement for closed drainage. However, the thoracic surgeon questioned the anesthesiologist’s qualifications and ability to diagnose pneumothorax via ultrasound and insisted on a chest CT for a definitive diagnosis before proceeding with treatment. The patient was then urgently transported to the CT examination room, where the diagnosis was confirmed and a right chest tube placement for closed drainage was performed, resulting in significant symptom improvement. The key events from PACU admission through follow-up are summarized in a timeline (Fig. 1). The chest CT indicated a large right pneumothorax and complete lung collapse, as shown in Fig. 2. Chest CT and CT 3D reconstruction imaging of the thoracic spine was conducted to clarify the cause, revealing that the spinal fixation screw had penetrated the vertebral body into the thoracic cavity, as shown in Fig. 3. The patient’s clinical symptoms improved after the right chest closed drainage treatment, and the drain was removed five days later, allowing for spontaneous recovery. The patient did not undergo any further surgical intervention subsequently, such as screw repositioning or removal. A follow-up in the outpatient clinic six months later revealed no complications related to the chest.

Fig. 1.

Case Timeline. Key events from PACU admission through follow-up, created by Figdraw. Note: The inset shows a representative M-mode stratosphere sign image from Xie et al. [8] (licensed under CC BY 4.0), marked with “*”. PACU: Post-Anesthesia Care Unit; POD: Postoperative Day

Fig. 2.

Imaging data of Case 1. A Post-admission MRI revealed a compression fracture of the T12 vertebral body with spinal cord compression. B Emergency postoperative chest CT showed severe right lung collapse with mediastinal shift, indicative of complete pneumothorax. C The chest X-ray reveals an indwelling closed drainage tube positioned in the right side of the thoracic cavity. D Follow-up imaging on postoperative day 2 after closed thoracic drainage demonstrated near-complete recovery of left lung function. All image captures preserved original metadata including essential information and dates, while patient names and institutional details were redacted in compliance with privacy protection protocols

Fig. 3.

Imaging evidence of complete pneumothorax etiology in Case 1. A Emergency postoperative chest CT showing complete right pneumothorax, with a pair of screws definitively breaching the T10 vertebral body and connecting to the right pleural cavity, showing fragmented bone tissue. B A pair of screws were suspected to have breached the T11 vertebral body. C and D Postoperative thoracolumbar CT 3D reconstruction showed conspicuously “exposed” screws at the T10 and T11 thoracic vertebrae. All image captures preserved original metadata including essential information and dates, while patient names and institutional details were redacted in compliance with privacy protection protocols

Scoping review

In May 2025, a targeted scoping search was conducted utilizing the electronic databases PubMed, Embase, and Cochrane Library. The search strategy employed the following combination of keywords: “Postoperative chest complications AND Pneumothorax NOT (Deep Vein Puncture OR Central Venous Catheterization OR Conference Abstract)”. This review was restricted to English-language literature published between January 1, 2005, and May 31, 2025, with full text accessibility. To clearly describe the specific surgical sites responsible for pneumothorax in surgeries at anatomically thoracic-adjacent and non-adjacent sites, additional screening was performed: procedures that required access to the thoracic cavity were excluded; cases of spontaneous pneumothorax were omitted; for identical surgical sites, only one reference was retained following discussion; and cases with unclear etiology of pneumothorax were excluded. During the literature screening phase, pneumothorax cases associated with deep vein puncture were excluded (such as internal jugular vein and subclavian vein catheterization), as they represent independent procedural complications that are not directly related to the surgical procedure itself. All literature searches were carried out independently by two anesthesiologists who had undergone specialized training, with the screening and exclusion criteria finalized through discussions among all researchers.

The scoping review summarizes iatrogenic pneumothorax associated with surgeries at anatomically thoracic-adjacent and non-adjacent sites, encompassing 18 different locations. Among these, eleven sites are anatomically adjacent to the thoracic cavity, and included surgeries involving the thyroid, clavicle, breast, spine, shoulder arthroscopy, back acupuncture, liposuction, pacemaker placement (hypoglossal nerve and heart), subdural-peritoneal shunt and the costal cartilage [9–19]. Seven sites are anatomically non-adjacent to the thoracic cavity include the spleen, stomach, liver, gallbladder, kidney, inguinal hernia, and maxillofacial region [20–26]. Five sites experienced carbon dioxide pneumothorax due to laparoscopic surgery, namely, the costal cartilage, liver, gallbladder, kidney, and inguinal hernia, of which four sites are not adjacent to the thoracic cavity [19, 22–25]. In a report on frontal facial surgery, the pneumothorax was explicitly linked to intubation, where the cause was related to mucous plug obstruction due to injury during nasal or oral intubation, leading to rupture of the alveoli during positive-pressure ventilation [26]. Detailed information is provided in Table 1, with the specific site distributions illustrated in Fig. 4.

Table 1.

Scoping review of iatrogenic pneumothorax associated with surgical sites anatomically adjacent and non-adjacent to the thoracic cavity

	Reference	Year	Surgery Type	Anatomical Location	Time of Detection	Related Conditions (Minimum SpO2)	Cause	Volume	Diagnostic Methods	Special Treatment
1	Solmaz et al. [9]	2011	Subdural-peritoneal shunt	Neck (adjacent)	Intraoperative	86%	Puncture tunnel injury mediastinal pleura	Bilateral, medium volume	Chest X-ray	Closed thoracic drainage
2	Lee et al. [10]	2010	Right thyroidectomy + left neck dissection	Thyroid (adjacent)	Postoperative PACU	65%	Injury of pleural roof during cervical lymphadenectomy	Right, medium	Chest X-ray	Emergency tracheal intubation + closed thoracic drainage
3	Beiriger et al. [11]	2025	Hypoglossal nerve stimulator implantation	Neck (adjacent)	Postoperative PACU	Not explicitly mentioned	Damage of pleural roof during electrode implantation	Small (0.6%)	Chest X-ray	Conservative observation (no intervention required for asymptomatic patients)
4	Darwich et al. [12]	2019	Pacemaker implantation	Subclavian vein (adjacent)	1 day after surgery	83%	Unilateral pneumothorax caused by puncture of left subclavian vein, spreading to both sides through congenital pleural channel	Bilateral (left 35%, right 20%)	Chest X-ray, CT	Closed left thoracic drainage, bilateral pneumothorax resolved
5	Kim et al. [13]	2019	Minimally invasive plate osteosynthesis for middle clavicular shaft fracture	Clavicle (adjacent)	Postoperative PACU	93%-94%	Injury to pleura during plate insertion	Right, medium	Chest X-ray	Closed thoracic drainage
6	Alwadai et al. [14]	2023	Liposuction + abdominoplasty	Upper torso (adjacent)	1 day after surgery	Not explicitly mentioned	Too deep liposuction cannula damages lung apex	Left, middle volume	CT	Closed thoracic drainage
7	Bamps et al. [15]	2016	Shoulder arthroscopic surgery	Shoulder (adjacent)	4 hours postoperatively	83%	Arthroscopic pump pressure is too high, causing air to enter the thoracic cavity through portals, forming tension pneumothorax	Left, Tonicity	Chest X-ray	Emergency closed thoracic drainage, extubation 4 days after operation
8	Oskarsson et al. [16]	2017	Acupuncture	Back (adjacent)	Postoperative Immediate / Delayed	84%	Excessive depth of acupuncture needle insertion (estimated to penetrate at least 5 cm to breach the parietal pleura, based on the patient's BMI of 30)	Bilateral, massive	Chest X-ray	Bilateral Seldinger chest tube insertion, removed after 3 days of hospitalization with discharge
9	Flynn et al. [17]	2022	Bilateral mastectomy + immediate reconstruction	Breast (adjacent)	Postoperative PACU	91%	Intraoperative puncture or anatomical injury	Bilateral, medium volume	Chest X-ray	Closed thoracic drainage
10	Lewis et al. [18]	2014	Posterior spinal fusion + osteotomy	Thoracic vertebrae (adjacent)	Postoperative PACU	86%	Pedicle screw misplacement or tear pleura during osteotomy	Unilateral medium volume	Intraoperative ultrasound, chest radiograph	Open pleural decompression during operation, closed drainage after operation
11	McCarn et al. [19]	2011	Costal cartilage harvesting	Rib (adjacent)	PACU	92%	Damage to the pleura during costal cartilage harvesting, with CO₂ pneumoperitoneum spreading into the pleural cavity.	Right, Moderate	Ultrasound (preferred), chest X-ray	Conservative observation, oxygen therapy, spontaneous absorption
12	Bonnard et al. [20]	2014	Laparoscopic splenectomy (children with sickle cell disease)	Spleen (not adjacent)	postoperation	Not explicitly mentioned	Surgical operation injured pleura, and postoperative pleural effusion induced ACS	Unspecified	Chest X-ray, clinical manifestations	Conservative treatment (closed thoracic drainage)
13	Cai et al. [21]	2010	Laparoscopic Nissen fundoplication	Stomach (not adjacent)	3 days after surgery	92%	Hernia of gastric fundus into thoracic cavity resulting in pleural tear	Left side	Chest X-ray, CT	Open repair of hernia contents, closed thoracic drainage
14	Chae et al. [22]	2020	Laparoscopic right hepatectomy (living donor for liver transplantation)	Liver (not adjacent)	Intraoperative	88%	Escape of intraperitoneal CO₂ into pleural cavity via potential congenital or iatrogenic micro-channel in the diaphragm	Large (right-sided dominant) + small left lung collapse	Point-of-care lung ultrasound, Chest X-ray	Temporary cessation of CO₂ insufflation, Right-sided chest tube insertion
15	Karayiannakis et al. [23]	2005	Laparoscopic cholecystectomy	Gallbladder (not adjacent)	Intraoperative	92%	CO ₂ enters the thoracic cavity through the congenital weak point of the diaphragm, forming right pneumothorax	Right, mass	Chest X-ray	Conservative observation, CO ₂ absorbs itself (subsides after 3 hours)
16	Zhao et al. [24]	2022	Retroperitoneal laparoscopic partial nephrectomy	Kidney (not adjacent)	Intraoperative	85%-88%	High pressure CO ₂ inflation causes diaphragm damage	Medium volume	Ultrasound, CT	Emergency thoracentesis decompression to reduce pneumoperitoneum pressure
17	Liu et al. [25]	2024	Laparoscopic inguinal hernia repair	Inguinal hernia (not adjacent)	Postoperative PACU	65%	High pressure CO ₂ enters the thoracic cavity through the diaphragm defect, and N ₂ O exacerbates emphysema	Bilateral, mass	Chest X-ray, CT	Emergency closed thoracic drainage
18	AI Shetawi et al. [26]	2016	Maxillofacial trauma repair	Maxillofacial region (not adjacent)	Intraoperative	60%	Repeated nasal intubation caused nasal hemorrhage, mucus plug obstructed the upper lobe of right lung, and alveolar rupture caused by positive pressure ventilation	Right, tension pneumothorax	Chest X-ray, bronchoscopy	Emergency closed thoracic drainage and postoperative extubation observation

Fig. 4.

Distribution map of specific anatomical sites for cases of iatrogenic pneumothorax associated with surgeries at anatomically thoracic-adjacent and non-adjacent sites. Note: The specific surgical sites of iatrogenic pneumothorax associated with surgeries at anatomically thoracic-adjacent and non-adjacent sites have been recorded (The specific site of the hypoglossal nerve stimulator implantation has not been marked.). Due to overlapping anatomical layers or their location in the back region, the images do not display the precise anatomical structures; however, specific locations have been indicated as “⊙”

Discussion

The anatomical proximity to the thoracic cavity is a major contributing factor to iatrogenic pneumothorax associated with surgeries at anatomically thoracic-adjacent and non-adjacent sites. Among the established 18 surgical sites, 11 (such as thyroid surgery, spinal procedures, and shoulder arthroscopy, etc.) are adjacent to the thoracic cavity, as detailed in Table 1 and illustrated in Fig. 4. Symptoms such as unexpected hypoxemia that occur clinically often indicate a severe pneumothorax. The delayed onset of these symptoms hours to days after surgery, along with the worsening of symptoms following positive pressure ventilation, highlights the critical need for early proactive monitoring [14, 21, 27, 28]. It is both surprising and noteworthy that laparoscopic surgery has become the most common non-adjacent etiology, primarily due to the entry of high-pressure carbon dioxide into the thoracic cavity through diaphragmatic defects [23–25]. In cases of CO₂ pneumothorax, for patients with a pneumothorax volume of less than 50% and oxygen saturation of 90% or higher, conservative management such as oxygen therapy and close monitoring should be prioritized [19, 23]. For cases of massive pneumothorax, persistent hypoxemia, or tension pneumothorax, the placement of a closed thoracic drainage system is necessary [22, 24, 25]. This invasive method continues to be the most effective and swift intervention available [29]. However, in comparison to the more prevalent gas pneumothorax, carbon dioxide pneumothorax is distinguished by a higher absorption rate within the body, and the clinically acceptable lower limit of oxygen saturation may indeed be lower [30]. Therefore, it is recommended that further prospective studies be undertaken in the future to validate the effectiveness of conservative management for CO₂ pneumothorax.

LUS should be highlighted as a valuable adjunct to facilitate earlier detection and triage, as illustrated in Case 1. Current techniques for ultrasound diagnosis of pneumothorax have reached a high level of maturity, characterized by the absence of pleural sliding, the presence of the “stratosphere sign” in M-mode ultrasound, and the lack of B-lines [31]. Xie et al. reported that the diagnostic accuracy of ultrasound for pneumothorax is on par with that of chest CT, achieving a sensitivity of 90.0% and a specificity of 98.9% [8]. Meanwhile, its portability and non-invasive nature make it a crucial diagnostic tool in the PACU, while also facilitating the identification of hemothorax and atelectasis, thus reducing the risk of reintubation [32–34]. However, the legal qualification requirements for anesthesiologists to perform LUS remain inconsistent around the world. Although anesthesiologists undergo specialized training in ultrasound, formal qualification certification is still required, such as the National Ultrasound Examination in China or the AFC certification in Canada [35–38]. The Canadian Society of Echocardiography (CSE) provides guidelines rather than formal certification [39]. Given the regional differences that exist, the establishment of a globally unified standard is both desirable and profoundly challenging.

In this case, the practice of postponing intervention for pneumothorax while awaiting imaging confirmation is inadvisable. This delay may have contributed to the progression of the condition to a right-sided complete pneumothorax (Fig. 2). This case highlights a crucial perioperative principle: when tension pneumothorax is suspected based on clinical evaluation (such as acute hypoxemia, absence of breath sounds, hemodynamic instability, or elevated airway pressure), immediate decompression should be performed without waiting for confirmatory imaging [40]. It also serves as a reminder that prompt and effective clinical management is critical, and imaging should not become a bottleneck that delays treatment. Furthermore, chest CT is instrumental in identifying the underlying cause (such as improper screw placement in this case), but it should only be conducted after the patient’s condition has stabilized and the emergency has been appropriately addressed. In this context, LUS can serve as a valuable bedside adjunct. While LUS offers high sensitivity, bedside accessibility, real-time diagnostic capabilities, and is widely used in perioperative and intensive care settings, only 4 out of the 18 studies included in this scoping review utilized ultrasound examination [41, 42]. This underscores that the application of LUS in diagnosing acute pneumothorax has not yet become widespread. Bedside chest X-ray remains the preferred method for diagnosing pneumothorax, however, clinical judgment must always take precedence in emergency situations [43].

A review of the literature reveals that 12 out of 18 cases of pneumothorax were identified either during the surgical procedure or in the PACU. In comparison to cases of pneumothorax identified in non-operating room settings (where oxygen saturation was consistently above 80%), the minimum oxygen saturation (SpO₂) levels recorded in intraoperative and PACU cases were significantly lower, with three cases falling below 80%. This disparity highlights the delays associated with symptom-driven monitoring. It has traditionally been considered that the detection rate of routine postoperative chest X-rays in the PACU is low and potentially harmful, thereby rendering their implementation unnecessary [44]. However, this perspective may require revision. In the future, striking a balance between enhancing early detection rates and managing the burden on healthcare resources will be a critical issue that urgently requires attention in the risk management of unexpected pneumothorax.

We need to acknowledge several limitations, including the small sample size, lack of long-term follow-up, absence of radiologist-confirmed ultrasound reports, and potential selection bias. Simultaneously, the scoping search was limited to English databases, potentially omitting studies in other languages. The false positive and negative rates of ultrasound screening were not specified, which may affect the assessment of the strategy’s risks and benefits.

In summary, we advocate for a clinically-driven, low-threshold LUS approach for high-risk surgical populations in the PACU, particularly for those undergoing thoracic or laparoscopic procedures. When encountering unexplained hypoxemia, dyspnea, diminished breath sounds, or suspicious signs of intraoperative pleural breach, a LUS examination should be promptly initiated. This targeted strategy aims to optimize diagnostic efficacy and resource utilization, which awaits validation through prospective studies.

Authors’ contributions

Conceptualization, HB.P.; Methodology, HB.P. and XP.W.; Writing – original draft, HB.P. and JX.O.; Writing – review & editing, HB.P., JJ.W., ML.C., Y.J. and JX.O. Literature search – HB.P. and XP.W.

Funding

The authors have not disclosed a specific grant for this research from any funding agency in the public, commercial, or non-profit sectors.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

The cases reported in this study have obtained the patient’s consent and have signed the informed consent form. All authors reviewed and agreed with the manuscript.

Consent for publication

We confirm that written informed consent has been obtained from the patient for the publication of their personal clinical details and any identifiable images in this study.

Competing interests

The authors declare no competing interests.