Objective: The goal of this study is to evaluate the sensitivity and specificity of the presence of pneumothorax on a 4-hour CXR after image-guided transthoracic lung biopsy (TTLB) for diagnosing subsequent clinically significant pneumothorax.

Hypothesis to be tested: The investigators hypothesize that an absence of pneumothorax on CXR at 4 hours can accurately predict an absence of clinically significant pneumothorax at 16 to 20 hours.

Design and subjects: This is a prospective, single-centre, diagnostic accuracy study conducted at a tertiary referral centre in Hong Kong. Patients will be recruited if an image-guided TTLB is arranged.

Study instruments: CXRs will be performed for patients at different time points. The presence and size of pneumothorax (if any) will be recorded.

Outcome: The diagnostic capacity of a 4-hour CXR after image-guided TTLB will be provide evidence regarding the safety and patient selection for same-day discharge.

Image-guided transthoracic lung biopsy (TTLB), by computed tomography (CT) or ultrasound guidance, is an important diagnostic modality for various pulmonary diseases. One of the most common and clinically important complications of TTLB is iatrogenic pneumothorax. The estimated incidence of iatrogenic pneumothorax following TTLB ranges between 12 and 45%, and 2 to 15% of these patients required chest drain insertion.

The majority of post-TTLB iatrogenic pneumothorax occurs immediately after the procedure, but delayed pneumothorax, which happened hours after the biopsy, is a recognized complication. The incidence rate of delayed pneumothorax ranges between 0.4 and 8.6%, and mostly could be detected at or within 4 hours after biopsy. Among these delayed pneumothorax, up to 60% were clinically significant and required chest drain insertion.

There is no high-quality evidence informing physicians of the best timing of CXR in detecting delayed pneumothorax. An ideal timing for this CXR should be late enough to capture the most delayed clinically significant pneumothorax yet promptly detect deterioration due to pneumothorax. The British Thoracic Society guideline recommends an erect CXR 1 hour after the biopsy, which is sufficient to detect the majority of post-biopsy pneumothorax. The guideline also mentions that patients should be warned of delayed pneumothorax but does not mandate the need for subsequent CXR. Overnight observation with a CXR the next day, i.e. 16 to 20 hours after biopsy, is a common local practice to detect the occurrence of delayed pneumothorax. This practice, although safe, has not been examined extensively and involves a longer hospital stay for all patients receiving TTLB. TTLB as a day procedure has been described instead. A series of CXRs were used to ensure adequate detection of enlarging iatrogenic pneumothorax, but it was labour-intensive.

Our group conducted an internal audit covering 3 months of hospital data in Prince of Wales Hospital, which includes 109 patients who underwent image-guided TTLB between Nov 2023 and Jan 2024. The incidences of pneumothorax and clinically significant pneumothorax requiring chest drain insertion were 23 (21.1%) and 7 (6.4%), respectively. Among those patients who had chest drain insertion, 4 had drains inserted within 3 hours after biopsy. The remaining 3 patients had chest drains inserted at 16 to 19 hours after biopsy. Their initial CXR at 1 hour showed no pneumothorax, but did not receive interim CXR in between. A separate group of 11 patients with CXR between 3 to 6 hours and no clinically significant pneumothorax were safely discharged the next day. These preliminary data suggest that an interim CXR at 4 hours may allow timely detection of clinically significant pneumothorax for early intervention, and an absence of pneumothorax at 4 hours can rule out the possibility of delayed pneumothorax at a later time point. This may facilitate early discharge in at least 80% of patients.

This prospective study aims to evaluate the diagnostic capacity of a 4-hour CXR after image-guided TTLB in predicting the clinically significant pneumothorax at 16 to 20 hours after biopsy. The investigators hypothesized that an absence of pneumothorax on CXR at 4 hours can accurately predict an absence of clinically significant pneumothorax at 16 to 20 hours. As the best way of measuring the size of iatrogenic pneumothorax has not been defined, several pragmatic methods with arbitrary cutoff values will be assessed in the study, including a binary outcome of pneumothorax (present or absent), apex-to-cupula distance (≤ 2 cm or > 2 cm), width of pneumothorax at the hilar level (≤ 1 cm or > 1 cm), Light's method (≤ 10% or > 10%) [11] and Collins method (≤ 10% or > 10%) [12]. The data collected from this study will provide evidence regarding the safety and patient selection for same-day discharge in patients undergoing image-guided TTLB.

Inclusion Criteria:

• Patients whose image-guided TTLB, including ultrasound or CT-guided, is arranged for diagnostic purposes
• Patients who are able to sign written informed consent to participate in the study

Exclusion Criteria:

• Patients with resolution of lung lesion prior to the scheduled image-guided TTLB
• Patients have active exacerbation of underlying diseases causing unstable respiratory conditions (e.g. exacerbation of COPD, pneumonia) which would lower the threshold for invasive interventions for iatrogenic pneumothorax
• Patients with psychiatric disease or cognitive impairment that may limit their ability to understand or give consent to the study
• Patients who decline to join the study

Patients who will require image-guided TTLB