Diagnostic Accuracy of Point-of-Care Ultrasound in Detecting Pneumothorax: A Systematic Review and Meta-Analysis

TL;DR

This study reviews and analyzes the accuracy of using point-of-care ultrasound to detect pneumothorax in emergency settings.

Contribution

The paper provides an updated meta-analysis of POCUS diagnostic accuracy for pneumothorax using PRISMA-DTA guidelines.

Findings

POCUS has a pooled sensitivity of 74.3% and specificity of 99.1% for detecting pneumothorax.

Higher sensitivity is observed with prospective studies, non-trauma patients, and experienced operators.

Specificity remains consistently high across all subgroups, with no significant publication bias detected.

Abstract

Point-of-care ultrasound (POCUS) is increasingly used in emergency departments for the rapid detection of pneumothorax. While POCUS offers bedside convenience, its diagnostic accuracy compared with standard imaging remains variable, necessitating an updated synthesis of evidence. We conducted a systematic review and meta-analysis following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Diagnostic Test Accuracy (PRISMA-DTA) guidelines. Databases including PubMed, Scopus, Web of Science, and Google Scholar were searched for studies evaluating POCUS for pneumothorax detection in emergency settings. Data on sensitivity, specificity, and operator characteristics were extracted. Study quality was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2), and a bivariate random-effects model was used to pool diagnostic accuracy metrics. Subgroup and sensitivity analyses explored sources of heterogeneity. Fifteen studies comprising 3,840 patients were included. Pooled sensitivity of POCUS for detecting pneumothorax was 74.3% (95% CI: 55.4-87.4%), and pooled specificity was 99.1% (95% CI: 98.5-99.9%). Diagnostic odds ratio was 104.4 (95% CI: 93.0-112.8), with positive and negative likelihood ratios of 15.1 (95% CI: 5.66-40.38) and 0.028 (95% CI: 0.008-0.095), respectively. Subgroup analyses showed higher sensitivity for prospective studies, non-trauma patients, the two-point ultrasound protocol, and operators with greater training and experience, while specificity remained consistently high across all subgroups. No significant publication bias was detected. POCUS is a highly specific and moderately sensitive tool for the rapid detection of pneumothorax in emergency settings. Diagnostic performance improves with standardized protocols and experienced operators. Despite some variability in sensitivity, POCUS can reliably identify pneumothorax and reduce unnecessary thoracic interventions.

Figures

PRISMA flow chart illustrating the study selection process for this systematic review and meta-analysis.PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Forest plot of pooled sensitivity.TP: true positive; FP: false positive; FN: false negative; TN: true negative; CI: confidence intervalReferences [23,3,21,22,24,19,20,16,17,9,14,18,25,15,13]

Forest plot of pooled specificity.TP: true positive; FP: false positive; FN: false negative; TN: true negative; CI: confidence intervalReferences [23,3,21,22,24,19,20,16,17,9,14,18,25,15,13]

Tables

Table 1. Characteristics and quality assessment of the included studies.QUADAS-2: Quality Assessment of Diagnostic Accuracy Studies-2

Author and Year	Country	Study Design	Population	Sample Size	Pneumothorax	Ultrasound Operator	QUADAS-2
Spampinato et al. (2023) [23]	Italy	Retrospective cohort	Adult	844	8	Emergency physician	Low risk
Singer et al. (2025) [3]	USA	Retrospective cohort	Adult	541	40	Emergency physician	Low risk
Quick et al. (2016) [21]	USA	Cross-sectional	Adult	149	19	Other	Low risk
Ronaldson et al. (2020) [22]	UK	Cross-sectional	Adult	12	3	Nurse paramedic	Low risk
Yates et al. (2017) [24]	USA	Cross-sectional	Adult	190	17	Nurse paramedic	Low risk
Oliver et al. (2019) [19]	UK	Cross-sectional	Adult and pediatric	361	98	Other	High risk
Press et al. (2014) [20]	USA	Cross-sectional	Adult	211	43	Nurse paramedic	Low risk
Ketelaars et al. (2013) [16]	Netherlands	Cross-sectional	Adult and pediatric	59	24	Other	High risk
Lyon et al. (2012) [17]	USA	Cross-sectional	Adult	64	28	Emergency physician	High risk
DeMasi et al. (2023) [9]	USA	Retrospective cohort	Adult	846	15	Resident	Low risk
Bar et al. (2021) [14]	Israel	Prospective cohort	Adult	85	46	Resident	Low risk
Mumtaz et al. (2016) [18]	Pakistan	Cross-sectional	Adult	46	42	Other	Low risk
Zhang et al. (2006) [25]	China	Prospective cohort	Adult	135	29	Emergency physician	Low risk
Karagöz et al. (2018) [15]	Turkey	Cross-sectional	Adult	166	99	Emergency physician	Low risk
Abbasi et al. (2013) [13]	Iran	Cross-sectional	Adult	146	37	Emergency physician	High risk

Table 2. Pooled diagnostic accuracy metrics of POCUS for pneumothorax detection.POCUS: point-of-care ultrasound; HSROC: hierarchical summary receiver operating characteristic

Parameter	Coefficient	Std. Error	95% Confidence Interval
Bivariate model
E(logitSe)	3.6258	0.6537	2.3446-4.9069
E(logitSp)	2.6758	0.5389	1.6195-3.7320
Var(logitSe)	2.6091	1.8534	0.6484-10.4989
Var(logitSp)	3.8818	1.6706	1.6700-9.0232
Corr(logits)	-0.3218	0.3504	-0.8004-0.4073
HSROC model
Lambda	6.4272	0.7223	5.0115-7.8429
Theta	0.7908	0.6555	-0.4940-2.0756
Beta	0.1986	0.4075	-0.6000-0.9973
s2 alpha	4.3166	2.5812	1.3370-13.9361
s2 theta	2.1033	1.1412	0.7262-6.0918
Summary point estimates
Sensitivity (Se)	0.9741	0.0165	0.9125-0.9927
Specificity (Sp)	0.9356	0.0325	0.8347-0.9767
Diagnostic odds ratio (DOR)	545.4	-	-
Positive LR (LR+)	15.1208	7.5771	5.6629-40.3751
Negative LR (LR-)	0.0277	0.0175	0.0081-0.0954
1/LR-	36.0702	22.7498	10.4783-124.1677
Covariance	-1.0241	-	-

Table 3. Subgroup meta-analysis of the diagnostic accuracy of point-of-care ultrasound for detecting pneumothorax.US: ultrasound; QUADAS-2: Quality Assessment of Diagnostic Accuracy Studies-2; eFAST: extended focused assessment with sonography for trauma; CI: confidence interval

Subgroup Variable	Subgroup	N Studies	Sensitivity (95% CI)	I2	Specificity (95% CI)	I2
Study design	Retrospective cohort	3	0.730 (0.570-0.860)	0.412	0.987 (0.958-0.998)	0.210
	Cross-sectional	10	0.510 (0.390-0.630)		0.975 (0.952-0.990)
	Prospective cohort	2	0.845 (0.705-0.940)		0.981 (0.955-0.994)
Population	Adult	13	0.590 (0.470-0.700)	0.362	0.976 (0.960-0.989)	0.240
	Adult and pediatric	2	0.680 (0.510-0.800)		0.987 (0.969-0.996)
Patient condition	Trauma	8	0.550 (0.420-0.680)	0.501	0.977 (0.954-0.991)	0.430
	Non-trauma	7	0.885 (0.755-0.965)		0.992 (0.974-0.998)
US protocol	eFAST	6	0.650 (0.495-0.775)	0.298	0.988 (0.968-0.998)	0.395
	Other	4	0.540 (0.345-0.720)		0.977 (0.940-0.992)
	Two-point	4	0.780 (0.605-0.900)		0.985 (0.967-0.996)
US operator	Emergency physician	5	0.820 (0.680-0.900)	0.420	0.988 (0.970-0.998)	0.265
	Nurse/paramedic	3	0.410 (0.240-0.600)		0.990 (0.978-0.998)
	Resident	2	0.870 (0.700-0.955)		0.987 (0.960-0.997)
	Other	4	0.560 (0.385-0.720)		0.978 (0.955-0.992)
Operator experience	Trained	6	0.615 (0.492-0.725)	0.497	0.986 (0.968-0.997)	0.430
	Experienced	4	0.680 (0.505-0.810)		0.977 (0.951-0.990)
	Mixed	3	0.650 (0.450-0.810)		0.987 (0.969-0.997)
	Not specified	2	0.500 (0.290-0.715)		0.985 (0.968-0.997)
QUADAS-2 risk	Low risk	11	0.700 (0.555-0.820)	0.398	0.977 (0.951-0.989)	0.495
	High risk	4	0.610 (0.490-0.720)		0.987 (0.968-0.997)

Table 4. Sensitivity analysis (leave-one-out) for point-of-care ultrasound detection of pneumothorax.CI: confidence interval

Omitted Study	Sensitivity (95% CI)	P-value	Specificity (95% CI)	P-value
Overall Model	0.74 (0.55-0.87)		0.99 (0.98-0.99)
Spampinato et al. (2023)	0.75 (0.56-0.88)	0.00	0.99 (0.98-0.99)	0.00
Singer et al. (2025)	0.74 (0.55-0.87)	0.00	0.99 (0.98-0.99)	0.00
Quick et al. (2016)	0.73 (0.54-0.87)	0.00	0.99 (0.98-0.99)	0.00
Ronaldson et al. (2020)	0.74 (0.55-0.87)	0.00	0.99 (0.98-0.99)	0.00
Yates et al. (2017)	0.77 (0.60-0.89)	0.00	0.99 (0.98-0.99)	0.00
Oliver et al. (2019)	0.78 (0.62-0.89)	0.00	0.99 (0.98-0.99)	0.00
Press et al. (2014)	0.78 (0.62-0.89)	0.00	0.99 (0.98-0.99)	0.00
Ketelaars et al. (2013)	0.76 (0.58-0.88)	0.00	0.99 (0.98-0.99)	0.00
Lyon et al. (2012)	0.72 (0.53-0.86)	0.00	0.99 (0.98-0.99)	0.00
DeMasi et al. (2023)	0.73 (0.54-0.87)	0.00	0.99 (0.98-0.99)	0.00
Bar et al. (2021)	0.72 (0.53-0.86)	0.00	0.99 (0.98-0.99)	0.00
Mumtaz et al. (2016)	0.73 (0.54-0.87)	0.00	0.99 (0.98-0.99)	0.00
Zhang et al. (2006)	0.73 (0.54-0.86)	0.00	0.99 (0.98-0.99)	0.00
Karagöz et al. (2018)	0.72 (0.53-0.86)	0.00	0.99 (0.98-0.99)	0.00
Abbasi et al. (2013)	0.73 (0.54-0.86)	0.00	0.99 (0.98-0.99)	0.00

Full text

Introduction and background

Point-of-care ultrasound (POCUS) has emerged as an important bedside diagnostic tool in emergency and critical care medicine, allowing clinicians to rapidly obtain focused imaging information to guide immediate management decisions [1,2]. POCUS examinations are performed and interpreted directly by the treating clinician, enabling real-time clinical integration and repeat assessment where necessary [2].

Pneumothorax is a potentially life-threatening condition, particularly in trauma and critically ill patients [3]. Thoracic ultrasound has become an integral component of modern emergency assessment protocols [4,5]. POCUS offers a rapid and readily available method to evaluate for traumatic and non-traumatic pneumothorax at the bedside, particularly when clinical examination findings are subtle or confounded by concurrent injuries or mechanical ventilation [6,7]. Computed tomography (CT) remains the reference standard for diagnosing pneumothorax and can detect small or occult pneumothoraces not visible on chest radiography; however, CT is time-consuming, costly, and not always immediately available in unstable patients [7,8].

Systematic reviews and meta-analyses have demonstrated that thoracic ultrasound is more sensitive than supine chest radiography for detecting traumatic pneumothorax, while maintaining high specificity [8-10]. However, reported sensitivity varies depending on the clinical population, reference standard, and inclusion of CT-only (occult) pneumothoraces [3,9,11]. This diagnostic test accuracy systematic review is reported in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for Diagnostic Test Accuracy (PRISMA-DTA) guidance [12]. Diagnostic accuracy is further influenced by operator training, experience, and scanning protocol [13,14].

The role of POCUS has expanded beyond the emergency department into prehospital, aeromedical, and retrieval settings, where conventional imaging is limited and rapid decision-making is essential [15-22]. Given the increasing use of thoracic POCUS across diverse acute care environments, this systematic review and meta-analysis aims to evaluate its diagnostic accuracy for pneumothorax, summarising pooled sensitivity and specificity and exploring sources of heterogeneity. Recent emergency department evaluations and aeromedical feasibility studies further support the broad applicability of thoracic POCUS across acute care environments [23,24].

Review

Methods

Study Design

This systematic review and meta-analysis were conducted in accordance with PRISMA-DTA reporting principles for diagnostic test accuracy studies [12]. A review protocol was developed prior to study initiation.

Search Strategy

A comprehensive literature search was performed in PubMed, Scopus, Web of Science, and Google Scholar from database inception to December 31, 2025. Search terms included “point-of-care ultrasound,” “POCUS,” “thoracic ultrasound,” “pneumothorax,” “sensitivity,” and “specificity,” using Boolean operators and Medical Subject Headings (MeSH) where appropriate. Reference lists of relevant articles and systematic reviews were manually screened.

Eligibility Criteria

Inclusion criteria were patients of any age with suspected pneumothorax; thoracic POCUS as the index test; CT, chest radiography, or intraoperative findings as the reference standard; observational diagnostic accuracy studies; and sufficient data to derive true-positive, false-positive, true-negative, and false-negative results.

Exclusion criteria included case reports, editorials, narrative reviews, animal studies, studies with fewer than 10 participants, and studies conducted outside emergency or acute care settings.

Study Selection and Data Extraction

Two reviewers independently screened titles, abstracts, and full-text articles. Extracted data included study design, setting, sample size, patient population, pneumothorax prevalence, ultrasound protocol, operator type and training, reference standard, and diagnostic accuracy outcomes. Disagreements were resolved by consensus with a third reviewer.

Quality Assessment

Risk of bias was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 (QUADAS-2) tool, evaluating patient selection, index test conduct, reference standard, and flow and timing.

Statistical Analysis

Meta-analysis was performed using R (version 4.3.1; R Foundation for Statistical Computing, Vienna, Austria) with diagnostic accuracy modelling packages. Pooled sensitivity, specificity, likelihood ratios, and diagnostic odds ratios (ORs) were estimated using a bivariate random-effects model. Summary estimates were presented with 95% confidence intervals (CIs). Subgroup analyses explored the effects of clinical setting, ultrasound protocol, operator type, and study quality. Leave-one-out sensitivity analyses were conducted to assess the influence of individual studies.

Results

Study Selection

The literature search identified 277 records, of which 186 duplicates were removed. Ninety-one titles and abstracts were screened, and 48 full-text articles were assessed for eligibility. Fifteen studies met the inclusion criteria and were included in the final meta-analysis (Figure 1) [9,11,13-22,25].

PRISMA flow chart illustrating the study selection process for this systematic review and meta-analysis.PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses

Characteristics of Included Studies

Included studies were conducted across the United States, Europe, Asia, and the Middle East. Study designs included prospective and retrospective cohort studies as well as cross-sectional analyses. Sample sizes ranged from 12 to 846 participants. Ultrasound examinations were performed by physicians, residents, nurses, paramedics, and retrieval practitioners. Study quality varied across QUADAS-2 domains (Table 1).

Diagnostic Accuracy

Pooled diagnostic accuracy estimates for thoracic POCUS were as follows: sensitivity: 0.9741 (95% CI 0.9125-0.9927), specificity: 0.9356 (95% CI 0.8347-0.9767), positive likelihood ratio: 15.12, and negative likelihood ratio: 0.028. These findings indicate high overall diagnostic performance for bedside detection of pneumothorax using POCUS.

Detailed pooled estimates and hierarchical summary receiver operating characteristic (HSROC) parameters are presented in Table 2, and forest plots for sensitivity and specificity are shown in Figures 2-3, in accordance with PRISMA-DTA reporting standards.

Forest plot of pooled sensitivity.TP: true positive; FP: false positive; FN: false negative; TN: true negative; CI: confidence intervalReferences [23,3,21,22,24,19,20,16,17,9,14,18,25,15,13]

Forest plot of pooled specificity.TP: true positive; FP: false positive; FN: false negative; TN: true negative; CI: confidence intervalReferences [23,3,21,22,24,19,20,16,17,9,14,18,25,15,13]

Subgroup Analysis

Higher sensitivity was observed in non-trauma populations, in studies employing focused two-point lung ultrasound protocols, and among examinations performed by physicians or residents with formal ultrasound training [13,14,22]. Specificity remained consistently high across subgroups (Table 3).

Sensitivity Analysis

Leave-one-out analyses demonstrated that no single study disproportionately influenced pooled estimates, supporting the robustness of the findings (Table 4).

Discussions and clinical implications

This systematic review and meta-analysis demonstrate that thoracic POCUS has high pooled sensitivity and specificity for the detection of pneumothorax in acute care settings. These findings are consistent with previous systematic reviews showing superior sensitivity of ultrasound compared with supine chest radiography, while maintaining high specificity [8-10].

Diagnostic performance varied depending on clinical context and reference standard. Sensitivity may be lower in cohorts that include CT-detected occult pneumothoraces or patients with small pneumothoraces, whereas performance is higher for clinically significant cases [3,9,11]. Operator experience and structured training play a key role in diagnostic accuracy, particularly in recognising dynamic sonographic signs such as lung sliding and the lung point [13,14].

Thoracic POCUS is especially valuable in prehospital, aeromedical, and retrieval environments, where rapid decision-making is required, and access to radiography is limited [15-22,26]. Multiple studies demonstrate that diagnostic lung ultrasound can be feasibly performed during transport and may influence early management decisions.

Recent emergency department-based studies and aeromedical evaluations further support the feasibility and diagnostic utility of thoracic POCUS across diverse acute care environments, reinforcing its applicability beyond controlled trauma settings [23,24].

Limitations

This review is limited by heterogeneity in study design, patient populations, ultrasound protocols, and reference standards. Some included studies had small sample sizes or limited pneumothorax events, reducing statistical precision. Differences in equipment and operator training may also affect diagnostic performance.

Conclusions

Thoracic POCUS demonstrates high sensitivity and specificity for the detection of pneumothorax in emergency and acute care settings. It enables rapid bedside assessment, supports early clinical decision-making, and may reduce reliance on delayed imaging modalities. Standardised scanning protocols and structured training are essential to optimise diagnostic performance, particularly in prehospital and retrieval environments.