Diagnostic chest ultrasound for acute respiratory failure

Diagnostic chest ultrasound for acute respiratory failure

Accepted Manuscript Diagnostic chest ultrasound for acute respiratory failure Peter Wallbridge, Daniel Steinfort, Tunn Ren Tay, Louis Irving, Mark Hew...

11MB Sizes 0 Downloads 9 Views

Accepted Manuscript Diagnostic chest ultrasound for acute respiratory failure Peter Wallbridge, Daniel Steinfort, Tunn Ren Tay, Louis Irving, Mark Hew PII:

S0954-6111(18)30209-9

DOI:

10.1016/j.rmed.2018.06.018

Reference:

YRMED 5468

To appear in:

Respiratory Medicine

Received Date: 20 February 2018 Revised Date:

19 May 2018

Accepted Date: 18 June 2018

Please cite this article as: Wallbridge P, Steinfort D, Tay TR, Irving L, Hew M, Diagnostic chest ultrasound for acute respiratory failure, Respiratory Medicine (2018), doi: 10.1016/j.rmed.2018.06.018. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

Diagnostic chest ultrasound for acute respiratory failure.

1,2

SC

RI PT

Peter Wallbridge Daniel Steinfort1,2 Tunn Ren Tay3 Louis Irving1,2 Mark Hew2,4

TE D

M AN U

1. Department of Respiratory & Sleep Medicine, Royal Melbourne Hospital, Melbourne, Australia 2. Department of Medicine, Royal Melbourne Hospital, University of Melbourne, Melbourne, Australia 3. Department of Respiratory and Critical Care Medicine, Changi General Hospital, Singapore 4. School of Public Health & Preventive Medicine, Monash University, Melbourne, Australia

EP

Corresponding author: Peter Wallbridge, Department of Respiratory and Sleep Medicine, Royal Melbourne Hospital, Grattan Street, Parkville, Victoria, Australia 3050 [email protected]

AC C

Word Count: 5161

1

ACCEPTED MANUSCRIPT Acute respiratory failure (ARF) is a common medical emergency with increasing incidence.[1] ARF has significant healthcare impact: in-hospital mortality is 30.2%, average hospital stay is 7.1 days, and ARF-related healthcare cost in the US is estimated at US$54 billion annually.[2] In patients under 45, the commonest causes of ARF are pneumonia, acute respiratory distress syndrome (ARDS), sepsis, asthma, drug ingestion, and trauma. In the elderly, pneumonia, heart failure, chronic obstructive pulmonary disease (COPD), ARDS, and sepsis predominate.[1]

SC

RI PT

In ARF, early and accurate diagnosis of the underlying cause improves outcomes. For example, timely antibiotic administration in community-acquired pneumonia reduces mortality, making early diagnosis vital.[3, 4] Accurate diagnosis is contingent on reliable chest imaging, which is problematic in the acute setting. Radiography is often limited to portable supine films, with diminished diagnostic accuracy.[5, 6] ‘Gold-standard’ imaging techniques such as computed tomography (CT) confer the risks of transport,[7] and are associated with radiation exposure[8] with its attendant risks.[9, 10] Data suggest that accurate diagnosis within the emergency department is difficult to achieve based on commonly available initial investigations, which impacts management[11] and mortality.[12]

M AN U

Ultrasound for ARF is widely used by emergency and intensive care specialists,[13, 14] and uptake by respiratory physicians is increasing.[15] In contrast to chest CT, ultrasound is portable, safe and ̶ when performed by the treating clinician ̶, allows integration of findings with clinical information. With the increasing use of ultrasound at the point-of-care in various clinical settings, it is timely to review the efficacy of ultrasound in assisting ARF diagnosis. Importantly, this paper will also review the impact of ultrasound downstream from the diagnostic process.

TE D

Diagnostic imaging efficacy

The efficacy of medical imaging is far broader than diagnostic accuracy. Six levels of diagnostic imaging efficacy were originally described,[16] which we have recently grouped into three (Figure 1):[17]

AC C

EP

a) Test attributes – measuring: i) Technical fidelity ii) Diagnostic accuracy b) Clinician behaviour - how the test alters: iii) Clinical diagnoses iv) Treatment decisions c) Health outcomes - how the test impacts: v) Patient outcomes vi) Societal outcomes

The value of diagnostic ultrasound in ARF is realised only to the degree to which it impacts clinical outcomes, by altering clinician behaviour. (Figure 2) This paper will discuss to what degree ultrasound for various aetiologies of ARF is supported by the literature at each efficacy level, and propose future directions for research in the field. TEST ATTRIBUTES: TECHNICAL FIDELITY AND DIAGNOSTIC ACCURACY

2

ACCEPTED MANUSCRIPT The interpretation of ultrasound artefacts reflecting changes in lung parenchyma have been well described, as has its diagnostic accuracy for various conditions in multiple patient cohorts.[18] (Table 1) Several common conditions seen in acute respiratory failure are outlined below. Pneumothorax

RI PT

In the intensive care unit (ICU) and trauma setting (Table 1), ultrasound is sensitive and specific for the diagnosis of pneumothorax[19-22], and is clearly superior to chest X-Ray (CXR). There are also data showing accuracy of ultrasound for the diagnosis [23-25] and monitoring [26] of postprocedure pneumothoraces, although specific ARF data are lacking.

SC

In contrast to traumatic and iatrogenic pneumothorax, sonographic detection of spontaneous pneumothorax in ARF may be less reliable, due to lower prevalence (3-4% vs 13-30%).[27] Secondly, some conditions seen in ARF such as COPD may mimic pneumothorax via loss of lung sliding secondary to hyperinflation. [21]

M AN U

In patients with ARF in the ICU, the positive predictive value (PPV) of absent lung sliding for pneumothorax has varied between studies, ranging from 55% to 100%.[6, 28, 29] Conversely, the negative predictive value approaches 100% when lung sliding and B-lines are present in this population. The wide range of PPV is likely related to the comorbidities present in different cohorts and cannot be explained by prevalence alone. Xirouchaki and colleagues included patients with COPD[6] which is recognised to significantly reduce the PPV of absent lung sliding for detection of pneumothorax,[30] likely explaining the 55% PPV. Detection of lung-point increases specificity, although is estimated to occur in only 66% of pneumothoraces.[31] (Figure 2)

TE D

Therefore, the main drawback of ultrasound for detection of pneumothorax in ARF outside of trauma and post-procedural settings is its limited PPV (unless lung point is demonstrated). The consequence of a false positive diagnosis is considerable, as it may lead to inappropriate chest tube insertion.

AC C

Pleural effusion

EP

It can be concluded that ultrasound is a sensitive tool for diagnosis of pneumothorax, although given the typically low prevalence in patients presenting with ARF, may be limited by a low positive predictive value.

Pleural effusion is characterised by a hypoechoic space between the parietal and visceral pleura[32] (Figure 3) and fluid characteristics may point to underlying aetiology.[33-35] Systematic reviews reported that ultrasound has greater sensitivity than CXR for diagnosing pleural effusion.[36, 37] (Table 1) These reviews included studies of patients with critical illness or ARDS (including intubated patients), suggesting that these test characteristics are representative of patients with ARF. The detection of pleural effusion was included in study protocols of ARF in ICU, most commonly considered in association with consolidation or pulmonary oedema.[27-29] Diagnostic characteristics for pleural effusion alone were not reported, however, effusions are typically seen in association with other pathologies in ARF[38-40] and pleural effusion as the sole aetiology for ARF is probably uncommon.[27]

3

ACCEPTED MANUSCRIPT We conclude that ultrasound is highly accurate for diagnosing pleural effusion, although specific data for ARF are lacking. Consolidation

RI PT

Alveolar consolidation is identified on ultrasound by increased echodensity of lung tissue, loss of the bright pleural line, air-bronchograms, hypoechoic vascular structures, and a serrated distal margin.[18] (Figure 4) Pneumonia

SC

The presence of US consolidation most commonly reflects infectious pneumonia, however other differentials include ARDS, lung contusion, organising pneumonia or pulmonary infarction. The aetiological pre-test probability is dependent on clinical context, which is usually known to the pointof-care sonographer. The diagnostic accuracy of ultrasound referenced to CT for detecting consolidation (regardless of aetiology) in ARF has been specifically examined in a systematic review and exceeds that of CXR. [41] (Table 1) Data limited to the ICU reveals similar results,[42, 43] with the presence of dynamic air-bronchograms particularly improving specificity.[44]

M AN U

When considering clinical diagnosis of pneumonia rather than sonographic consolidation in patients with and without ARF, ultrasound sensitivity is slightly reduced. [45-47] Atelectasis

TE D

There are variable appearances of atelectatic lung on ultrasound varying from B-lines to consolidation with air bronchograms, thought to reflect the degree of collapse.[48] There are data to suggest that lung recruitment strategies can be guided by ultrasound,[48, 49] although the impact on outcomes are unknown. We conclude that ultrasound is accurate for the detection of consolidation in acute respiratory failure, with atelectasis having a variable ultrasound appearance. Pulmonary embolism

EP

Some forms of peripheral consolidation-- specifically wedge-shaped or rounded— have been reported to indicate pulmonary infarction due to pulmonary embolus (PE).

AC C

Studies examining this sonographic finding have focused on populations with a high pre-test probability of PE. Systematic reviews [50, 51] report inferior sensitivity to gold-standard CT pulmonary angiography (CTPA) (Table 1), with one reporting a false negative rate of 11% in patients with a high pre-test probability of PE.[51] Given the grave consequences of a false negative result, chest ultrasound alone is not recommended in place of CTPA, although it may suggest alternate diagnoses in this population.[52] Combining chest ultrasound with lower limb venous Doppler and targeted echocardiography improves accuracy and may be a reasonable diagnostic pathway in patients where CTPA is contraindicated.[53] Findings suggestive of PE in the absence of other pathology in acute respiratory failure may alert the treating clinician to the diagnosis and assist in risk stratification.[53] The addition of focused echocardiography and lower limb Doppler ultrasound improves US specificity, and a negative multi-organ examination with an alternative diagnosis suggested has high NPV, and may obviate the need for CT imaging.[53]

4

ACCEPTED MANUSCRIPT None of the studies pooled in the systematic reviews specifically examined patients in ARF, although ARF may have been present based on inclusion criteria. Additionally, early studies used single-slice CT technology,[54] with lower sensitivity for diagnosis of PE compared to modern multi-slice scanners.[55] We conclude that the accuracy of ultrasound for diagnosing pulmonary embolus in ARF remains unclear and requires further study.

RI PT

Interstitial syndrome

The ultrasonographic interstitial syndrome is an umbrella term, encompassing diffuse and localised interstitial changes characterised by increased B-lines [56] through a mechanism that is not wellestablished.[57] (Figure 5)

M AN U

SC

Focal B-lines may represent pulmonary contusion[58], focal fibrosis (such as post radiotherapy), pneumonia[29, 59] or atelectasis.[48] Diffuse B-lines may be caused by increased lung water or diffuse pulmonary fibrosis. B-lines have been shown to correlate with CT-defined lung-weight[60] and brain natriuretic peptide in pulmonary oedema,[61, 62] and US has been used to determine extra-vascular lung water volume.[63-65]

TE D

Different methods for measurement and quantification of B-lines have been described, and definitions used impact on the sensitivity and specificity of ultrasound for detection of interstitial changes. Clear definitions of B-lines are often not used, which makes comparison of studies difficult. Anderson and colleagues prospectively assessed the intraclass correlation of different methods of Bline identification in patients suspected to have cardiogenic pulmonary oedema, reporting ICC coefficients of greater than 0.8 regardless of method utilised,[66] suggesting method may ultimately not be important. In patients with clinically-defined ARDS admitted to the ICU, the inter-observer agreement for USdetected interstitial syndrome was reported as 0.74,[43] suggesting US is reliable in this setting.

EP

Cardiogenic pulmonary oedema

AC C

Increased B-lines are sensitive and specific for diagnosis of cardiogenic pulmonary oedema(CPE) in breathless patients presenting to hospital, or in hospital with high pre-test probability of CPE.[67, 68] (Table 1) There is however limited data regarding differentiation between pulmonary fibrosis[69-73] and pulmonary oedema with significant overlap on ultrasound.[38] In contrast to the high accuracy in breathless patients in emergency settings, the accuracy of B-lines in patients with ARF is lower. In a study of 136 patients with ARF of whom 14% were intubated, the sensitivity and specificity of B-lines to diagnose pulmonary oedema were only 63% and 74% respectively. [29] The poor sensitivity may reflect treatment effects, as B-lines have been shown to change in real-time with fluid-shifts[64]and treatment for cardiac failure.[74] The reported PPV of only 42% was likely a result of the low cohort prevalence of cardiac failure, with a high proportion of pneumonia. ARDS

5

ACCEPTED MANUSCRIPT ARDS may be differentiated from cardiogenic pulmonary oedema on the basis of inhomogeneous interstitial pattern, pleural line changes and presence of lung consolidation.[38] Data suggest that these changes are dynamic, responsive to changes in ventilatory parameters,[48] with US interstitial syndrome diagnostic accuracy of 95% when compared to CT-defined ARDS.[43] Interstitial lung disease

RI PT

Findings such as an irregular pleural line, sub-pleural cysts and reduced lung sliding have been reported to occur in interstitial lung disease.[69, 75] The distance between B-lines may reflect underlying patient characteristics and parenchymal processes[76] and the presence of B-lines correlates with CT fibrosis scores.[77] The diagnostic accuracy of ultrasound for ILD in the context of ARF is unknown.

SC

We conclude that the accuracy of lung ultrasound diagnosis of pulmonary oedema in acute respiratory failure is limited. Ultrasound appears accurate for the diagnosis of ARDS, with insufficient data on the diagnosis of interstitial lung disease in ARF to establish accuracy.

M AN U

Diaphragm assessment

The role of ultrasound in musculoskeletal assessment is well established, with increasing interest in the evaluation of the diaphragm both in acute and chronic respiratory diseases, as well as neuromuscular conditions. Generally, ultrasound measurements of the diaphragm are structural (thickness) and functional (excursion and respiratory variation in thickness). Additionally, the position of each hemidiaphragm relative to the other can be assessed. There are multiple causes of diaphragm dysfunction,[78] each resulting in a distinctive ultrasound appearance.[79] (Figure 6)

EP

TE D

Reduced diaphragm thickness indicates atrophy, which can develop rapidly once mechanical ventilation is commenced.[80] Additionally, the ratio of change in thickness during inspiration (thickening ratio) discriminates between paralysed and normally functioning diaphragms accurately,[81] with strong correlation between the thickening ratio and the maximal generated inspiratory pressure.[82] Ultrasound may assist in predicting weaning outcomes in mechanically ventilated patients,[83-87] performing best in subpopulations with a high pre-test probability of dysfunction.[87]

AC C

The diagnostic utility of ultrasound diaphragm assessment in patients with ARF is unknown, and although the presence of diaphragm dysfunction in the absence of alternate explanations for ARF may alter the diagnostic approach, specific data is lacking. IMPACT ON DIAGNOSTIC THINKING AND THERAPEUTIC EFFICACY OF COMBINED ULTRASOUND ASSESSMENT: CLINICIAN BENEFITS Although ultrasound is accurate when compared to a reference standard, a more clinically meaningful measure is whether it can aid diagnostic thinking and therefore influence therapeutic decisions. (Figure 1). Several studies show that multimodality ultrasound in ARF influences clinician behaviour on both these levels (Table 2).

6

ACCEPTED MANUSCRIPT Ultrasound on admission to the intensive care unit

RI PT

In patients admitted to ICU with ARF, multiple diagnoses may coexist and differentiation may be difficult. Diagnostic confidence may be low for certain presentations.[88] Improved diagnostic sensitivity for common aetiologies may assist clinicians both with regards to decision making and diagnostic confidence. There are several studies assessing the diagnostic accuracy of multimodality US in patients with ARF which have shown improved diagnostic accuracy when compared to a standard approach, although have largely been limited to patients with single diagnoses within specific diagnostic categories, leaving the question of performance of ultrasound in a broader population with ARF open.

M AN U

SC

In the landmark ’Blue Protocol’, lung and deep vein ultrasound were performed on 301 ICU patients with ARF, 35 (12 %) of whom were intubated.[27]. Within 20 minutes of ICU admission Lichtenstein or Meziere scanned 12 lung segments, the internal jugular, subclavian, and ileo-femoro-popliteal veins. Based on ultrasound findings, patients were categorised to have one of the following; pulmonary oedema, pneumonia, pulmonary embolus, pneumothorax, or airways disease. Ultrasound diagnoses were then compared to final ICU discharge diagnoses. The results were impressive. Based on clinical assessment 76 % of patients received a correct diagnosis referenced to the final ICU diagnosis. The accuracy of ultrasound diagnosis was 90.5 %, an absolute increase of 15 %. This shows the potential of ultrasound to change diagnosis and alter management. US was most helpful in the detection of consolidation and interstitial oedema, often in the setting of underlying airways disease.

EP

TE D

There were some limitations to these data, now more than ten years old. Firstly, although not involved in direct patient management, ultrasonographers were not blinded to clinical information. Those who did not have a clear US pattern on the background of any chronic respiratory disease other than asthma, or with ‘simple bronchial superinfection’ were considered to have COPD, with authors reporting COPD was confirmed by ‘functional tests’. Patients with multiple or no clear ICU discharge diagnosis, or rare diagnoses (frequency of <2%) were excluded. Inclusion of these 41 patients would have reduced the diagnostic accuracy from 90% to 78%.

AC C

Secondly, ARDS did not form a final diagnostic category. This is unusual, because ARDS is considered a common cause of ARF,[1] and may be responsible for up to two-thirds of cases of ARF with pulmonary infiltrates.[89] In a recent randomised study of patients with ARF admitted to ICU, 79% had bilateral infiltrates on CXR.[90] Importantly, sonographic findings of ARDS overlap with both pulmonary oedema and pneumonia. In Lichtenstein’s study, seven patients with “beginning ARDS” were included, but placed in the pneumonia category. Another ten patients with a discharge diagnosis of pneumonia and pulmonary oedema were excluded, some of whom may have had ARDS. Thirdly, based on autopsies, the accuracy of final clinical diagnoses in critically ill patients is only approximately 80%, so the reference standard in this study is necessarily imperfect.[91] In a subsequent study, Silva et al added focused echocardiography to lung and deep vein ultrasound to evaluate 78 consecutive patients admitted to ICU with ARF.[28] Three patients (4%) were excluded for multiple discharge diagnoses. Scanning by one of two experts was performed on admission and took mean duration of only 12 minutes. A third of patients were intubated within 8

7

ACCEPTED MANUSCRIPT hours of admission, suggesting severe respiratory failure in this cohort. The same five diagnostic categories from Lichtenstein’s study were used. Compared to final diagnosis, diagnostic accuracy was higher in patients undergoing combined ultrasound (83% vs 63%). (Table 2).

RI PT

The authors suggest that ultrasound could have improved initial treatment in patients with pneumonia, pulmonary oedema, COPD and PE, although the rates of appropriate therapy were not reported. Similar caveats apply with regards to exclusion of those with multiple diagnoses, which occurred at a surprisingly low frequency in this cohort. It is unclear whether the sonographer was blinded, and again, ARDS was not included as a diagnostic category.

M AN U

SC

In a follow-up study from the same group, Bataille and colleagues compared [29] the diagnostic accuracy of lung ultrasound, with and without focused echocardiography, referenced to final diagnosis. Nineteen (14%) of 136 included patients were intubated. Patients were scanned after a mean of 16 minutes from admission, and ultrasound took 9 minutes. 55% of patients also underwent CT scanning, and 26% had formal echocardiography during their admission. The authors found the addition of focused echocardiography to lung ultrasound increased diagnostic accuracy from 63% to 81%. Several factors may have increased reported diagnostic accuracy. Patients were assigned only one of four possible diagnoses – cardiogenic pulmonary oedema, pneumonia, pulmonary embolism, or pneumothorax. Four patients with COPD were excluded, as were an undisclosed number with multiple final diagnoses. No cases of ARDS were diagnosed.

TE D

In this cohort, the finding of B-lines within a lung region had poor sensitivity and specificity for cardiogenic pulmonary oedema, being absent in 37% of cases, and detectable in 33% of pneumonia cases. The definition of B-lines was not particularly restrictive, which may explain the high number of patients with pneumonia considered to have a B-line pattern, although the low sensitivity is more difficult to explain.

AC C

EP

In all three studies, the applicability of the results to clinical practice is limited by the exclusion of those with multiple diagnoses, which is commonly encountered in intensive care.[2] Additionally, excluding COPD patients may alter the specificity of ultrasound for pneumothorax; bullous disease and hyperinflation resemble pneumothorax on sonography.[30] It is also remarkable that none of the studies reported ARDS, so results may not be applicable in cohorts where this is prevalent. Additionally, lung ultrasound alone, without concurrent deep venous or cardiac examination, is accurate in only 63%.[29] Finally, all these data were obtained in centres with expert operators, with scans performed by small numbers of highly skilled practitioners with extensive ultrasound experience. The impact of experience on accuracy is not well described, although ultrasonographer experience has previously been suggested to explain study heterogeneity in systematic reviews.[19] With these caveats, the use of lung ultrasound as an initial test in ARF, when supplemented by venous ultrasound and focused cardiac examination, clearly increases the diagnostic accuracy for the underlying cause. There are less data focusing on the therapeutic impact of ultrasound performed on admission in the intensive care unit.

8

ACCEPTED MANUSCRIPT

RI PT

Manno and colleagues, performed a prospective study of non-blinded multi-organ US performed within 12 hours of admission in the ICU, which included a proportion with ARF.[92] In this cohort of 125 patients, admission diagnosis was modified in 25.2%, management changed in 39.2% and further testing was arranged in 18.4%. Diagnostic accuracy was not reported, although where goldstandard imaging was performed, US false negative rate was 4.9%. These data suggest that ultrasound on admission may alter management, although the lack of blinding requires consideration. It is also worth noting that ultrasound was particularly helpful in those with multiple organ failure, with the specific impact in ARF not described.

Targeted testing in the intensive care unit

M AN U

SC

A different approach, taken among ventilated ARF patients already in the ICU by Xirouchaki et al was to deploy targeted lung ultrasound only when a clinical question arose. A single experienced operator performed 253 lung scans in response to specific questions relating to five pre-specified diagnoses (pneumothorax, pleural effusion, unilateral lobar or total lung consolidation/atelectasis, pneumonia or diffuse cardiogenic or non-cardiogenic pulmonary oedema) or unexplained deterioration in gas exchange.[5] Ultrasound performed for a specific question confirmed the suspected diagnosis in approximately 70%, with a single diagnosis not supported in 23%, with findings such as bilateral effusions, basal consolidation and diffuse B-lines identified. An alternate (unsuspected) single diagnosis was suggested in 7%.

EP

TE D

Of relevance to this review, whenever ultrasound was performed for an unexplained deterioration in gas exchange in a mechanically ventilated patient (worsening of respiratory failure), ultrasound findings consistent with a specific process were identified in 40%, and changed management in every case. Non-specific findings of dependent lung consolidations, diffuse B lines and bilateral pleural effusions were identified in the remaining 60%, which were thought to be related to ventilation itself.

AC C

Overall, ultrasound results directly changed patient management in 47% of cases in this cohort, of which two thirds were ‘invasive’ interventions (bronchoscopy, chest tube insertion and thoracentesis) and one third ‘non-invasive’ (medication, ventilator strategy or positioning). Management was especially influenced when pneumothorax or pulmonary embolism were suggested by ultrasound (all cases), with less impact for pulmonary oedema (44% change in management) or atelectasis (60%). The outcomes of these management changes were not reported, although all deemed appropriate when retrospectively reviewed. A final point in this study was that 252/253 of all scans showed findings of consolidation, interstitial syndrome or an effusion, leading the authors to conclude that routine and indiscriminate scanning in the ICU would be of limited benefit due to the high prevalence of non-specific findings related to mechanical ventilation. The clinical context is therefore paramount. Initial testing in the emergency department

9

ACCEPTED MANUSCRIPT Laursen and colleagues performed a prospective randomised trial where multimodality point of care ultrasound (including lung ultrasound) was undertaken in patients presenting with respiratory symptoms to the emergency department.[93] This study randomised half of 320 patients with breathlessness, desaturation, chest pain or cough, to ultrasound of the heart, lung and deep veins. The sonographer was not blinded, and results were provided to the treating clinician to impact management and diagnosis. Patients who could not be scanned within an hour of presentation were excluded.

SC

RI PT

The primary outcome was the proportion with a correct diagnosis at four hours, referenced to subsequent blinded panel audit diagnosis, and this was superior in the ultrasound group (88%) compared to control (63.7%). This accelerated treatment, with 78% of the ultrasound group commencing appropriate therapy within 4 hours (control 57%). In-hospital mortality was similar between groups, with unchanged length of stay, although this study was not powered to look at such outcomes.

M AN U

In particular, when compared to control subjects, at four hours ultrasound improved the detection of pleural effusion (89% vs 14%), systolic heart failure (91% vs 45%), pulmonary oedema (79% vs 38%) and malignancy (80% vs 38%). Timeliness of diagnosis of empyema and parapneumonic effusion was also improved. Patients who had ultrasound were more likely to have further tests ordered within four hours, suggesting that the diagnostic process is hastened. An unknown proportion of patients presumably had ARF, as inclusion criteria included patients started on oxygen or who were tachypnoeic. However, the applicability of this study to a broader range of ARF patients in emergency departments is unknown.

TE D

In summary, there are high quality data in the emergency department to support the use of expertperformed US in patients with ARF with improved diagnostic accuracy when compared to standard clinical assessment. There are data to suggesting benefit with use of US within the ICU, although randomised data are lacking.

EP

Initial testing in respiratory wards

Very limited data exist for ultrasound in ARF outside the ICU.

AC C

Our group has performed a prospective cohort study in patients admitted to a high dependency respiratory ward with ARF, performing lung ultrasound within 24h of presentation in addition to standard clinical workup. Ultrasound was performed by one of five respiratory specialists and trainees, rather than a single expert operator. In this cohort of elderly patients with multi-morbidity and often multiple underlying diagnoses, clinician management was changed in 30% with improved clinician diagnostic confidence in 44% of patients.[94] Of the 50 included patients, 34 had a single admission diagnosis, with 16 having multiple diagnoses. Initial single diagnoses most commonly included pneumonia, exacerbations of asthma or COPD, or decompensated cardiac failure (left or right ventricular failure). Multiple diagnoses typically involved left and/or right heart failure with pneumonia. Unsurprisingly there was only a single case of ARDS in this less critically ill cohort, and there were no pneumothoraces. Ultrasound confirmed the clinical diagnosis in 66%, and improved treating clinician confidence in more than 50% when this was the

10

ACCEPTED MANUSCRIPT case. Overall clinical diagnosis was changed in 10%, with additional diagnoses added in the remaining 24% of cases.

IMPACT ON HEALTH OUTCOMES: PATIENT AND SOCIETAL BENEFIT

RI PT

The major limitation to this study was the lack of a gold-standard diagnosis. This limits validity, although the inclusion of patients with multiple clinical diagnoses not admitted to an ICU suggests that ultrasound may be able to be used in a broader context in those with ARF than previously studied. Additionally, echocardiography was also not performed.

SC

In patients with ARF, although authors suggest that management would have been changed with the increased diagnostic accuracy over standard approaches[28], there are not yet data to show that immediate patient outcomes such as intubation rates, treatment failure, in-hospital mortality or inhospital complications or longer-term patient outcomes such as long-term mortality or impairment in quality of life are improved. Additionally, whether there is a reduction in harm from incorrect or un-necessary procedures and/or treatments remains to be seen.

M AN U

There are some data showing that use of lung ultrasound in the intensive care unit reduces the number of chest x-rays performed, with associated reductions in cost and radiation, although rates of CT scanning are no different. [95, 96] This occurs on a background of general decrease in rates of utilisation of advanced imaging techniques in the ICU, with a significant increase in uptake of US.[97] Obviously, the effect of US introduction on the use of other imaging modalities does not occur in isolation and is highly dependent on the resource and medicolegal aspects of a specific practice setting.

EP

TE D

Whether the use of thoracic ultrasound ultimately impacts resource utilisation more broadly such as reducing bed-days both for high dependency areas and overall length of stay and therefore societal outcomes such as total healthcare expenditure is unknown,[17] although radiation risks associated with standard imaging are well described,[10] with any reduction in overall exposure likely to be beneficial. The use of thoracic ultrasound may be particularly beneficial in resource-limited settings, where access to advanced imaging is difficult,[98, 99] including in those with acute respiratory failure.[100]

AC C

Concluding Discussion

Unanswered Questions

The above data show the potential role for thoracic ultrasound to augment diagnostic accuracy and possibly therapeutic efficacy in patients with ARF. There are however several important limitations and issues that have not been addressed by current literature. Significant gaps remain in our understanding of the role of ultrasound in ARF, firstly relating to factors that may affect accuracy, and secondly, whether patients and health care systems are likely to benefit from its use. Precise patient selection will be crucial in determining the practical utility of ultrasound in ARF. Many of the published studies have focused on clinical situations where the likelihood of the target condition(s) being present is high, and whether these findings are applicable in other populations and settings is unknown (context bias). Secondly, studies recruiting patients with severe respiratory failure (requiring mechanical ventilation and/or ICU admission) introduce a form of spectrum bias,

11

ACCEPTED MANUSCRIPT whereby only those with severe disease are included. This may increase the detection of lung changes on ultrasound. Thirdly, patients with multiple aetiologies, which can occur in up to 47% of elderly patients with ARF,[12] have been poorly studied. The characteristics and results of excluded patients in the ICU studies were also not well described. Some of these deficits may be addressed by well-designed RCTs of patients with ARF, which is the next step for lung ultrasound research. (Figure 2))

M AN U

SC

RI PT

Another practical consideration is whether ultrasound can be performed in a timely fashion as part of a standard clinical workload. The fact that less than half eligible patients in our cohort study[94] were able to be scanned within a generous 24 hour timeframe suggests that practice patterns outside of the ICU may need to be substantially altered to enable integration of US to be maximally effective. Studies of multimodality ultrasound in ICU are reported as being performed ‘on admission’, albeit by a clinician not directly involved in patient care. The timing of ultrasound examination is important, as for certain pathologies (particularly interstitial syndromes), imaging results rapidly change [64] and the time-lapse between the initial clinical assessment and/or goldstandard chest imaging may have impact on the reported sensitivity and specificity of lung ultrasound examination.

TE D

Ultrasound is undeniably an operator-dependent imaging modality. All published series have used a single operator (or small group of operators) with significant ultrasound expertise. In the pneumothorax literature, some of the heterogeneity between studies was related to operator experience,[19] and therefore the clinical applicability of these results should be interpreted cautiously. Additionally, the specific protocol and method of analysis utilised may impact on results. In our own systematic review of consolidation,[41] using lung regions as opposed to lungs as the primary unit of analysis increased specificity and reduced sensitivity, and also inflated the apparent precision of ultrasound.

AC C

EP

Finally, the diagnostic role of ultrasound in clinical practice needs to be considered – how and where does US fit with current investigations such as x-ray and CT? It is yet to be determined whether US should replace or supplement standard investigations in adult populations, although in paediatric populations where pneumonia is suspected, US appears able to replace CXR.[101] The utility of ultrasound as an add-on investigation in conjunction with a panel of tests such as validated biomarkers by a single clinician is likely where the value of ultrasound will ultimately lie. This is a strength of the study by Laursen et al in the emergency department,[93] and we hope this approach will be replicated in future. Such ‘real-life’ implementation of ultrasound more accurately reflects clinical practice, and studies utilising this approach are likely to be more applicable for clinicians. Future directions

Ideally before a clinical test is integrated into practice, studies should confirm that it can impact patient health and outcomes, rather than simply demonstrating accuracy.[102] A carefully designed randomised trial of ultrasound in acute respiratory failure assessing diagnostic decision making, therapeutic impact and patient outcomes is therefore the next step. Including patients with multiple underlying aetiologies will improve clinical relevance, and extending these studies to include patients outside the ICU will increase applicability. The use of hard

12

ACCEPTED MANUSCRIPT endpoints focusing on both patient and societal outcomes will ensure that performing ultrasound examinations in the critically ill truly improves patient care. Final Comments

RI PT

Chest ultrasound is an accurate tool for the diagnosis of conditions commonly underlying ARF, and has been shown to impact diagnostic reasoning in critical care settings and expedite appropriate management in the emergency department when combined with standard diagnostic approaches. There is a role for well-designed randomised controlled trials of chest ultrasound in patients with acute respiratory failure to confirm definite benefit with regards to patient and societal outcomes, and to direct the optimal implementation of chest ultrasound for this indication.

AC C

EP

TE D

M AN U

SC

Competing Interests: None to declare

13

ACCEPTED MANUSCRIPT REFERENCES

7. 8. 9. 10.

11.

12. 13.

14.

15. 16. 17. 18. 19. 20.

RI PT

6.

SC

5.

M AN U

4.

TE D

3.

EP

2.

Stefan, M.S., et al., Epidemiology and outcomes of acute respiratory failure in the United States, 2001 to 2009: a national survey. J Hosp Med, 2013. 8(2): p. 76-82. Walkey, A.J. and R.S. Wiener, Use of noninvasive ventilation in patients with acute respiratory failure, 2000-2009: a population-based study. Ann Am Thorac Soc, 2013. 10(1): p. 10-7. Mandell, L.A., et al., Infectious Diseases Society of America/American Thoracic Society consensus guidelines on the management of community-acquired pneumonia in adults. Clin Infect Dis, 2007. 44 Suppl 2: p. S27-72. Kumar, A., et al., The duration of hypotension before the initiation of antibiotic treatment is a critical determinant of survival in a murine model of Escherichia coli septic shock: association with serum lactate and inflammatory cytokine levels. J Infect Dis, 2006. 193(2): p. 251-8. Xirouchaki, N., et al., Impact of lung ultrasound on clinical decision making in critically ill patients. Intensive Care Med, 2014. 40(1): p. 57-65. Xirouchaki, N., et al., Lung ultrasound in critically ill patients: comparison with bedside chest radiography. Intensive Care Med, 2011. 37(9): p. 1488-93. Waydhas, C., Intrahospital transport of critically ill patients. Crit Care, 1999. 3(5): p. R83-9. Slovis, B.H., et al., Significant but reasonable radiation exposure from computed tomography-related medical imaging in the ICU. Emerg Radiol, 2016. 23(2): p. 141-6. Mayo, J.R., J. Aldrich, and N.L. Muller, Radiation exposure at chest CT: a statement of the Fleischner Society. Radiology, 2003. 228(1): p. 15-21. Smith-Bindman, R., et al., Radiation dose associated with common computed tomography examinations and the associated lifetime attributable risk of cancer. Arch Intern Med, 2009. 169(22): p. 2078-86. Laursen, C.B., et al., Focused sonography of the heart, lungs, and deep veins identifies missed life-threatening conditions in admitted patients with acute respiratory symptoms. Chest, 2013. 144(6): p. 1868-1875. Ray, P., et al., Acute respiratory failure in the elderly: etiology, emergency diagnosis and prognosis. Crit Care, 2006. 10(3): p. R82. Price, S., et al., Expert consensus document: Echocardiography and lung ultrasonography for the assessment and management of acute heart failure. Nat Rev Cardiol, 2017. 14(7): p. 427440. Rybicki, F.J., et al., 2015 ACR/ACC/AHA/AATS/ACEP/ASNC/NASCI/SAEM/SCCT/SCMR/SCPC/SNMMI/STR/STS Appropriate Utilization of Cardiovascular Imaging in Emergency Department Patients With Chest Pain: A Joint Document of the American College of Radiology Appropriateness Criteria Committee and the American College of Cardiology Appropriate Use Criteria Task Force. J Am Coll Cardiol, 2016. 67(7): p. 853-79. Denton, E.J., L.M. Hannan, and M. Hew, Physician-performed chest ultrasound: progress and future directions. Intern Med J, 2017. 47(3): p. 306-311. Fryback, D.G. and J.R. Thornbury, The efficacy of diagnostic imaging. Med Decis Making, 1991. 11(2): p. 88-94. Hew, M. and T.R. Tay, The efficacy of bedside chest ultrasound: from accuracy to outcomes. Eur Respir Rev, 2016. 25(141): p. 230-46. Volpicelli, G., et al., International evidence-based recommendations for point-of-care lung ultrasound. Intensive Care Med, 2012. 38(4): p. 577-91. Ding, W., et al., Diagnosis of pneumothorax by radiography and ultrasonography: a metaanalysis. Chest, 2011. 140(4): p. 859-66. Alrajhi, K., M.Y. Woo, and C. Vaillancourt, Test characteristics of ultrasonography for the detection of pneumothorax: a systematic review and meta-analysis. Chest, 2012. 141(3): p. 703-8.

AC C

1.

14

ACCEPTED MANUSCRIPT

27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37.

38.

39. 40. 41.

42.

RI PT

26.

SC

25.

M AN U

24.

TE D

23.

EP

22.

Alrajab, S., et al., Pleural ultrasonography versus chest radiography for the diagnosis of pneumothorax: review of the literature and meta-analysis. Crit Care, 2013. 17(5): p. R208. Ebrahimi, A., et al., Diagnostic Accuracy of Chest Ultrasonography versus Chest Radiography for Identification of Pneumothorax: A Systematic Review and Meta-Analysis. Tanaffos, 2014. 13(4): p. 29-40. Shostak, E., et al., Bedside sonography for detection of postprocedure pneumothorax. J Ultrasound Med, 2013. 32(6): p. 1003-9. Kumar, S., et al., Role of ultrasonography in the diagnosis and management of pneumothorax following transbronchial lung biopsy. J Bronchology Interv Pulmonol, 2015. 22(1): p. 14-9. Vezzani, A., et al., Ultrasound localization of central vein catheter and detection of postprocedural pneumothorax: an alternative to chest radiography. Crit Care Med, 2010. 38(2): p. 533-8. Galbois, A., et al., Pleural ultrasound compared with chest radiographic detection of pneumothorax resolution after drainage. Chest, 2010. 138(3): p. 648-55. Lichtenstein, D.A. and G.A. Meziere, Relevance of lung ultrasound in the diagnosis of acute respiratory failure: the BLUE protocol. Chest, 2008. 134(1): p. 117-25. Silva, S., et al., Usefulness of cardiothoracic chest ultrasound in the management of acute respiratory failure in critical care practice. Chest, 2013. Bataille, B., et al., Integrated use of bedside lung ultrasound and echocardiography in acute respiratory failure: a prospective observational study in ICU. Chest, 2014. 146(6): p. 1586-93. Slater, A., et al., COPD can mimic the appearance of pneumothorax on thoracic ultrasound. Chest, 2006. 129(3): p. 545-50. Lichtenstein, D., et al., The "lung point": an ultrasound sign specific to pneumothorax. Intensive Care Med, 2000. 26(10): p. 1434-40. Lomas, D.J., S.G. Padley, and C.D. Flower, The sonographic appearances of pleural fluid. Br J Radiol, 1993. 66(787): p. 619-24. Qureshi, N.R., N.M. Rahman, and F.V. Gleeson, Thoracic ultrasound in the diagnosis of malignant pleural effusion. Thorax, 2009. 64(2): p. 139-43. Yang, P.C., et al., Value of sonography in determining the nature of pleural effusion: analysis of 320 cases. AJR Am J Roentgenol, 1992. 159(1): p. 29-33. Abramowitz, Y., et al., Pleural effusion: characterization with CT attenuation values and CT appearance. AJR Am J Roentgenol, 2009. 192(3): p. 618-23. Grimberg, A., et al., Diagnostic accuracy of sonography for pleural effusion: systematic review. Sao Paulo Med J, 2010. 128(2): p. 90-5. Yousefifard, M., et al., Screening Performance Characteristic of Ultrasonography and Radiography in Detection of Pleural Effusion; a Meta-Analysis. Emerg (Tehran), 2016. 4(1): p. 1-10. Copetti, R., G. Soldati, and P. Copetti, Chest sonography: a useful tool to differentiate acute cardiogenic pulmonary edema from acute respiratory distress syndrome. Cardiovasc Ultrasound, 2008. 6: p. 16. Kataoka, H. and S. Takada, The role of thoracic ultrasonography for evaluation of patients with decompensated chronic heart failure. J Am Coll Cardiol, 2000. 35(6): p. 1638-46. Dean, N.C., et al., Pleural Effusions at First ED Encounter Predict Worse Clinical Outcomes in Patients With Pneumonia. Chest, 2016. 149(6): p. 1509-15. Hew, M., et al., The diagnostic accuracy of chest ultrasound for CT-detected radiographic consolidation in hospitalised adults with acute respiratory failure: a systematic review. BMJ Open, 2015. 5(5): p. e007838. Lichtenstein, D.A., et al., Ultrasound diagnosis of alveolar consolidation in the critically ill. Intensive Care Med, 2004. 30(2): p. 276-81.

AC C

21.

15

ACCEPTED MANUSCRIPT

49. 50. 51. 52. 53. 54. 55.

56.

57. 58. 59. 60. 61. 62.

63. 64. 65.

RI PT

48.

SC

47.

M AN U

46.

TE D

45.

EP

44.

Lichtenstein, D., et al., Comparative diagnostic performances of auscultation, chest radiography, and lung ultrasonography in acute respiratory distress syndrome. Anesthesiology, 2004. 100(1): p. 9-15. Lichtenstein, D., G. Meziere, and J. Seitz, The dynamic air bronchogram. A lung ultrasound sign of alveolar consolidation ruling out atelectasis. Chest, 2009. 135(6): p. 1421-5. Hu, Q.J., et al., Diagnostic performance of lung ultrasound in the diagnosis of pneumonia: a bivariate meta-analysis. Int J Clin Exp Med, 2014. 7(1): p. 115-21. Chavez, M.A., et al., Lung ultrasound for the diagnosis of pneumonia in adults: a systematic review and meta-analysis. Respir Res, 2014. 15: p. 50. Ye, X., et al., Accuracy of Lung Ultrasonography versus Chest Radiography for the Diagnosis of Adult Community-Acquired Pneumonia: Review of the Literature and Meta-Analysis. PLoS One, 2015. 10(6): p. e0130066. Bouhemad, B., et al., Bedside ultrasound assessment of positive end-expiratory pressureinduced lung recruitment. American journal of respiratory and critical care medicine, 2011. 183(3): p. 341-7. Gardelli, G., et al., Using sonography to assess lung recruitment in patients with acute respiratory distress syndrome. Emerg Radiol, 2009. 16(3): p. 219-21. Squizzato, A., et al., Diagnostic accuracy of lung ultrasound for pulmonary embolism: a systematic review and meta-analysis. J Thromb Haemost, 2013. 11(7): p. 1269-78. Jiang, L., et al., Role of Transthoracic Lung Ultrasonography in the Diagnosis of Pulmonary Embolism: A Systematic Review and Meta-Analysis. PLoS One, 2015. 10(6): p. e0129909. Koenig, S., et al., Ultrasound assessment of pulmonary embolism in patients receiving CT pulmonary angiography. Chest, 2014. 145(4): p. 818-23. Nazerian, P., et al., Accuracy of point-of-care multiorgan ultrasonography for the diagnosis of pulmonary embolism. Chest, 2014. 145(5): p. 950-7. Mathis, G., et al., Thoracic ultrasound for diagnosing pulmonary embolism: a prospective multicenter study of 352 patients. Chest, 2005. 128(3): p. 1531-8. Carrier, M., et al., Subsegmental pulmonary embolism diagnosed by computed tomography: incidence and clinical implications. A systematic review and meta-analysis of the management outcome studies. J Thromb Haemost, 2010. 8(8): p. 1716-22. Lim, J.H., et al., Ring-down artifacts posterior to the right hemidiaphragm on abdominal sonography: sign of pulmonary parenchymal abnormalities. J Ultrasound Med, 1999. 18(6): p. 403-10. Soldati, G., R. Copetti, and S. Sher, Sonographic interstitial syndrome: the sound of lung water. J Ultrasound Med, 2009. 28(2): p. 163-74. Soldati, G., et al., Chest ultrasonography in lung contusion. Chest, 2006. 130(2): p. 533-8. Hew, M. and S. Heinze, Chest ultrasound in practice: a review of utility in the clinical setting. Intern Med J, 2012. 42(8): p. 856-65. Baldi, G., et al., Lung water assessment by lung ultrasonography in intensive care: a pilot study. Intensive Care Med, 2013. 39(1): p. 74-84. Gargani, L., et al., Ultrasound lung comets for the differential diagnosis of acute cardiogenic dyspnoea: a comparison with natriuretic peptides. Eur J Heart Fail, 2008. 10(1): p. 70-7. Cibinel, G.A., et al., Diagnostic accuracy and reproducibility of pleural and lung ultrasound in discriminating cardiogenic causes of acute dyspnea in the emergency department. Intern Emerg Med, 2012. 7(1): p. 65-70. Volpicelli, G., et al., Lung ultrasound predicts well extravascular lung water but is of limited usefulness in the prediction of wedge pressure. Anesthesiology, 2014. 121(2): p. 320-7. Noble, V.E., et al., Ultrasound assessment for extravascular lung water in patients undergoing hemodialysis. Time course for resolution. Chest, 2009. 135(6): p. 1433-9. Jambrik, Z., et al., Usefulness of ultrasound lung comets as a nonradiologic sign of extravascular lung water. Am J Cardiol, 2004. 93(10): p. 1265-70.

AC C

43.

16

ACCEPTED MANUSCRIPT

72.

73.

74. 75. 76. 77. 78. 79. 80. 81. 82. 83.

84. 85.

86. 87.

RI PT

71.

SC

70.

M AN U

69.

TE D

68.

EP

67.

Anderson, K.L., et al., Inter-rater reliability of quantifying pleural B-lines using multiple counting methods. J Ultrasound Med, 2013. 32(1): p. 115-20. Al Deeb, M., et al., Point-of-care ultrasonography for the diagnosis of acute cardiogenic pulmonary edema in patients presenting with acute dyspnea: a systematic review and metaanalysis. Acad Emerg Med, 2014. 21(8): p. 843-52. Martindale, J.L., et al., Diagnosing Acute Heart Failure in the Emergency Department: A Systematic Review and Meta-analysis. Acad Emerg Med, 2016. 23(3): p. 223-42. Sperandeo, M., et al., Transthoracic ultrasound in the evaluation of pulmonary fibrosis: our experience. Ultrasound Med Biol, 2009. 35(5): p. 723-9. Lo Giudice, V., et al., Ultrasound in the evaluation of interstitial pneumonia. J Ultrasound, 2008. 11(1): p. 30-8. Tardella, M., et al., Ultrasound in the assessment of pulmonary fibrosis in connective tissue disorders: correlation with high-resolution computed tomography. J Rheumatol, 2012. 39(8): p. 1641-7. Aghdashi, M., B. Broofeh, and A. Mohammadi, Diagnostic performances of high resolution trans-thoracic lung ultrasonography in pulmonary alveoli-interstitial involvement of rheumatoid lung disease. Int J Clin Exp Med, 2013. 6(7): p. 562-6. Mohammadi, A., S. Oshnoei, and M. Ghasemi-rad, Comparison of a new, modified lung ultrasonography technique with high-resolution CT in the diagnosis of the alveolo-interstitial syndrome of systemic scleroderma. Med Ultrason, 2014. 16(1): p. 27-31. Volpicelli, G., et al., Bedside ultrasound of the lung for the monitoring of acute decompensated heart failure. Am J Emerg Med, 2008. 26(5): p. 585-91. Reissig, A. and C. Kroegel, Transthoracic sonography of diffuse parenchymal lung disease: the role of comet tail artifacts. J Ultrasound Med, 2003. 22(2): p. 173-80. Hasan, A.A. and H.A. Makhlouf, B-lines: Transthoracic chest ultrasound signs useful in assessment of interstitial lung diseases. Ann Thorac Med, 2014. 9(2): p. 99-103. Gargani, L., et al., Ultrasound lung comets in systemic sclerosis: a chest sonography hallmark of pulmonary interstitial fibrosis. Rheumatology (Oxford), 2009. 48(11): p. 1382-7. Gerscovich, E.O., et al., Ultrasonographic evaluation of diaphragmatic motion. J Ultrasound Med, 2001. 20(6): p. 597-604. Sarwal, A., F.O. Walker, and M.S. Cartwright, Neuromuscular ultrasound for evaluation of the diaphragm. Muscle Nerve, 2013. 47(3): p. 319-29. McCool, F.D. and G.E. Tzelepis, Dysfunction of the diaphragm. N Engl J Med, 2012. 366(10): p. 932-42. Gottesman, E. and F.D. McCool, Ultrasound evaluation of the paralyzed diaphragm. Am J Respir Crit Care Med, 1997. 155(5): p. 1570-4. Ueki, J., P.F. De Bruin, and N.B. Pride, In vivo assessment of diaphragm contraction by ultrasound in normal subjects. Thorax, 1995. 50(11): p. 1157-61. Lu, Z., et al., Diaphragmatic Dysfunction Is Characterized by Increased Duration of Mechanical Ventilation in Subjects With Prolonged Weaning. Respir Care, 2016. 61(10): p. 1316-22. Zambon, M., et al., Assessment of diaphragmatic dysfunction in the critically ill patient with ultrasound: a systematic review. Intensive Care Med, 2017. 43(1): p. 29-38. Blumhof, S., et al., Change in Diaphragmatic Thickness During the Respiratory Cycle Predicts Extubation Success at Various Levels of Pressure Support Ventilation. Lung, 2016. 194(4): p. 519-25. DiNino, E., et al., Diaphragm ultrasound as a predictor of successful extubation from mechanical ventilation. Thorax, 2014. 69(5): p. 423-7. Llamas-Alvarez, A.M., E.M. Tenza-Lozano, and J. Latour-Perez, Diaphragm and Lung Ultrasound to Predict Weaning Outcome: Systematic Review and Meta-Analysis. Chest, 2017. 152(6): p. 1140-1150.

AC C

66.

17

ACCEPTED MANUSCRIPT

94. 95. 96. 97. 98. 99. 100.

101.

102.

RI PT

93.

SC

92.

M AN U

91.

TE D

90.

EP

89.

Boots, R.J., et al., Predictors of physician confidence to diagnose pneumonia and determine illness severity in ventilated patients. Australian and New Zealand practice in intensive care (ANZPIC II). Anaesth Intensive Care, 2005. 33(1): p. 112-9. Bellani, G., et al., Epidemiology, Patterns of Care, and Mortality for Patients With Acute Respiratory Distress Syndrome in Intensive Care Units in 50 Countries. JAMA, 2016. 315(8): p. 788-800. Frat, J.P., S. Ragot, and A.W. Thille, High-Flow Nasal Cannula Oxygen in Respiratory Failure. N Engl J Med, 2015. 373(14): p. 1374-5. Tejerina, E., et al., Clinical diagnoses and autopsy findings: discrepancies in critically ill patients*. Crit Care Med, 2012. 40(3): p. 842-6. Manno, E., et al., Deep impact of ultrasound in the intensive care unit: the "ICU-sound" protocol. Anesthesiology, 2012. 117(4): p. 801-9. Laursen, C.B., et al., Point-of-care ultrasonography in patients admitted with respiratory symptoms: a single-blind, randomised controlled trial. Lancet Respir Med, 2014. 2(8): p. 63846. Wallbridge, P.D., et al., A prospective cohort study of thoracic ultrasound in acute respiratory failure: the C3PO protocol. JRSM Open, 2017. 8(5): p. 2054270417695055. Brogi, E., et al., Could the use of bedside lung ultrasound reduce the number of chest x-rays in the intensive care unit? Cardiovasc Ultrasound, 2017. 15(1): p. 23. Zieleskiewicz, L., et al., Implementation of lung ultrasound in polyvalent intensive care unit: Impact on irradiation and medical cost. Anaesth Crit Care Pain Med, 2015. 34(1): p. 41-4. Lee, J., et al., Advanced imaging use in intensive care units has decreased, resulting in lower charges without negative effects on patient outcomes. J Crit Care, 2015. 30(3): p. 460-4. Shah, S.P., et al., Focused cardiopulmonary ultrasound for assessment of dyspnea in a resource-limited setting. Crit Ultrasound J, 2016. 8(1): p. 7. Shah, S.P., et al., Impact of the introduction of ultrasound services in a limited resource setting: rural Rwanda 2008. BMC Int Health Hum Rights, 2009. 9: p. 4. Riviello, E.D., et al., Hospital Incidence and Outcomes of the Acute Respiratory Distress Syndrome Using the Kigali Modification of the Berlin Definition. Am J Respir Crit Care Med, 2016. 193(1): p. 52-9. Jones, B.P., et al., Feasibility and Safety of Substituting Lung Ultrasonography for Chest Radiography When Diagnosing Pneumonia in Children: A Randomized Controlled Trial. Chest, 2016. 150(1): p. 131-8. Ferrante di Ruffano, L., et al., Assessing the value of diagnostic tests: a framework for designing and evaluating trials. BMJ, 2012. 344: p. e686.

AC C

88.

18

ACCEPTED MANUSCRIPT TABLE 1. Diagnostic characteristics of thoracic ultrasound for common ARF aetiologies Clinical Pattern

Ultrasound Findings

Differential Diagnoses

Pneumothorax [19-22]

Absent lung sliding

Lung bullae

88% (85-91) [19] 91% (86-94) [20]

98% (97-99%) [20]

ARDS

79% (68-98) [21]

98% (97-99) [21]

Ruled out by presence of ‘lung pulse’

Mainstem intubation

Hypoechoic space: ‘quad sign’

Pleural thickening

87% (81-92%) [22]

93% (89-96) [36] Intra-parenchymal fluid

Increased echogenicity

Atelectasis Pulmonary infarction

Air-bronchograms

All comers: 97% (93-99) [45]

Hypo-echoic/serrated distal edge Hypoechoic vascular structures

Neoplasm

Peripheral wedgeshaped consolidation

Pneumonia Neoplasm

Lower limb DVT

TE D

Right ventricular dysfunction

99% (98-99) [22]

96% (95-98) [36]

Studies typically in trauma (high prevalence) Heterogeneity on meta-analysis Clinical significance of pneumothorax not detected on CXR unclear High prevalence of effusion in included studies

98% (92-100) [37] All comers: 94% (85-98) [45]

94% (92-96) [46]

96% (94-97) [46]

95 (93-97) [47]

90% (86-94) [47]

ARF: 91% (81-97) to 100% (95-100) [41]

ARF: 78% (52-94) to 100% (99-100) [41]

M AN U

Pulmonary contusion

94% (88-97) [37]

Comments

RI PT

Localised fibrosis

Loss of pleural line

Pulmonary embolism [50, 51]

99% (98-99%) [19]

Lung-point ‘Stratosphere sign’

‘Sinusoid sign’ Pneumonia [41, 45-47]

Specificity* (inc 95% CI)

SC

Pleural effusion [36, 37]

Absence of B-lines

Sensitivity* (inc 95% CI)

87% (76-92) [50]

82% (71-89) [50]

85% (78-90) [51]

83% (73-90) [51]

Studies report high prevalence of pneumonia (clinical context important)

Sensitivity of US in ARF depends on unit of analysis (per patient or per region)[41] Poor performance as both ‘rule-in’ and ‘rule out’ test [51]

Pleural effusion

Increased B-lines (“interstitial syndrome”)

Diffuse: Cardiogenic pulmonary oedema, ARDS, Infection, Interstitial lung disease Non-cardiogenic pulmonary oedema

AC C

EP

Pulmonary oedema and interstitial lung disease (ILD) [67, 68]

Diaphragm dysfunction [79, 81, 82]

Reduced thickness

Reduced thickening ratio

Reduced/paradoxical excursion

Focal: Pneumonia, pulmonary contusion, fibrosis, lymphangitis Direct trauma Surgery Adjacent consolidation, malignancy or atelectasis Fluid (pleural/ascites) COPD Neuromuscular disease Denervation (neck/chest)

Cardiogenic pulmonary oedema: 94% (81-98) [67]

Cardiogenic pulmonary oedema: 92% (84-96.) [67]

85%(83-88) [68]

93% (91-94) [68]

Non-cardiogenic pulmonary oedema/ILD:

Non-cardiogenic pulmonary oedema/ILD:

N/A

N/A

N/A

N/A

Poorly predicts pulmonary occlusion pressure [37] Predicts extravascular lung water [37-39]

Thickening ratio has modest predictive value for weaning outcome, with lower accuracy for excursion.[87]

* Reported by systematic reviews Abbreviations:

19

ACCEPTED MANUSCRIPT

RI PT

ARDS – acute respiratory distress syndrome ARF – acute respiratory failure COPD – chronic obstructive pulmonary disease CXR – chest x-ray US - ultrasound

TABLE 2. Diagnostic and therapeutic impact of ultrasound in acute respiratory failure Design

Subjects

Setting

Inclusion

Exclusion

Blinding

Intervention

Comparator

Laursen et al [93]

Single blind RCT

320

ED

Resp Sx*

Cognition Age <18 Delayed US (>1h)

Clinician: Yes Operator: No

Combined lung and cardiac US on presentation

Discharge diagnosis (blinded)

Lichtenstein Meziere [27]

Prosp cohort

260

ICU

ARF req ICU

Clinician: yes Operator: unclear

Lung and compressive DVT US at ICU admission

ICU clinical diagnosis (treating clinician blinded to US)

Diagnostic accuracy

Silva et al [28]

Prosp cohort

No definite discharge diagnosis Multiple or rare discharge diagnoses Age <18

Clinician: yes Operator: unclear

Combined lung, cardiac and DVT US at ICU admission Combined lung and cardiac ultrasound at ICU admission

Diagnosis causing ARF (blinded panel)

Diagnostic accuracy

US: 83% Standard: 63%

Diagnosis causing ARF (blinded panel)

Diagnostic accuracy lung US (LUS) vs LUS + TTE (TUS)

LUS: 63% TUS: 79%

N/A

Impact on decision making and change in management

N/A

Change in: Dx Clinician confidence Mx

N/A

Prosp cohort

136

ICU

Xirouchaki [5]

Prosp cohort

189

ICU

Prosp cohort

50

Resp HDU

AC C

Wallbridge et al [94]

ARF req ICU

IMV with clinical question

EP

Baitaille et al [29]

ARF req ICU RR > 25

M AN U

ICU

Multiple final diagnoses Age <18 Multiple final diagnoses COPD (n=4)

Clinician: yes Operator: unclear

TE D

78

ARF req HDU

SC

Study

Unclear

No US within 24h

No

No

Lung ultrasound

Lung ultrasound Estimation of CVP

N/A

Primary outcome Correct diagnosis at 4h

Diagnostic accuracy US: 88% Standard: 64%

(at 4h) US: 90.5% Standard: 76%

Appropriate treatment at 4h: US: 78% Standard: 57% 25% of patients had incorrect or uncertain diagnosis at 2h

Sig improved diagnostic accuracy for pulmonary oedema and pneumonia, no change in PE and pneumothorax Net reclassification index: 85.6% Change in Mx: 47% Change in Dx: 34% Increased clinician conf: 44% Change in Mx: 30%

*: one or more of: respiratory rate >20 breaths/min, oxygen saturation <95%, started on oxygen therapy; one or more of: current or previous dyspnoea, cough, chest pain Abbreviations: ARF – acute respiratory failure Conf – confidence CVP – central venous pressure Dx – diagnosis DVT – deep vein thrombosis ICU – intensive care unit

Other results

PE – pulmonary embolism Prosp – prospective Req – requiring RR – respiratory rate Sig - significant Sx - symptoms

20

ACCEPTED MANUSCRIPT TTE – transthoracic echocardiogram US - ultrasound

AC C

EP

TE D

M AN U

SC

RI PT

IMV – invasive mechanical ventilation Mx - management N/A – not applicable

21

ACCEPTED MANUSCRIPT Figure 1. Levels of diagnostic imaging efficacy and associated clinical impact. Adapted from Hew and Tay [17] Figure 2. Clinical impact of diagnostic imaging. Image of ultrasound machine used with permission (MindRay Medical Australia).

RI PT

Figure 3. Lung point. Transition between normal lung sliding (solid white arrow – B-line) and pneumothorax (dashed white arrow). Figure 4. Pleural effusion. D – diaphragm, E – effusion, L – lung edge.

SC

Figure 5. Consolidated lung. Air bronchograms (white arrows) on background of increased echodensity with serrated distal margin. Figure 6. Increased B-lines (white arrows) consistent with interstitial syndrome.

AC C

EP

TE D

M AN U

Figure 7. Diaphragm. Image of diaphragm (white arrow) with visible muscle layers in anterior axillary plane between two ribs (dashed white lines).

22

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT Diagnostic chest ultrasound for acute respiratory failure: Highlights

EP

TE D

M AN U

SC

RI PT

Ultrasound (US) accurately diagnoses causes of acute respiratory failure (ARF) US improves diagnostic reasoning and decision-making for ARF in intensive care However, the impact of US for ARF on patient and societal outcomes is unknown Further data are also needed to examine US for ARF outside critical care settings

AC C

• • • •

ACCEPTED MANUSCRIPT Professor J Virchow Editor in Chief Respiratory Medicine

RE: Review paper – Chest ultrasound for acute respiratory failure

Dear Professor Virchow,

RI PT

20/2/18

SC

We wish to confirm that there are no known conflicts of interest associated with this publication and there has been no significant financial support for this work that could have influenced its outcome.

M AN U

We confirm that the manuscript has been read and approved by all named authors and that there are no other persons who satisfied the criteria for authorship but are not listed. We further confirm that the order of authors listed in the manuscript has been approved by all of us. We confirm that we have given due consideration to the protection of intellectual property associated with this work and that there are no impediments to publication, including the timing of publication, with respect to intellectual property. In so doing we confirm that we have followed the regulations of our institutions concerning intellectual property.

TE D

We understand that the Corresponding Author is the sole contact for the Editorial process (including Editorial Manager and direct communications with the office). He is responsible for communicating with the other authors about progress, submissions of revisions and final approval of proofs.

AC C

Regards,

EP

We confirm that we have provided a current, correct email address which is accessible by the Corresponding Author and which has been configured to accept email form ([email protected])

Peter Wallbridge1,2 Daniel Steinfort1,2 Tunn Ren Tay3 Louis Irving1,2 Mark Hew2,4

1. Department of Respiratory & Sleep Medicine, Royal Melbourne Hospital, Melbourne, Australia 2. Department of Medicine, Royal Melbourne Hospital, University of Melbourne, Melbourne, Australia 3. Department of Respiratory and Critical Care Medicine, Changi General Hospital, Singapore 4. School of Public Health & Preventive Medicine, Monash University, Melbourne, Australia