Impact of integrated use of diagnostic ultrasound examinations in respiratory intensive care units

Background Implementing point-of-care multiorgan ultrasound (POCUS) to the initial assessment of ICU patients allows intensivists to immediately integrate ultrasound findings with the patient history, physical, and laboratory results, yielding a powerful clinical synergy, improving diagnostic accuracy, and ameliorating further management plans. The aim of this work was to assess the diagnostic performance and therapeutic effect of POCUS in patients admitted to respiratory ICU (RICU). Patients and methods A prospective study was carried out on patients admitted to the RICU. POCUS examination was performed to the patients within 12 h of admission that included echocardiography, lung ultrasound, abdominal ultrasound including inferior vena cava assessment and lower limb venous duplex. Results A total of 102 patients were included. The total number of sonographic findings was 320, of which 94 (29.3%) were new findings. This resulted in confirmation of the admitting diagnosis, modification of the admitting diagnosis, prompted further testing, change in medical therapy prescribed, and prompted invasive procedures in 35, 51, 11, 41, and 14% of patients, respectively. However, it was ineffective in confirming or modifying diagnosis, provided wrong diagnosis, and missed a diagnosis in 29.4, 2, and 11.7% of patients, respectively. Conclusion Integrating POCUS in the initial assessment of critically ill RICU patients together with standard diagnostic tests lead to diagnostic and therapeutic changes in most of patients which affected the management of these patients. Thus, it seems reasonable to consider the routine use of POCUS as a new respiratory examination option in the armamentarium of the intensivists.


Introduction
The inaccuracy of physical examination at admission to the ICU has been extensively reported, and lifethreatening conditions could be missed at the primary assessment, especially in patients presented with acute respiratory symptoms [1][2][3]. Thus, intensivists need to make rapid, accurate, and appropriate decisions in situations where there is a high degree of stress and uncertainty and when patients possess little physiologic reserve [4].
Implementing point-of-care multiorgan ultrasound (POCUS) allows the intensivists to personally perform and interpret the ultrasound (US) examination results at the bedside and immediately integrate ultrasound findings with the patient history, physical, and laboratory results, hence yielding a powerful clinical synergy, improving diagnostic accuracy and ameliorating further management plans [5][6][7][8].
Results of several studies have shown that application of POCUS in patients admitted with acute respiratory symptoms to emergency department or ICU was superior to standard diagnostic tests alone for establishing an accurate diagnosis, leading to changes in medical therapy and prompting further invasive procedures [3,5,8]. Therefore, now it seems crucial to use POCUS as part of the standard diagnostic tests in critically ill patients.
Lung US has been recently introduced in our respiratory ICU (RICU). Its effect on diagnosis and management has been established in several studies [9,10]. Combing lung US examination with other organ US examinations, such as heart, inferior vena cava (IVC), abdomen, and deep venous system, using POCUS concept into standard diagnostic assessment of critical ICU patients, remains scarcely applied on the international level and has never been applied in our RICU [11,12]. The actual levels of POCUS implementation and contribution to ICU patient management need further studies [11,13].
The aim of our study was to assess the diagnostic performance and therapeutic effect of POCUS in patients admitted to RICU.

Patients and methods
Setting and study population This prospective study was carried out on patients admitted to the RICU of Ain Shams University hospitals, Egypt, in the period from October 2016 to July 2017. The study design has been approved by the research ethics committee of our institution, and a written informed consent was obtained from the patients or their relatives.

Inclusion and exclusion criteria
All adult patient admitted to the RICU were included when the sonographic examinations could be performed within 12 h after the primary assessment. Exclusion criteria were age younger than 18 years, discharge form RICU within 24 h, and the presence of patient-related conditions that strongly hamper ultrasound examination such as morbid obesity, the presence of subcutaneous emphysema and severe edema, and the presence of environmentrelated limitations that hamper proper examination such as patient isolation for fear of infection.

Primary assessment
On admission, the diagnosis was made by the attending resident and senior registrar in charge, who had at least 3 years of experience in RICU, depending on history taken in the emergency department or referring ward, clinical examination, laboratory findings, and imaging findings without beside ultrasonographic evaluation. The severity of the patient's condition on admission was graded using the new Simplified Acute Physiology Score (SAPS II) [14] Sonographic examination Within 12 h of RICU admission, POCUS of the heart, lungs, deep veins, and abdomen was performed by a single physician (M.F.) who received a 2-month comprehensive training in bedside ultrasonography and echocardiography under supervision of an expert radiologist and an expert cardiologist. This was followed by a 2-month period of directly supervised practice after which he started to work independently.
The operator was aware of the patient's clinical picture but was blinded from the provisional diagnosis and any initial radiological assessment.
The examinations included Lung ultrasound 'Six' ultrasound areas were examined on each side: the anterior, lateral, and posterolateral views in the upper and lower thoracic regions [2]. Interpretation was done following the principles described by Lichtenstein [15]. The low-frequency curvilinear probe was used to examine the lung parenchyma at a depth of 10 cm and the high-frequency linear probe for examination of pleural sliding on 2D and M-mode.

Focused echocardiography
A goal-directed transthoracic echocardiography was done using the basic views, including parasternal short axis and long axis views, apical views (fourchamber, five-chamber, and two-chamber), and subcostal views [2].

Pelvi-abdominal ultrasound
Pelvi-abdominal ultrasound was done according to 'the focused assessment with sonography for trauma' examination [16].

Assessment of the inferior vena cava
The subcostal view was used to measure maximum diameter, estimate the percent of respiratory collapsibility (Caval index) [17], and visualize the intraluminal thrombosis. The curvilinear 2-5-MHz probe was used. Measurements were made at a distance not less than 2 cm caudal from the junction of the right atrium [18].

Venous system
Mild compression maneuver was used to assess the lower limb (right and left femoral and popliteal veins) and neck vessels (right and left jugular veins). Doppler study was used when needed.
The following specific diagnostic points (Table 1) were prospectively defined as previously described by others [2,3].
We used the criteria previously described by Manno et al. [2] to define ultrasound-induced modification, confirmation, wrong evaluation, and lack of confirmation of admitting diagnosis ( Table 2).
Ultrasound findings not previously known to the attending resident and the senior registrar in charge, which was unrevealed by ultrasonographic examination, will be defined as a 'new' finding. Changing the admitting diagnosis or medical therapy or to perform invasive procedures was decided by the senior physician and the RICU consultant.

Statistical analysis
Data were collected, revised, coded, and entered to the Statistical Package for Social Science (IBM SPSS) version 20.0. (SPSS Inc., Chicago, Illinois, USA). Descriptive statistics were performed, including demographic characteristics, medical history, and symptoms, at admission. The effect of point of care ultrasonography on the diagnosis and treatment was calculated. Percentages were computed for the categorical variables. Data analysis was conducted using SPSS software, version 20.0 (SPSS Inc., Chicago, Illinois, USA).

Baseline patient characteristics
A total of 102 patients were enrolled in our study. Patient characteristics, presenting symptoms, and  medical history are presented in Table 3. Overall, 18 (17.6%) patients required invasive mechanical ventilation and nine patients needed noninvasive positive pressure ventilation.
The admitting and final diagnosis among studied patients are reported in Table 4; Acute exacerbation of chronic obstructive pulmonary disease was the most common diagnosis among our patients.
Different sonographic findings are shown in Table 5. The total number of sonographic findings was 320, of which 94 (29.3%) were new ultrasound findings (Figs 1-3). The number of new finding per patient ranged from 0 to 4 with a median of one finding per patient.

Diagnostic effect
The effect of ultrasound examination among studied patients is detailed in Table 6. Ultrasound examination modified the admitting diagnosis in 52/102 cases (50.9%) (Figs 1 and 5). In 14 (13.7%) cases, more than one modality was used. Ultrasound also confirmed the diagnosis (

Sonographic examination
The time required to complete the examination ranged from 15 to 45 min, being least in the last 20 patients, with a mean of 38.4 min.

Discussion
Adding POCUS of the heart, lungs, abdomen, IVC, and deep veins to the standard initial diagnostic tests within 12 h of RICU admission resulted in confirmation of the admitting diagnosis, modification of the admitting diagnosis, prompted further testing, change in medical therapy prescribed, and prompted invasive procedures in 35,  Several previous studies evaluated POCUS in patients presented with respiratory symptoms to the emergency room (ER) or general ICUs, but up to our knowledge, none of them assessed it specifically in RICU [2,3,5,8,12,13]. RICU patients are a heterogeneous group presenting with either primary respiratory disease or secondary respiratory disease to other illness. They are characterized by respiratory failure, need for mechanical ventilation, severe illness, multiple system dysfunction, and multiple coexisting comorbidities [19].
In a prospective multicentric study in 142 ICUs in France, Belgium, and Switzerland by Zieleskiewicz et al. [13] to describe the diagnostic and therapeutic effects of POCUS performed during a 24-h period, the use of POCUS changed the diagnosis in 21% of cases, led to confirmation of a suspected diagnosis in 63% of cases, and was associated with interventions including treatment, imagery ordering, and patient triage in 69% of cases [13]. Manno et al. [2] also investigated the use of POCUS in ICU patients, and it was found that ultrasound examination confirmed the admitting diagnosis in 58.4% of cases, modified it in 25.6% of cases, was ineffective in confirming or modifying it in 13.6% of cases, and missed it in 2.4% of cases. The ultrasonographic findings prompted further testing in 18.4% of patients, led to changes in medical therapy in 17.6% of patients, and led to invasive procedures in 21.6% patients [2].

Echocardiography
Transthoracic echocardiography in critically ill has been associated with an improvement in the diagnosis of patients with acute respiratory failure and/or shock [20,21]. It also increased treating physicians ability to pick up subtle acute decompensated heart failure cases initially misdiagnosed as COPD or other diagnoses [8]. Similarly, in our study, echocardiography had the greatest effect on diagnosis in 41% (42/102) of patients, of whom 13 patients had ejection fraction less than or equal to 45%.
Tricuspid gradient measurement for estimation of right ventricular systolic pressure is a useful and practical method for noninvasive prediction of pulmonary artery pressure and correlates strongly with invasive pulmonary artery systolic pressure assessment [22,23]. A clinical diagnosis of pulmonary hypertension (PHTN) was made if the peak systolic pressure was greater than 30 mmHg. Secondary PHTN is a common complication of chronic lung disease, which is always associated with poor prognosis and usually progresses to right heart   A known breast cancer female patient with severe dyspnea, and chest radiograph showed massive effusion confirmed by lung ultrasound (a). Echocardiography (b) showed pericardial effusion (star) with signs of tamponade adding a new etiological diagnosis for her dyspnea.

Figure 3
A 35-year-old IV addict male patient presented with toxic symptoms and respiratory failure, and chest radiograph and computed tomography showed septic emboli (a). Lung ultrasound (b) revealed confirmed pleural-based hypoechoic lesions (arrow); echo showed vegetation over tricuspid valve (star) and rim of pericardial effusion (arrow), leading to modification of diagnosis and medical treatment.
failure [24]. However, diagnosis may be delayed in many cases as the dyspnea is usually attributed to the primary lung disease [24]. This was clearly obvious in our study, as 14 (74%) of 19 patients were newly diagnosed as having PHTN and/or right-sided heart dysfunction, and PHTN was secondary to their primary lung disease (nine patients with COPD, four patients with bronchiectasis, one patient with interstitial lung disease). This finding greatly influenced the workup of the patients and had a direct effect on the therapeutic plan.
Noncardiologists who learn focused critical care echocardiography can adequately interpret basic information and successfully incorporate it into advanced cardiopulmonary life support [25]. On the contrary, noncardiologists with minimal training failed to identify important cardiac abnormalities such as valvular heart diseases, regional wall abnormalities, and acute cor pulmonale [26,27]. This obstacle was surmounted in our study by the comprehensive training given to our sonographer, where severe mitral stenosis as a cause of massive hemoptysis was diagnosed in a patient with known systemic lupus erythematosus and severe mitral regurgitation causing pulmonary venous congestion in another patient, in addition to the previously stated patients with PHTN (Fig. 1).
A variety of complications attributed to septic pulmonary emboli have been described in right-sided endocarditis [28]. This makes transthoracic echocardiography an important initial investigation   in these patients. POCUS enabled rapid, bedside, noninvasive diagnosis of infective endocarditis in two patients with history of intravenous drug abuse who presented with septic embolic and empyema (Fig. 3).
These data highlight the importance of transthoracic point-of-care echocardiography by respiratory intensivists, as performing an echocardiography and getting immediate interpretation by a cardiologist is not always available in the RICU [29].

Inferior vena cava ultrasound
Integrating the IVC analysis with a multiorgan ultrasound approach, which includes evaluation of the dimensions and function of the right and left cardiac chambers, with basic evaluation of the pulmonary congestion by assessing lung ultrasound for B lines led to changes in the medical therapy in 10/102 (8%) of our patients. This is consistent with the results of other studies that reported fluid status adjustments to be a frequent therapeutic action performed following a focused ultrasound examination [12].

Lung ultrasound
In one study, focused LUS identified a missed life-threatening condition in 23 (17%) patients presenting with acute respiratory symptoms [3]. In another study, LUS pointed out 55 (40.3%) new findings in patients admitted to medical ICU enabling the differentiation of the etiologic diagnosis in patients with an admitting diagnosis of acute respiratory insufficiency [2].
In our study, LUS was able to assess the lungs in all included patients. A total of 130 findings were seen, of which 12 were identified as new findings; causing modification of the admitting diagnosis in 12/102 (11.7%) patients. LUS also led to changes in medical therapy in 23 (22.5%) patients, including fluid management and diuretic therapy. We attributed this small number of new finding to the nature of our patients, as most of them have a significant respiratory disorder that was usually apparent in plain chest radiograph.
We describe a pattern of bronchiectasis in LUS, in the form of intersecting comet tail artifacts not erasing A-lines with or without an irregular pleural line detected by linear probe, a finding that was not previously mentioned and needs further analysis (VEDIO).
At bedside, initial chest ultrasound is more sensitive than chest radiographies in the detection of small pleural effusions that are misdiagnosed as parenchymal opacities or are not seen [9,30]. A total of 32 pleural effusions were detected in our study (Fig. 2), and with the help of bedside LUS, the nature of the fluid could be assessed and aided the change in the management in three cases. Overall, LUS had a greater effect on the therapeutic plan of our patients rather than a diagnostic effect.

Pelviabdominal ultrasound
Abdominal ultrasound evidenced 14 new pathologic findings and modified admitting diagnosis in 7/102 (6.8%) cases (Fig. 1). This nearly matched the results of Manno et al. [2], where abdominal examination as a part of the ICU 'sound protocol' evidenced 20 new pathologic findings and induced changes in therapy in 3/125 (2.4%) cases.

Duplex
Central line insertion is a daily practice in ICU. Diagnosing IJV thrombosis in two patients using ultrasonography helped in selecting the site of A peripheral hypoechoic lesion (arrow), which was suspected to be a pulmonary infarction in a patient with chronic obstructive pulmonary disease, computed tomography pulmonary angiogram showed an atelectatic band.

Figure 6
A 30-year-old male presented with severe dyspnea, fever, and right sided D-shaped homogenous opacity (b). Lung ultrasound revealed complex septated pleural effusion (a) managed by medical treatment followed by video-assisted thoracoscopy.  central line insertion in addition to initiation of therapeutic anticoagulation (Fig. 1).
Among the three patients in which a new lower limb DVT was diagnosed, two cases were diagnosed as having pulmonary embolism. This is in accordance with previous studies that combined echocardiography and venous ultrasonography to chest sonography as a reliable method for screening patients with suspected pulmonary embolism at bedside [31].

Wrong diagnosis
In two cases, ultrasonography provided a wrong diagnosis; a full stomach appearing below the diaphragm was misdiagnosed as a peritoneal fluid collection, which is a common pitfall in point-ofcare ultrasound [32]. Another patient with systolic heart failure, a pleural-based triangular wedgeshaped opacity suggested the diagnosis of pulmonary embolism (Fig. 4), but appeared to be an atelectatic band in computed tomography pulmonary angiography.

Missed diagnosis
Transthoracic ultrasound achieves only poor visualization of the mediastinum as compared with computed tomography scan [33]. Thus, it was expected to miss central, mediastinal, or hilar lesions in 5/102 (4.9%) of studied patients.
The limitation of our study was that the examiner has not been blinded to the clinical picture of the patient, which is difficult to eliminate in any ultrasound examination. Being a single-center study, which included only patients with respiratory diseases, the results cannot necessarily be applied to other ICUs.
In conclusion, integrating POCUS in the initial assessment of critically ill RICU patients together with standard diagnostic tests led to diagnostic and therapeutic changes in most of patients, which affected these patients' management. Thus, it seems reasonable to consider routine use of POCUS as a new respiratory examination option in the armamentarium of the intensivists.

Financial support and sponsorship
Nil.

Conflicts of interest
There are no conflicts of interest.