Routine lung ultrasound to detect postoperative pulmonary complications following major abdominal surgery: a prospective observational feasibility study

There is increasing interest in early detection of postoperative pulmonary complications (PPCs) to reduce patient morbidity and mortality [1]. Furthermore, in patients unexpectedly admitted to the intensive care unit (ICU), insufficient diagnostic imaging was often performed on the general ward and inadequate treatments were initiated [2]. Early PPC detection would enable physicians to start treatment in time and, therefore, prevent its negative impact on patient outcome.

Respiratory failure following anaesthesia and surgery typically starts with atelectasis, due to diaphragmatic dysfunction, inability to clear secretions and immobility. Chest auscultation and chest X-ray (CXR) are frequently used as diagnostic modalities to detect PPCs, but have limited diagnostic accuracy [3‐6]. Moreover, ideally, radiography requires an upright position of the patient, and multiple concomitant lung abnormalities further complicate the two-dimensional interpretation [7‐9]. Computed tomography (CT) is the gold standard for pulmonary pathology but requires significant ionizing radiation and the need for transfers within the hospital, which is not without risk [3].

Alternatively, lung ultrasound (LUS) is a quick, point-of-care and radiation free technique. It has an excellent diagnostic accuracy for atelectasis, pulmonary effusions, pulmonary edema, and/or pneumonia in critically ill patients compared to the gold standard CT [3, 4, 6]. Therefore, LUS is well established in these pathologies, and practical approaches to perform LUS have been described extensively elsewhere [10, 11]. We previously showed that, in cardiothoracic surgery patients, LUS detected more (clinically relevant) PPCs compared to CXR and at an earlier time point [12]. Following major abdominal surgery, the feasibility could be limited by dressings, chest drains, or subcutaneous emphysema.

Daily routine CXR has been largely abandoned in ICU and cardiothoracic patients due to its limitations, costs, and radiation burden. On demand, CXR performance has led to a significant decrease in the number of CXRs without any increase in adverse events in ICU patients [13]. Studies comparing on-demand strategies with routine strategies are lacking in patients after major abdominal surgery. However, performing CXR on demand does not facilitate early detection of postoperative pulmonary complications. Potentially, routine LUS can detect PPCs early and with higher accuracy to improve perioperative management in major abdominal surgery patients [4, 10, 14].

The primary objective of this pilot study was to test the feasibility of routine LUS following major abdominal surgery and to report the detected PPCs with LUS in these postoperative patients. Subsequently, PPC detection rates for LUS and CXR were compared.

We conducted this prospective, observational feasibility study in the Amsterdam UMC—VU University Medical Center Amsterdam (VUmc Amsterdam, The Netherlands), a tertiary hospital. The local Human Subjects Committee of the VUmc approved the study (METc 16/128), and written informed consent was obtained from all participants. Patient inclusion started June 2016 and ended March 2017. Patients were included on the general ward. The study included consecutive adult patients (age ≥ 18 years) scheduled for elective major abdominal (e.g., gastrointestinal, vascular, or renal) surgery with an intermediate or high risk for the development of postoperative pulmonary complications according to the Assess Respiratory risk In Surgical patients in CATalonia (ARISCAT) risk score ≥ 26 (https://​www.​mdcalc.​com/​ariscat-score-postoperative-pulmonary-complications) [15, 16]. Exclusion criteria were trauma or emergency surgery.

115 patients were considered eligible for inclusion. 110 patients signed informed consent, 12 were excluded because of postoperative correction of initial ARISCAT score (n = 7) or cancelling informed consent (n = 5). Overall, 98 consecutive patients were included in the study. Table 1 shows the patient characteristics of included patients. One patient died within the 30 day follow-up period.

This pilot study demonstrates that routine LUS is highly feasible and frequently detects PPCs in patients following major abdominal surgery. In addition, LUS detected more PPCs when compared to CXR when both modalities were performed on POD 0. Importantly, the discordant observations for LUS and CXR were highest for atelectasis and pleural effusion.

We confirm that LUS is highly feasible in postoperative patients [12, 20, 21]. Goudy et al. found that LUS was feasible in 97% of the hemi thoraxes in 252 ultrasound examinations after cardiothoracic surgery, despite the presence of chest drainage tubes or applied dressings [21]. To the best of our knowledge, we are the first to report high feasibility following major abdominal surgery. Following major abdominal surgery, dressings, chest drains, or subcutaneous emphysema or limited patient mobility could have prohibited the feasibility in this setting.

We cannot compare our results with the previous research, because this is the first study to report PPC rates detected by LUS following major abdominal surgery. The rates found with LUS are high in comparison with PPC rates previously reported with other diagnostic modalities [15]. However, in the cardiothoracic surgery population, it was previously reported that the rate of PPCs detected with LUS was up to 100% [12, 22]. Detection of PPCs with CXR in our study (37%) is in line with preoperative risk scores [16]. In addition, Mazo et al. recently showed incidence rates of PPCs ranging from 38 to 50% in a large validation study in Western Europe in a comparable study population [16].

We found that LUS and CXR, when performed at the same time, identified pleural effusion and atelectasis in different patients (discordant pairs) following major abdominal surgery. These discordant pairs were detected in up to 50% of the patients who had both imaging modalities performed on the same day. This difference can be explained by a higher sensitivity and specificity of LUS in detecting pleural effusion and atelectasis compared to CXR [3, 6, 20]. In a recent systematic review and meta-analysis, LUS had an overall sensitivity of 95% (92–96%) and specificity of 94% (90–97%) compared to CT in adult critically ill patients with respiratory symptoms [6]. CXR had an overall sensitivity of 49% (40-58%) and specificity of 92% (86–95%) in the same systematic review and meta-analysis. Therefore, LUS is being considered by others as the bedside gold standard for pulmonary pathology in critically ill patients with respiratory symptoms [4, 14].

In a recent pilot study performed in cardiothoracic surgery evaluating effectiveness of LUS as the primary diagnostic imaging technique, LUS was exhaustive in up to 80% of the patients. Subsequent CXR was not required in the postoperative management of these patients to assist decision making in this study and LUS was considered effective in perioperative patient management. Furthermore, LUS allowed further discrimination of lung abnormalities mainly between atelectasis and pleural effusions. [22, 23].

Other advantages of LUS are that the time to perform LUS was significantly shorter than for CXR in the postoperative setting [12] and LUS had excellent inter-observer agreement [4, 12]. Furthermore, LUS can be used to guide central venous catheter placement and position [24], thoracentesis, and the management of fluid administration in acute circulatory failure [11]. However, if routine LUS findings are inconclusive in patients with respiratory symptoms, additional imaging is warranted; we do suggest using additional CT imaging to make a final diagnosis.

Our study has important limitations. LUS examinations could not be performed on all postoperative days. On POD 0, patients were frequently admitted after working hours, and, on POD 3, was frequently during weekends. Consequently, researchers were not always available to perform LUS. Therefore, the predictive value of detected complications to become clinically relevant PPCs could not be analysed. Another important limitation is the absence of the gold standard (computed tomography) for lung pathology to relate our LUS findings. We chose to use and adjust the most practical, succinct, but well validated and accurate LUS protocol (BLUE protocol) [14, 19]. However, detection rates could even be increased more by further quantifying B lines and to scan 8 or even 28 separate intercostal spaces instead of 4 [25]. The clinical importance of using LUS is unknown with regard to detecting PPCs that do not require therapy according to the treating physician. This should be studied in further studies. We strongly believe that not all atelectasis or pleural effusion need treatment. However, LUS findings, similar to other clinical findings, should always be seen in light of patient’s symptoms and health status. Our findings indicate to further study LUS as a primary screening tool for detecting PPCs in major abdominal surgery patients. Future studies should focus on patient outcome and cost effectiveness.

Strength of this study is that we have demonstrated that LUS is indeed feasible and frequently detects PPCs in patients following major abdominal surgery. The protocol we propose is easy to use, also for physicians newly trained in LUS. The protocol is based on limited, standardized signs, a major advantage of LUS, because the risk of wrong interpretations is thereby decreased.

Developing competence in LUS is considered straightforward. Surgeons and anaesthesiologists active in the postoperative period should consider training in LUS as part of their point-of-care ultrasound skills. Furthermore, adding this useful tool into resident training would parallel its broader use in other medicine specialties. Teaching the basics of LUS to perioperative physicians and residents would be fairly easy. For example, in nephrology, a 4-h course that includes pre-course cognitive preparation, a didactic lecture and training in image acquisition/interpretation is sufficient to provide the learner a strong foundation in LUS. However, the course alone is not sufficient to provide competence [26]. Competence in LUS requires additional bedside scanning of patients under the supervision of capable faculty as outlined before [10].

This pilot study shows that LUS is highly feasible and detects PPCs frequently in patients after major abdominal surgery. LUS findings should be evaluated in combination with all available clinical data when considering initiating treatment. Closely monitoring patients and detecting PPCs with LUS could create a window of opportunity to limit the impact of PPCs on patient outcome. This needs to be studied in future trials.