Intensive care unit physician-delivered point of care renal tract ultrasound in acute kidney injury is feasible

Abstract

Acute kidney injury is common in critically ill patients, with ultrasound recommended to exclude renal tract obstruction. Intensive care unit clinicians are skilled in acquiring and interpreting ultrasound examinations. Intensive Care Medicine Trainees wish to learn renal tract ultrasound. We sought to demonstrate that intensive care unit clinicians can competently perform renal tract ultrasound on critically ill patients. Thirty patients with acute kidney injury were scanned by two intensive care unit physicians using a standard intensive care unit ultrasound machine. The archived images were reviewed by a Radiologist for adequacy and diagnostic quality. In 28 of 30 patients both kidneys were identified. Adequate archived images of both kidneys each in two planes were possible in 23 of 30 patients. The commonest reason for failure was dressings and drains from abdominal surgery. Only one patient had hydronephrosis. Our results suggest that intensive care unit clinicians can provide focussed renal tract ultrasound. The low incidence of hydronephrosis has implications for delivering the Core Ultrasound in Intensive Care competencies.

Introduction

Acute kidney injury (AKI) has an incidence of 57% in critically ill patients.¹ Renal ultrasound is a first-line investigation to help determine the cause of AKI.² Arranging urgent ultrasound imaging for intensive care unit (ICU) patients can be challenging and time consuming as frequently a portable bedside scan is required and there is escalating pressure on hospital imaging services, especially outside normal working hours. Early bedside ultrasound is recommended in AKI where the cause has not already been identified or to exclude the presence of renal tract obstruction,² with earlier detection enabling expedient relief of obstruction via nephrostomy or ureteric stenting. Focussed ultrasound of the kidneys by ICU physicians for the assessment of hydronephrosis can potentially improve patient outcomes with quicker time to diagnosis and intervention, reduce the need for transfer of critically ill patients to the ultrasound department and reduce the burden of work for the radiology department.

Basic renal tract imaging forms part of the syllabus for ‘Core Ultrasound in Intensive Care’ (CUSIC).³ Focused renal tract ICU ultrasound can be limited to identifying the kidneys, assessing for hydronephrosis, examining for a full bladder and differentiating the anechoic space of the bladder from free pelvic fluid. Focused renal tract imaging is quick and when performed to answer specific questions (such as presence of hydronephrosis), can be an invaluable tool in the ICU Physician’s armoury to identify AKI aetiology. Development of point of care ultrasound has been suggested for many clinical specialties, with various accreditation packages available.^4,5 Many ICU physicians have ultrasound competency in different modalities, ranging from echocardiography, lung, intra-abdominal, vascular and regional anaesthesia. Feasibility of extension of these skill sets to focussed renal tract imaging is an area that this study wished to assess.

We performed this single-centre feasibility study by performing bedside ultrasound imaging of the renal tract by two consultant intensivists in critically ill patients on the ICU, with assessment for the presence of hydronephrosis. All images were reviewed by the same radiologist for adequacy of images and whether the findings were in correlation with their own.

Although other groups have examined point of care renal imaging by other specialists in a range of clinical areas,^6–10 we believe this is the first study assessing focused renal tract imaging by intensivists on the ICU.

Method

All patients admitted to the ICU of a University teaching hospital with an AKI were prospectively scanned by one of two intensive care consultants over a six-week period. Both consultants had prior experience with ultrasound and echocardiography. The scanning was performed using the current departmental ICU ultrasound machine, a Sonosite EDGE. A curvilinear probe was used (Sonosite C60) starting with the manufacturer’s default ‘abdominal’ setting (with a frequency of 2.2 MHz and use of tissue harmonic imaging). Images were optimised for depth and gain as necessary. Each consultant was blinded to the patient history and probable aetiology of the AKI so as not to bias interpretation. Ethical approval was obtained for this study, and patients or their advocates gave verbal consent.

The patients were scanned in whichever position they were being nursed at the time, so as to not affect any aspect of the patients’ treatment and nursing needs, and with the aim of minimising any discomfort to the patients. A curvilinear low-frequency probe was used, with acoustic windows identified usually posterior to the mid-axillary line, and in the suprapubic area.

Longitudinal and transverse views of both kidneys and of the pelvis were obtained and renal images stored. Pelvic images were only stored if there was any evidence of free fluid or a distended bladder, as all patients studied had a urinary catheter in situ. Doppler studies were not included within the study protocol. Duration of scan, adequacy of views of both kidneys, presence of hydronephrosis, other significant abnormalities and suggested need for further departmental ultrasound scan were all noted for each patient (Figure 1) . Hydronephrosis was defined as 30 patients were scanned in total.

Figure 1.

Data collection document for each scan performed.

All images were reviewed by a senior radiology registrar with specialist interest in ultrasound for adequacy and accuracy of findings.

Results

Thirty patients were scanned in total, of which two had only a single kidney (one by previous nephrectomy and one as an incidental finding on previous computed tomography (CT) scan performed for a separate indication). Nineteen patients (63%) were male.

Both kidneys were identified in 26 of 28 patients with two kidneys, and the one kidney was identified in two of two patients with a single kidney. The reason for failing to identify one kidney in the two patients was extensive drains and dressings from recent major abdominal surgery, which obscured the acoustic window for imaging in the supine position (Figure 2).

Figure 2.

Chart showing findings on the 30 patients scanned.

Views of all kidneys were deemed adequate by the ICU physician in 23 of 30 patients, and inadequate for at least one kidney in 7 of 30 patients. The inadequate views were frequently due to our decision to scan the patients without significantly moving them, meaning that surgical dressings obscured the acoustic windows in the supine position, or pain from intra-abdominal pathology hindered our ability to archive high-quality images.

Hydronephrosis was noted only in one patient and was graded as mild to moderate in a patient who presented with urinary sepsis and bladder outflow obstruction. Subsequent CT scanning (delayed by over 48 h) revealed resolution of these changes. Colour Doppler was used to confirm the presence of hydronephrosis in this case (although was not used routinely in all cases when it was clear that there was no evidence of renal tract obstruction).

Pelvic free fluid was seen in 4 of 30 patients, and no patients had a bladder catheter which was either misplaced or ineffective at draining the bladder.

Median duration of all 30 scans was 7 min. The scan duration fell during the study period, as confidence with the technique and with quality image acquisition grew. The median duration for scans 1–10 was 9 min, for scans 11–20 was 8.5 min and for scans 21–30 was 5 min.

A further scan by the specialist radiologist was recommended in six patients (Figure 3) . The reasons were inability to adequately image one kidney, with worsening or non-resolving AKI (n = 3), abnormal kidney shape or echotexture (n = 2) and incidental findings suggestive of pelvic pathology (n = 1).

Figure 3.

Improvement in scan duration over the course of the study (x = scan number; y = scan duration (min)).

No images that were deemed adequate by the scanning intensivist were rejected as being inadequate by the reviewing radiologist who, importantly, was blinded to any prior interpretation of the images.

Discussion

This study highlights that ICU physicians, with relevant experience in ultrasound, can successfully perform point of care renal tract ultrasound to the standard specified by CUSIC. The major impediment to a successful and complete scan in this study population was the presence of painful intra-abdominal pathology and post-operative dressings and drains. These factors would normally be mitigated by moving or turning the patient to find alternative acoustic windows or by using a different imaging modality, such as CT scanning. Frequently, neither of these is appropriate for ICU patients, especially if obstruction is not felt to be the likely cause of AKI. In our series, there was only one case of hydronephrosis seen on ultrasound, which may likely reflect the aetiology of AKI in critical care settings, where ‘pre-renal’ causes are much more prevalent than in general hospital admissions.^1,9 The low frequency of hydronephrosis may also be an effect of initial treatment and resuscitation: all patients had a urinary catheter placed as part of their critical care management, which should rapidly cause resolution of upper renal tract dilatation when the cause of AKI is lower urinary tract obstruction.

In this study, a modern Sonosite EDGE ultrasound machine was used. In high-resource settings, the quality of ultrasound technology and image quality available in a hospital radiology department would exceed that which we used for our scans. Notwithstanding this, the images acquired were sufficient for the purposes of the basic CUSIC assessment of the renal tract. Image quality generally improved through the study, as ICU physicians became more adept at finding the best acoustic windows and at optimising the technical quality of the ultrasound images. We suggest that introduction of CUSIC-level abdominal ultrasonography should not mandate a critical care department to update their ultrasound technology, but close liaison with radiology colleagues (who may also be CUSIC supervisors) with regard to the assessment of image quality and procurement of ultrasound technology is recommended.

Duration of scan time shortened with increasing experience, with a median time for scanning and archiving images for practitioners with over 10 scans completed being just 5 min. Focussed renal ultrasound on ICU should be quick, with no change in patient positioning made for any of the scans, to ensure that the patient’s nursing care is not affected nor is the patient exposed to further discomfort.

There has recently been guidance published on number of scans deemed adequate for competency,¹¹ and this study suggests 15 scans as a minimum is adequate for achieving competency with image acquisition as well as duration of scan. However, one must bear in mind that in our study only one case of hydronephrosis was observed in 30 patients, in keeping with the incidence of hydronephrosis in AKI in hospitalised patients.¹² Other studies in critical care settings have identified an even lower prevalence.⁹ This means that the trainee intensivist may not obtain sufficient exposure to this finding within even 30 scans. Considering that a primary reason to perform focused renal ultrasound on the ICU is to detect hydronephrosis, the low incidence of this condition may hamper efforts to develop and demonstrate competency. The sub-optimally positioned patient and use of a basic ICU ultrasound machine may present an issue with regard to recognition of this finding, particularly in those patients with lesser grades of hydronephrosis which can be difficult to identify. On the other hand, extending the minimum number of scans to more than 30 may present challenges with regard to training time and will also depend on the size and case-mix of an individual ICU. These challenges have been surmounted with training in echocardiography, where libraries of archived images and cine clips are available to allow learners to experience the range of potential pathologies. We suggest both regular peer-review of ultrasound images for the different degrees of hydronephrosis, possibly in conjunction with the local radiology department as a joint radiology meeting, in order to raise awareness of this finding for those intensivists who wish to perform focussed renal tract ultrasound. Departments should keep a store of anonymised images to facilitate training and retention in competency.

This study was performed by two intensivists with experience of ultrasound and echocardiography. We aim to extend this study by training Intensive Care Medicine trainees with no/limited experience of renal ultrasound and assessing the feasibility of this, with regard to how long it would take to train the trainee to a competent level, with their scans being reviewed and critiqued by a radiologist in a similar way.

We aim to put forward our recommendation to CUSIC that 15 scans performed independently as a minimum should be obtained in order to achieve competency for focused renal tract ultrasound on the ICU. We also recommend the use of colour Doppler in those cases where hydronephrosis is thought to be present to differentiate blood flow in vessels from true hydronephrosis.

In this study, only one case of hydronephrosis was identified and was graded. Clearly recognising this pathology is crucial for the scanning intensivist, although the grading can prove difficult. Indeed there can be discrepancies amongst radiologists regarding this, and we suggest grading of hydronephrosis should not be a core competency for focused renal tract ultrasound on ICU, but rather the presence of hydronephrosis should be.

We feel that focused renal ultrasound on ICU is an achievable goal for ICU physicians and trainees. As with all forms of ultrasound, the initial learning curve can be steep, although as this study showed, with increasing experience, the duration of scan length significantly shortens making this a practical and useful aid to daily working practice.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.