Article Text

Download PDFPDF
Case report
Point-of-care ultrasound to evaluate volume status in severe hyponatremia
  1. Connor Evins1 and
  2. Aniel Rao2
  1. 1Department of Internal Medicine, University of South Carolina School of Medicine Greenville Campus, Greenville, South Carolina, USA
  2. 2Department of Internal Medicine, Prisma Health, Greenville, South Carolina, USA
  1. Correspondence to Dr Aniel Rao; anielrao{at}gmail.com

Abstract

A 51-year-old man was hospitalised for severe hyponatremia. Initial history and physical examination suggested hypovolemia, and he was treated with normal saline at 100 mL/hour. After several days, his hyponatremia failed to improve, and then worsened without resolution of presenting ataxia and fatigue. He had no new complaints including no cough or orthopnea. He had no jugular venous distention or oedema, and his lungs were clear to auscultation. Point-of-care ultrasound was used, revealing a distended inferior vena cava, pulmonary oedema and pleural effusion, suggesting hypervolemia. Based on ultrasound findings, we treated with 60 mg oral torsemide two times per day. Hyponatremia resolved without complication within 48 hours. In this case, physical examination failed to recognise volume status change from hypovolemic to hypervolemic, increasing hospitalisation and morbidity. The point-of-care ultrasound proved to be an accurate tool for proper volume evaluation, and may be used as an adjunct to physical examination for hyponatremic patients.

  • ultrasonography
  • fluid electrolyte and acid-base disturbances
  • general practice / family medicine

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Background

Hyponatremia is a common electrolyte abnormality representing a decreased sodium concentration below 135 mmol/L that when defined in this way is present in ‘up to 8% of the general population, and up to 60% of hospitalised patients’.1 More specifically, it is estimated to be affecting ‘4% of patients presenting to emergency departments and up to 30% of patients in general medicine wards’.2 Hyponatremia requires urgent correction and treatment, as it has been associated with a 30% higher risk of death and 14% longer hospital stays, even when accounting for comorbid conditions.1 3 To further complicate this matter, severe complications can occur with rapid correction of hyponatremia such as central pontine demyelination syndrome especially if the patient has chronic hyponatremia.4 These poor outcomes, both from the abnormality and from the correction of the abnormality, combined with the frequency of its development have led to the development of several protocols for evaluation, diagnosis and treatment. However, current guidelines often differ from one another further showing the difficulty of its management.1

One of the first steps in identifying the cause of hyponatremia is ruling out pseudohyponatremia (isotonic hyponatremia) and hypertonic hyponatremia. Pseudohyponatremia can be caused by massive elevations in lipids or proteins, which can lead to laboratory error in serum sodium measurement due to these larger molecules occupying more plasma volume.4 These particles do not change the amount of blood solute, and therefore plasma osmolality will be normal in these cases. In hypertonic hyponatremia, effective solutes other than sodium (such as glucose and mannitol) are present in the plasma causing an osmotic fluid shift from the intracellular to extracellular space, which can lead to an appearance of a low sodium concentration.4 In these cases, plasma osmolality will be elevated. Inability to rule out isotonic and hypertonic causes of hyponatremia are two of the major pitfalls associated with diagnosis.5

The most common presentation of hyponatremia is a patient who has hypotonic hyponatremia. By definition, this means there is a decreased number of total osmoles relative to blood volume. This can be caused either by excessive loss of sodium and other electrolytes, or by elevation in free water in the blood.4 Determining which of those things caused the abnormality relies heavily on laboratory tests such as the urine osmolality, urine sodium and serum uric acid levels which can provide clues to the patient’s diuretic use and antidiuretic hormone status.5 Along with these laboratory tests, the primary diagnostic step is determining the patient’s volume status, so as to categorise and treat the abnormality appropriately.4

Volume status is difficult to determine however, and all 10 clinical practice guidelines referenced in a 2014 study suggested that it is primarily analysed through clinical evaluation and physical examination.1 Unfortunately, physical examination has been shown to be a poor indicator of both volume status and volume responsiveness. A 2013 study showed ‘findings on physical examination correlate poorly with true volume status’.6 This study used an invasive pulmonary catheter to measure Global End-Diastolic Volume Index, and compared these findings with specific clinical evaluation.6 Specifically, they found poor correlation with findings of dry mucous membranes, dry axillae and decreased skin turgor for a diagnosis of hypovolemia.6 Delayed capillary refill and decreased mean arterial pressure also showed low sensitivity for hypovolemia.6 Another study of cardiologists managing heart failure showed that jugular venous distention had only an 81% predictive accuracy for assessing volume overload, and absence of rales was only 35% negatively predictive.7 Other estimates show that visibility of jugular venous pressure is not visible up to 80% of the time by resident physicians, and rales on lung auscultation recognise hypervolemia less than 50% of the time by resident physicians, regardless of level of training.8 In a 2014 survey, resident physicians of all levels of training were on average self-rating their own confidence in assessing volume status via physical examination to be approximately 3.5 out of a score of 6.8 They also stated that a major barrier to their confidence is conflicting examination findings with colleagues.8 These studies suggest that both resident and attending physicians, regardless of amount of time spent in training, are consistently inaccurate in measuring both hypervolemia and hypovolemia using physical examination findings.

Studies also suggest the use of ultrasound to help improve analysis of volume status by physical examination, in particular the use of point-of-care ultrasound serving as an adjunct to physical examination. Point-of-care ultrasound is the use of mobile ultrasound technology, often at the patient’s bedside, to be interpreted directly by the physician.9 As stated by Choi et al, it has been used frequently in intensive care or emergency department settings to confirm differential diagnoses, and further research is still needed to determine its total scope of use.9 As the results and examination are performed directly by the physician, it decreases the time, cost and potential miscommunication involved in ordering a formal ultrasound via a radiology department. It is also a non-invasive test that decreases risk to the patient. Another study that uses point-of-care ultrasound in the critical care setting is Pourmand et al, which specifically used ultrasound to analyse large veins such as the inferior vena cava to guide volume resuscitation.10 This study was evaluating for fluid responsiveness during urgent volume resuscitation, and recognises and validates the use of large veins for volume status changes while giving intravenous fluids.10

One validated method for checking volume status on a large vein is measuring the Caval Index on the inferior vena cava. The Caval Index is defined as the difference between the inspiratory and expiratory inferior vena cava diameters, divided by the expiratory diameter.11 Pasquero et al used healthy blood donors to establish a baseline Caval Index, and then measured this against a hypovolemic hyponatremic population.2 They discovered that clinical evaluations of these patients were not correspondent with the ultrasound evaluations in approximately 50% of cases, with the clinical evaluations being more likely to misdiagnose hypovolemia.2 In this study, hypervolemic causes of hyponatremia were not used, which is a major limitation given the case described below. However, it still supports the idea that using inferior Vena Caval Index can help determine volume status in hyponatremic patients if used in conjunction with physical examination findings.

This case serves to highlight the use of point-of-care ultrasound as an adjunct to physical evaluation, laboratory and other imaging studies to improve volume status assessment on an individual with severe hyponatremia refractory to initial treatment.

Case presentation

The patient is a 51-year-old white man with a medical history significant for severe coronary artery disease resulting in cardiogenic shock (at one point requiring a short-term left ventricular assist device), heart failure with reduced ejection fraction estimated at 20%–25%, hypertension and schizophrenia who presented with his brother to the emergency department with 5-day history of fatigue, ataxia, altered mental status and abnormal behaviour.

His brother states that he has been getting steadily worse, with multiple near falls and unsteadiness in 2 days leading up to admission. He has never had any symptoms like this before, and has poor insight into his symptoms. His behaviour and mental status have also been inappropriate, as the brother states he has been unable to complete activities of daily living such as bathing or feeding himself. His affect has also been more flat than normal, with the patient’s brother reporting significantly longer time to respond to questions than his baseline. He also reported one episode of hyper elation and agitation that resolved after several hours. He has been alert and oriented to only person and place, but not time (normally he is fully oriented). The patient denies any shortness of breath, chest pain, nausea, headache, vomiting, diarrhoea or decreased oral intake. He and his brother deny any trauma. The patient denies any hallucinations, delusions, depressed mood or worsening negative symptoms of schizophrenia. His medication regimen is notable for spironolactone 25 mg daily, torsemide 200 mg two times per day and metolazone 2.5 mg daily for diuresis. This large dose of diuretics was due to a recent hospitalisation for congestive heart failure exacerbation with volume overload. It was during this hospitalisation that torsemide was increased, and spironolactone was added with metolazone. His other home medications included atorvastatin 80 mg daily, clopidogrel 75 mg daily, hydroxyzine 25 mg as needed for itching, mirtazapine 30 mg daily, olanzapine 10 mg daily, trazodone 50 mg as needed for sleep, metoclopramide 5 mg as needed for nausea and pantoprazole 40 mg two times per day for heartburn. While the patient does have a history of poor medication adherence and health literacy, his brother states that he has been taking his medications as prescribed leading up to the onset of these symptoms.

The patient had a blood pressure of 91/73 mm Hg on admission, with his normal baseline systolic blood pressure being around 100 mm Hg. His pulse, temperature, oxygen saturations and respiratory rate were within normal limits. On physical examination, he was generally in no acute distress. He was alert and oriented, with no focal neurologic deficit, and a flat affect. He had very dry mucous membranes, decreased skin turgor and no oedema. His cardiac examination was notable for an irregular rhythm, no appreciated jugular venous distention with equal peripheral pulses. His pulmonary examination was overall clear to auscultation. Abdomen was non-distended, with normal bowel sounds.

Basic metabolic panel on admission was notable for a serum sodium of 108 mmol/L, potassium of 1.5 mmol/L, chloride of 65 mmol/L, blood urea nitrogen (BUN) of 82 mg/dL and a creatinine of 2.32 mg/dL (baseline 1.2 mg/dL). His serum osmolality was 242 mOsm/Kg H2O. His urine sodium was 20 mmol/L, urine potassium 25.6 mmol/L, urine chloride 20 mmol/L and urine osmolality 204 mOsm/kg H2O. His brain natriuretic peptide (BNP) was 2113 pg/mL, which is around his baseline. His glucose on admission was 119 mg/dL. His most recent lipid panel from 8 months prior had a total cholesterol of 263 mg/dL, high-density lipoprotein (HDL) of 37 mg/dL, low-density lipoprotein (LDL) of 183 mg/dL and triglycerides of 217 mg/dL.

Notable imaging includes a non-contrast CT head that was unrevealing for acute abnormalities, and his ECG was found to have no changes from the previous studies. In the setting of the patient’s presenting symptoms and severe hyponatremia, he was admitted to the internal medicine teaching service.

Outcome and follow-up

Given his significant electrolyte abnormalities, prerenal azotemia, extensive and recently advanced home diuretic regimen and hypovolemic physical examination, the patient was assessed to be in severe hypotonic hyponatremia due to hypovolemia from over-diuresis as the cause of his presenting symptoms. In concordance with cardiology consultation, he was treated with gradual sodium correction with isotonic fluids. Normal saline was introduced at 100 mL/hour, with sodium checks in every 4 hours to avoid correction greater than 8–10 mmol/L in 24 hours. His home diuretics were held while correcting sodium. We also held his other home medications, including his schizophrenia medication olanzapine due to the risk of syndrome of inappropriate antidiuretic hormone (SIADH), which may worsen his hyponatremia. In addition, with the patient’s hypokalemia requiring aggressive repletion, we were concerned with continuing his olanzapine and its inherent risk of QT-interval prolongation. His electrolytes improved on this therapy, with a peak in sodium to 119 mmol/L on the morning of hospital day 4. However, by that afternoon, his sodium actually decreased to 116 mmol/L. At this point, his potassium and chloride had corrected, and his creatinine returned close to baseline. His presenting symptoms of ataxia, altered mental status and fatigue had slightly improved. He was able to walk around the room with supervision and had no falls. He was awake throughout the day and did not report fatigue. However, given that he was a poor historian with a very flat affect secondary to his schizophrenia history, it was difficult to ascertain his insight into his symptoms. It was also difficult to tell if his negative symptoms had worsened while holding olanzapine, but his brother confirmed that he was around his baseline in this regard. While he was oriented to person, place and time, his lack of insight and inability to describe his concerns remained a challenge throughout the hospitalisation.

Following his initial presentation and volume repletion, his physical examination, symptoms, labs and imaging supported an euvolemic aetiology of his hyponatremia. He remained clear on pulmonary auscultation and did not have any abdominal distention, jugular venous distention or lower extremity oedema. He also did not report any cough, orthopnea, or paroxysmal nocturnal dyspnoea. His BNP had decreased to 1837 pg/mL. A chest radiograph showed right basilar atelectasis with possible infiltrate, but was mostly stable compared with baseline. In conjunction with an euvolemic hyponatremia, an evaluation for SIADH was started. This was not only supported by his current examination status, but he was also being treated with olanzapine prior to admission, which has a risk for SIADH. His intravenous fluids were stopped and he was placed on a 1000 mL/day fluid restriction. However, his hyponatremia failed to significantly improve with a value of 117 mmol/L on the morning of hospital day 5, and 118 mmol/L on the morning of day 6. His urine electrolytes included a urine sodium of less than 20 mmol/L, urine potassium 26.9 mmol/L, urine chloride 23 mmol/L and urine osmolality 315 mOsm/kg H2O. His low urine sodium suggested that SIADH was not the aetiology of his hyponatremia. He remained clinically euvolemic on physical examination on hospital day 6, including normal pulmonary examination to auscultation, no jugular venous distention, no oedema, normal capillary refill and no abdominal distention. He also did not report cough, orthopnea or paroxysmal nocturnal dyspnoea, and continued to state that he had been resting comfortably every night.

In the setting of likely mismanagement of this patient’s hyponatremia, the primary inpatient teaching service used point-of-care ultrasound on the morning of hospital day 6 to better assess the patient’s volume status. The patient was found to have a moderate-sized right pleural effusion and diffuse pulmonary parenchymal B-lines suggesting pulmonary oedema. His inferior vena cava appeared plethoric and its diameter measured 2.5 cm with no respiratory variation. His estimated Caval Index was 1, given this lack of respiratory change, which is significantly higher than the estimated 0.35 in euvolemic patients.2 In all, the ultrasound findings supported hypervolemia. As a result, a diuretic regimen was started with oral torsemide 60 mg two times per day and his volume response and sodium were monitored closely. His sodium improved to 126 mmol/L on hospital day 7, and then he was successfully discharged home with a sodium of 132 mmol/L on hospital day 8. He was stable at time of discharge, with discharge summary stating to monitor volume status closely in upcoming appointments. He was also counselled on medication adherence, frequent weight monitoring and fluid restriction. Home health services were called prior to his discharge to discuss assisting his mother and brother with his care.

While the internal medicine teaching service and the cardiology service recommended discharge to skilled nursing facility to help manage his medications and chronic debility, he wanted to be discharged home to his mother and brother. He followed up with his cardiologist 2 weeks after discharge, and was scheduled with his primary care physician but never arrived at that appointment. At the time of his cardiology follow-up, his diuretic regimen was torsemide 100 mg two times per day (down from 200 mg two times per day), spironolactone 25 mg daily (same as prior to hospitalisation) and metolazone 2.5 mg daily (same as prior to hospitalisation). In setting of his heart failure with reduced ejection fraction, he was started on 25 mg daily of metoprolol succinate. He had also resumed his home medications that were held during the hospitalisation, most notably olanzapine 10 mg daily and mirtazapine 30 mg daily for his schizophrenia. He had no complaints of fatigue, altered mental status, ataxia or abnormal behaviour. His sodium at this visit was 130 mmol/L, somewhat stable since discharge. Medication non-adherence remained an issue at follow-up, with patient having little understanding of his medical history or current medications. Home healthcare services assisted with medication adherence.

Discussion

This particular case is an example of the complexities and limitations of physical examination, imaging and laboratory values as an evaluation of volume status when managing hyponatremia. This patient initially presented with presumed hypovolemic hyponatremia in the setting of his recent advancement in diuretic regimen for his heart failure. He was given intravenous fluids for his hypovolemia with initial improvement in sodium values. However, his sodium failed to normalise and his physical examination (lack of jugular venous distention, lower extremity oedema, pulmonary auscultatory changes or clinical respiratory changes) incorrectly supported an euvolemic hyponatremia. Chest radiograph also incorrectly supported euvolemia. From hospital day 3 to 6, the patient received improper treatment for his hyponatremia despite thorough physical examinations and imaging studies. This was an unnecessary cost and lengthened the patient’s hospital length-of-stay. The use of point-of-care ultrasound to discover a pleural effusion, pulmonary parenchymal B-lines and a plethoric inferior vena cava directly led to a change in the patient’s hyponatremia mismanagement. The patient quickly improved with appropriate diuresis for his previously undiscovered hypervolemic hyponatremia.

Given his speed of recovery with the appropriate diagnosis, and the validated use of point-of-care ultrasound to measure Caval Index to serve as an adjunct to physical examination in determining volume status, this case serves to highlight how point-of-care ultrasound can be used in the setting of hyponatremia. As described in the Introduction section, physical examination skills of physicians can be limited in determining volume status, as in some cases the changes are subtle. In this case, even chest radiography did not support his volume change. Point-of-care ultrasound is time efficient and non-invasive, and has been shown to potentially decrease total cost of care to the patient. Previous studies have used Caval Index to analyse hypovolemic hyponatremia.2 However, our case suggests that it can be used to help determine hypervolemia as well, as an elevated Caval Index helped us in transition therapy to drastically improve our patient’s hyponatremia. As the technology of ultrasound devices continues to improve to allow for ease of transport during inpatient rounds, and as physicians continue to train in both performing and interpreting this examination, it is a useful tool to be used in cases of hyponatremia where volume status is difficult to determine.

Learning points

  • There are shortcomings of the physical examination for evaluation of volume status in assessing hyponatremia.

  • Point-of-care ultrasound can be used as an effective adjunct to physical examination, laboratory and other imaging studies to evaluate volume status, and thereby guide effective therapy for severe hyponatremia.

  • Future volume status assessment for patients with complicated hyponatremia would benefit from the use of point-of-care ultrasound.

References

Footnotes

  • Contributors AR and CE contributed to the planning, conduct and reporting of the work described in the article.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Next of kin consent obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.