I am a hospitalist at a small rural hospital, and I’ve read both the hyponatremia and SIADH articles published in The Hospitalist. Our lab does not do any urine testing beyond a UA [urinalysis] in house, so the urine osmol and urine Na+ tests are send-outs, which take several days to come back. I’m having difficulty with diagnosing the reason for and treating hyponatremia. I find it complicated, and when the urine tests are not readily available, it’s difficult to use the algorithm. Do you know of any basic tips that could help? If our hospital does not have the right urine tests readily available, should I be sending these patients to the larger hospital from the ED if sodium is <125, instead of admitting them?
—Carleigh Wilson, DO
Dr. Hospitalist responds:
I too can recall moonlighting in a small rural hospital in southern Georgia 25 years ago; I remember having to improvise when taking care of patients with hyponatremia. Fortunately, even though we have developed more sophisticated equipment to help in the diagnosis (e.g. electrodes that are not hampered by excess triglycerides or proteins), my basic approach to evaluating these patients hasn’t changed much.
I still begin with an assessment of the patients’ serum osmolarity, which is usually low in most hospitalized patients. If it is elevated, then, of course, hyperglycemia is the most common cause, but we also must consider alcohol or, way less common these days, mannitol. Hypoosmolar hyponatremia most often occurs when the kidneys are overwhelmed by the intake of water and cannot excrete it as free water. Even though it can occur when there is pathologic consumption of large amounts of water (psychogenic polydipsia) or excess consumption of beer, it most often occurs when the kidney fails to fully dilute the urine, which would be evidenced by urine osmolality >100mmol/l.
A good history and physical will help categorize the patients with poorly diluted urine; in the absence of measured urine osmolality, however, urine specific gravity (USG) can be used with some caveats. Studies have shown that USG done by refractometry and reagent strip (both very inexpensive) have a correlation of 0.75-0.80 with urine osmolality. Although there are many variables affecting both (pH, ketones, glucose, urobilinogen, bilirubin, and protein for the reagent strips), I would use the refractometer, which only seems to be affected by bilirubin, ketones, and hemoglobin. So, at a pH of 7, with all the variables considered and a USG of 1.010, predicted osmolality is approximately 300 mosm/kg/H20. Also, while osmolarity and specific gravity change in parallel, the two measurements diverge when there are large particles in solution (e.g. glucose or proteins), so be careful … Dr Kokko would be so proud of me!
The next step is to determine the patients’ volume status. Because I’m old school, I still believe this can be done at the bedside. On most occasions, it is only when I’m trying to decide whether the primary sodium loss in hypovolemic patients is due to a renal or extra-renal cause that the urine sodium is helpful. In truly hypovolemic and asymptomatic patients, I usually start with normal saline (NS) to correct the volume status and follow the serum sodium closely to avoid correcting too rapidly. Urine sodium is not particularly helpful in caring for euvolemic and hypervolemic hyponatremic patients.
Due to multiple variables (e.g. acute vs. chronic, co-morbidities, nursing and lab support, quick and safe transfer to higher level of care), it is difficult to quantify a “safe” level of hyponatremia to treat at a small rural hospital. Considering the clinical variables and presence of symptoms, I probably would not be comfortable with a serum sodium less than 123 meq/L. However, you have to understand and appreciate your limitations and develop your own level of comfort.
Do you have a problem or concern that you’d like Dr. Hospitalist to address? Email your questions to firstname.lastname@example.org.