Postoperative Hyponatremia in Menstruant Women

TO THE EDITOR:

Ayus and colleagues [1] correctly emphasize the potential dangers of acute postoperative hyponatremia. No one should dispute that neurologic symptoms warrant a set of electrolyte studies, that symptomatic hyponatremia deserves urgent treatment, or that hypotonic fluids are best avoided in the postoperative period. Future editions of surgical textbooks should stress these important points. Unfortunately, the authors also leave the impression that hyponatremia in young women is associated with a high mortality rate. This conclusion is not supported by their findings.

The only valid mortality statistics are found in the authors' survey of 674 consecutive patients with postoperative hyponatremia. There were 39 menstruant women in this unbiased sample and none of them died, developed brain damage, or even became encephalopathic. Contrast this experience with the authors' 65 case patients, who were referred to them because of hyponatremic encephalopathy. Of the 28 menstruant women in this sample, 82% died and nearly half the survivors developed brain damage. What explains the extraordinarily high morbidity and mortality rates in the case patients?

Ayus and colleagues tell us that 32 of the 65 case patients were from hospitals affiliated with either Baylor University or the University of California and that 33 were from other institutions. Is it a coincidence that female case patients who died or had permanent brain damage also numbered 33? I hope not. Iatrogenic deaths from postoperative hyponatremia should be vanishingly rare in university teaching hospitals.

In previous reports of postoperative brain damage caused by hyponatremia, Dr. Arieff noted that cases were collected over a period of several years from multiple hospitals (one hospital per patient) and that patients were referred to him by physicians, family members, and lawyers [2, 3]. Thus, it seems likely that many of the case patients in the present series were referred to the investigators because the patients had fatal brain damage and not just because they had symptomatic hyponatremia. Obviously, mortality statistics in a patient sample with these selection criteria are uninterpretable and the sample's demographics merely reveal what is most likely to be referred to a national expert for review; it is easy to imagine why young patients would predominate.

All we should conclude is that some patients with postoperative hyponatremia suddenly die with serum sodium concentrations that most other patients tolerate. How often this occurs and why the victims are usually female are subjects for further study.

• Copyright 2004 by the American College of Physicians