PREDICTORS OF OUTCOME IN HOSPITALIZED PATIENTS: DISCUSSION

Posted by James

The major aim of this study was to assess the outcome of severe hyponatremia and characterize factors influencing outcome in hospitalized patients with this disorder. This study suggests that severe hyponatremia is associated with high mortality. It is unclear if this association derives from hyponatremia per se, or the associated comorbidity in these subjects.

One of the major patient characteristics associated with increased mortality was the presence of symptoms. Our data fails to confirm a previous association of hyponatremia in general with increased mortality in HIV-infected patients, CHF, thiazide diuretic use and psychogenic polydipsia. Our prevalence data of 2.8% supports previous observations showing that severe hyponatremia is indeed quite rare. A recent study found an incidence of 1% among South African patients. Even though mortality did not differ across etiologic groups, morbidity differences were observed between etiologic groups of patients. While morbidity was 100% in patients with psychogenic polydipsia, mortality was zero in this group.
erectalis 20

Patients in sickle-cell pain crisis who developed acute hyponatremia following rapid administration of 5% dextrose in water, therapy with morphine and NSAIDS (for pain control) also had 100% survival with only 10% morbidity. While mild, induced hyponatremia appears beneficial in therapy of sickle cell crisis, as it causes swelling of red blood cells which helps correct the sickling effect, our data shows that this therapy, when combined with morphine and NSAIDS which impair water excretion, can cause severe hyponatremia. This study is the first to document the association between therapy for sickle-cell crisis and the occurrence of severe hyponatremia.

Another major finding of this study was the association of slow rates of correction with mortality. Our data supports previous observations by Ayus et al. that a slow correction rate (<0.7mmol/L/hour) was associated with high mortality. Overall, our study revealed a reluctance to use hypertonic saline in symptomatic patients with severe hyponatremia. The reasons for this reluctance are unknown. One possible explanation could be a desire to avoid overly rapid correction. Another explanation could be the high proportion of patients in our study population with hypovolemia, in whom normal saline may be considered adequate therapy. However, this would be an inadequate explanation, as normal saline also was prescribed to symptomatic patients with SIADH, leading to worsening of hyponatremia, and uncal herniation in one patient with tuberculous meningitis. The safety of hypertonic saline in symptomatic patients is well established.
prescription drugs from canada

Our finding of an association between hypoxia and mortality in hyponatremia has been previously observed by others. The precise basis for this association between hypoxia and mortality in hyponatremia is unclear, but some authors suggest that hypoxia plays a critical role in the genesis of hyponatremic brain injury in experimental animals. Vexlar et al. suggest, in fact, that hypoxia may interfere with cerebral adaptive responses to hyponatremia. Our definition of hypoxia in this study was fairly liberal, and yet this association remained very strong. Recently Ayus et al. reported on the occurrence of pul­monary edema in marathon runners with hyponatremia. In our series, pulmonary edema was found in only 25 patients (14.8%). Death in one patient was clearly associated with overly rapid correction. Interestingly, this patient was treated with fluid restriction alone. That fluid restriction alone can lead to overly rapid correc­tion has been reported by other investigators. This suggests that clinical vigilance is still vital when a severely hyponatremic patient is man­aged with fluid restriction alone.

This study also revealed an association between sepsis and increased mortality in severe hyponatremia. This was intriguing, since none of these patients were in frank septic shock, or had evidence of multi-organ systems failure at the outset. The mechanism by which sepsis would increase mortality with severe hyponatremia is unknown. Some data suggest that alterations of arginine vassopressin levels may occur in sepsis, though results have been variable depending on the phase of sepsis studied. Others have previously noted the association of sepsis with increased mortality in hyponatremia.

One possible explanation for the association between sepsis and increased mortality in hyponatremia is the interference with the Na⁺K⁺ ATPase, an enzyme involved in early cerebral adaptation to hyponatremia. Furthermore, there are suggestions that hyponatremia during sepsis is associated with progressive failure of the energy-dependent transport involving this enzyme. canadian pharmacy generic viagra

Though our study has the unique strength of being one of the largest studies of outcome of severe hyponatremia in hospitalized patients, it has some important limitations. Firstly, the retrospective, observational nature of our study, and the sample size limits the robustness with which we can analyze the data, and hence limits the power of some of our observations. Second, by evaluating patients with severe hyponatremia, an unintended element of bias is introduced. Third, our patients had a predominance of acute severe hyponatremia and the applicability of our findings to the general population of patients with hyponatremia is unknown. In spite of these limitations, our findings are noteworthy, and prospective studies would be needed to validate these findings. In conclusion, this study has shown that severe hyponatremia in hospitalized patients is associated with high mortality. Hypoxia, symptoms of encephalopathy, slow correction rates and sepsis appear to increase mortality.