Hyponatraemia: common in elderly in-patients.

Clinical bottom line (level 4)

  1. Less than 1% of elderly patients admitted to hospital had hyponatraemia. 20% of patients developed hyponatraemia during their admission.
  2. Patients were at increased risk of hyponatraemia if they had:
    • hepatic failure
    • GI losses
    • sweating
    • diabetes
    • ascites
  3. Patients were at increased risk of hyponatraemia if they took:
    • NSAIDs
    • potassium-sparing diuretics
    • trimethoprim/sulphamethoxazole
Rosenblatt et al: Journal of Geriatric Drug Therapy 1996; 11 (2): 71-84
Expires July 2005

The study

Case-control study with objective outcomes, not adjusted for confounding factors, not validated in an independent set of patients.

Setting: university hospital, USA

847 patients (aged mean 71 years, ?% male) older than 60 years and admitted to hospital

Excluded if
  • no baseline serum sodium on admission
  • on dialysis


    • Cases: patients (% male, mean age ):
    Controls: ?patients :


    Outcomes studied:
  • hyponatraemic during admission
  • hyponatraemic on admission

    • The evidence

    Patient expected event rate for hyponatraemic during admission: 20.2%
    risk factor for
    hyponatraemic during admission     		unadjusted OR
    (95% CI)    		 NNH
    (95% CI)
    sweating 5.39
    (2.40 to 12.1)     		13
    (6 to 45)
    hepatic failure 2.91
    (1.63 to 5.21)     		11
    (5 to 35)
    ascites 2.63
    (1.12 to 6.17)     		29
    (9 to 397)
    GI losses 1.99
    (1.27 to 3.11)     		9
    (4 to 33)
    diabetes mellitus 1.62
    (1.09 to 2.41)     		9
    (4 to 63)
    trimethoprim/sulphamethoxazole 3.51
    (1.54 to 7.97)     		21
    (8 to 97)
    NSAIDs 1.86
    (1.10 to 3.14)     		15
    (6 to 130)
    potassium-sparing diuretics 1.93
    (1.13 to 3.30)     		15
    (6 to 108)

    Patient expected event rate for hyponatraemic on admission: 0.9%
    • Na (mmol/l):
      • 130-134: 129 (75%; 95% CI: 69% to 82%)
      • 121-129: 37 (22%; 95% CI: 16% to 28%)
      • <121: 5 (2.9%; 95% CI: 0.4% to 5.4%)

    Comments

    1. Lack of distinction between sever (<125 mmol/L) and mild cases makes the results less useful.
    2. Failure to adjust for confounding factors makes these results much less certain.

    Citation

    1. Rosenblatt DE, Opdycke RA, Kustrzyk BM, et al: Incidence of hyponatremia in elderly patients on hospital admission. Journal of Geriatric Drug Therapy 1996; 11 (2): 71-84
    Search Terms: reference in review article
    Contributor: Chris Ball and Musab Hayatli, July 2000
    Reviewer: Malcolm Man-Son-Hing

    Clinical Question.
    Patient
    Intervention or Exposure
    Outcome