Retrospective evaluation of fluid overload and relationship to outcome in critically ill dogs.

Objective - To determine if critically ill dogs have an increased risk of fluid overload (FO) during hospitalization compared to less ill dogs, and to determine if FO is associated with increased mortality during hospitalization. Design - Observational, case-control study. Setting - University teaching hospital. Animals - Thirty-four critically ill dogs and 15 comparatively healthy stable postoperative dogs with neuro-orthopedic disease. Interventions - None. Measurements and Main Results - Data recorded included underlying disease, body weight, and APPLE(fast) score at admission, single-day and composite APPLE(full) scores during hospitalization, total fluid volume administered (L), total fluid volume output (L), and outcome. Percent FO (%FO) was calculated using the equation 100 x ([fluid volume administered - fluid volume lost]/1000 mL/L) - (% dehydration at admission), with fluid volume expressed as mL/kg of baseline body weight. Critically ill dogs developed greater %FO during hospitalization compared to stable postoperative dogs (12.1 +/- 11.7% vs 0.5 +/- 5.2%, P = 0.001), and half (8 out of 16) of the dogs with %FO >= 12% died. Composite APPLE(full) scores were weakly positively correlated with %FO, whereas APPLE(fast) and single-day APPLE(full) scores recorded at admission were not. The odds ratio for death was 1.08 for every percent increase in FO during hospitalization (95% confidence limits 1.012-1.59, P = 0.02). Conclusions - Critically ill dogs are at increased risk for FO during hospitalization, and a weak but significant association exists between %FO, illness severity, and mortality. Prospective studies are warranted to confirm the findings of this retrospective study.