Is Volume Toxic? Part 2

Jul 06 2011 Published by under Journal Club

Over at the blogspot place in May 2010 I reviewed an article for journal club showing that children requiring continuous renal replacement therapy did worse with greater volume overload at the start of therapy. Query: Is saline toxic?

The current New England Journal of Medicine reports a trial of children in Africa who presented with a febrile illness and impaired perfusion who were randomized to first receive a bolus of saline or albumin or to proceed directly to maintenance fluid therapy. Bolus therapy is given to rapidly restore circulating volume. Children with hypotension (low blood pressure) were assigned to either bolus group, but not the maintenance fluid group. Children with gastroenteritis were excluded from the study; 57% of participants had malaria. The primary endpoint was 48 hour mortality, with secondary endpoints including pulmonary edema; increased intracranial pressure; 4 week mortality; or 4 week neurologic sequelae.

The trial stopped ~500 patients before the initial estimated enrollment on the advice of the Data and Safety Monitoring Board (DSMB). In stratum A (patients with impaired perfusion but without frank hypotension) bolus therapy of either type increased 48 hour mortality (Relative risk 1.45; 95% confidence interval, 1.13 to 1.86; p=0.003) and mortality at 4 weeks (12% vs. 8.7%; p=0.004). No other secondary endpoints showed any differences. Stratum B (patients with frank hypotension) included only 29 patients, and no difference between albumin (69% mortality) and saline (56% mortality; p=0.45) was demonstrated.

Impaired perfusion can be diagnosed by physical exam. Prolonged capillary refill time (after skin compression, blanching should resolve in 2-3 seconds; see figure) provides the most common assessment of this state, although cyanotic extremities may also be a sign. In the presence of normal blood pressure, bolus therapy appears to be contraindicated in children with these signs, although generalizing the findings from this study of children primarily affected by malaria may not be valid to children in countries such as the US where the underlying conditions may be different.

Other questions arise. First, do boluses help children with frank hypotension? Perhaps they might do better with pressors rather than volume expansion? What about children with gastrointestinal volume losses, a common cause of shock? A recent Clinical Pathological Conference from NEJM demonstrated the apparent benefits of intravenous fluid boluses in a child with cholera. Are these benefits real for children with all gastrointestinal illnesses?

Volume expansion, with saline or colloid, has been a mainstay of therapy for as long as I can remember. Restoration of circulating volume has a place; however, in other cases, additional volume could be toxic. Further studies are needed to define the risk:benefit ratio and conditions of fluid resuscitation.

Performing these studies in the US may be extremely difficult. First, getting informed consent for critically ill children can be fraught with difficulty, especially when one might get randomized to nonstandard care. Do I want to take that chance with my child's life on the line? For a condition like infant diarrhea that is far less likely to lead to death in this country than in Africa?

I am afraid that the initial exploration of these questions will have to happen in other countries with fewer resources and laxer protection of research subjects. Perhaps once we have more information we can perform a meaningful study in a "first-world" country.

2 responses so far

  • D. C. Sessions says:

    Ok, so I have a thing about blood volume. Sue me.

    Does anyone have a hypothetical mechanism for this effect yet? Anything more suggestive than differential survival?

    • whizbang says:

      There is utter speculation that volume may turn off some beneficial adaptive process in nonhypotensive normovolemic shock. Duh. Bottom line: we have no idea why these kids did better without the bolus.

      It is important to remember that children with gastroenteritis (and certain hypovolemia) were excluded from the population studied. That is likely why so few children with hypotension were included.

Leave a Reply