Aim

To determine if a protocolised management of hypotensive sepsis, including fluids, blood transfusion and vasopressors will improve survival. 

Rationale

There is ongoing controversy regarding the optimal management of sepsis. Protocols have been shown to be minimally effective in high resource settings but there have been limited studies of this approach in lower resource settings.    

Design

Randomised controlled trial 

Setting

Single centre- Zambian hospital (1500 bed referral hospital) 

Population

212 patients

Adults with >2 SIRS criteria and systolic bp <90 or MAP <65

Exclusion; severe hypoxia or tachypnoea or obvious alternative pathology (e.g bleeding or heart failure).

Mean age 36 years old, the vast majority testing HIV positive, ⅔ having TB and many malnourished. 

Intervention

2 litres of isotonic fluid over first 2 hours, then 2 litres of next 4 hours

Dopamine infusion if MAP <65 after first 2 litres of fluid.

Blood transfusion if Hb less than 7

Control

Usual clinican care; these patients received less than 2 litres of fluid in the first 6 hours. Only 2 percent received vasopressors. 

Primary outcome

In hospital mortality 

Results

51of 106 patients (48.1%) in the sepsis protocol group vs 34 of 103 patients (33.0%) in the usual care group (between-group difference, 15.1% [95% CI, 2.0%-28.3%]; RR, 1.46 [95% CI, 1.04-2.05]; P = .03)

Author's Discussion

In this patient population (high levels of TB, HIV and malnutrition) a protocol-based approach increased mortality. Despite improvements in fluid administration, vasopressor use and lactate clearance there were higher rates of respiratory distress and mortality. They speculate that these patients' comorbidities, as well as lack of access to ventilatory support made these patients more vulnerable to fluid overload and cautiously suggest that risks of intravenous fluids outweigh the benefits.

Strengths

Randomised 

Power calculation based on previous trials 

Intention to treat analysis 

Similar baseline characteristics in both groups

Weakness

Weekday recruitment only 

Only 15% received vasopressors so I am not sure sufficient numbers to make definitive statements regarding this intervention from this trial.

Very malnourished and immunocompromised cohort of patients, this limits generalisability to low-resourced environments with lower burdeons on these conditions.

 My humble opinion

The marked increase in mortality with the increased use of fluids is remarkable and concerning. It deepens my uncertainty of how much fluid to prescribe for septic patients. I will be significantly more cautious about my approach to fluids in the malnourished, immunocompromised patients.