In the October 2017 issue of the Townsend Letter I reviewed evidence that the combination of intravenous vitamin C, thiamine, and hydrocortisone can decrease mortality by nearly 80% in patients with severe sepsis and septic shock. These findings were reported in a retrospective study by Dr. Paul Marik and coworkers at Sentara Norfolk General Hospital (which is affiliated with Eastern Virginia Medical School). Beginning in January 2016, these doctors began administering this combination treatment to all patients with severe sepsis or septic shock.¹ Their protocol was instituted after the doctors observed a dramatic recovery in three patients with fulminant sepsis who were almost certainly going to die from septic shock. After treatment with intravenous vitamin C and hydrocortisone, all three patients recovered rapidly with no residual organ dysfunction. Hydrocortisone (without vitamin C) is frequently given to patients with severe sepsis, although a randomized controlled trial found that it does not improve outcomes.² Marik and associates included hydrocortisone in their protocol because of evidence that it may act synergistically with vitamin C in patients with severe sepsis. Thiamine was also given, because thiamine deficiency is common in septic patients and is associated with an increased risk of death.

The treatment protocol was as follows (all treatments were given intravenously): Vitamin C: 1.5 g every six hours for four days or until discharge from the intensive care unit (ICU). Hydrocortisone: 50 mg every six hours for seven days or until ICU discharge, followed by a tapering dose over three days. Thiamine: 200 mg every 12 hours for four days or until ICU discharge. The 47 patients who received this treatment (intervention group) were compared with a control group of 47 patients with severe sepsis or septic shock treated at the same hospital between June 2015 and December 2015. None of the patients in the control group received intravenous vitamin C or thiamine, but 60% received intravenous hydrocortisone at the discretion of the treating physician.

The hospital mortality rate was 8.5% (4 of 47) in the intervention group and 40.4% (19 of 47) in the control group (78.9% reduction; p < 0.01). None of the patients in the intervention group died of complications related to sepsis; rather, the four patients who died succumbed to complications of their underlying disease. The Sequential Organ Failure Assessment (SOFA) score decreased (improved) in all patients in the intervention group, and none developed progressive organ failure. Approximately two-thirds of the patients in each group had acute kidney injury at presentation. However, significantly fewer patients in the intervention group than in the control group ended up requiring dialysis or other renal replacement therapy (10% vs. 37%; p = 0.02). All patients in the intervention group were weaned off vasopressors at a mean of 18.3 hours after starting vitamin C, and the dose of vasopressors was typically reduced at two-to-four hours after the first vitamin C infusion. In contrast, nine patients in the control group required increasing doses of vasopressors and died of refractory septic shock. In the intervention group, renal function improved in all patients with acute kidney injury.

This report created great excitement, as it potentially represented a major advance in the treatment of critically ill patients. More than 50 medical centers around the world began using Marik’s protocol, and many reportedly observed the same dramatic benefits.³ Marik’s report also led numerous research groups around the world to conduct randomized controlled trials to determine whether the benefits of vitamin C, thiamine, and hydrocortisone could be confirmed. The results of some trials were positive, while others found no benefit. In none of the randomized trials were the results even nearly as impressive as those reported by Marik and associates. Some of these studies are reviewed below.

Randomized Controlled Trials

One hundred thirty-nine patients in New Jersey with sepsis or septic shock were randomly assigned to receive, in double-blind fashion, the Marik protocol or placebo for a maximum of four days. The mean time that patients required vasopressors was significantly less in the intervention group than in the placebo group (27 vs. 53 hours; p<0.001). Hospital mortality was 16.4% in the intervention group and 19% in the placebo group (p = 0.65).⁴

Eighty patients in China with sepsis or septic shock were randomly assigned to receive, in single-blind fashion, the Marik protocol or placebo for four days (with hydrocortisone continued for 7 days). Twenty-eight day mortality was nonsignificantly lower in the intervention group than in the placebo group (27.5% vs. 35%; 21.4% reduction p = 0.47). The mean improvement in the SOFA score at 72 hours was significantly greater in the intervention group than in the placebo group (3.5 vs. 1.8; p = 0.02).⁵

Two hundred five patients at 14 centers in the United States who had septic shock were randomly assigned to receive, in double-blind fashion, the Marik protocol or placebo for four days. Patients were enrolled within 24 hours of diagnosis. Median time from initiation of vasopressor therapy to initiation of study medication was 14.5 hours in the intervention group and 13.0 hours in the placebo group. The proportion of patients who died within 30 days was nonsignificantly higher in the intervention group than in the placebo group (34.7% vs. 29.3%; p = 0.26). There were no significant differences between groups with respect to mean change in SOFA score during the first 72 hours or incidence of kidney failure.⁶

One hundred patients in northern India diagnosed with sepsis (85% of whom had septic shock) received standard treatment and were randomly assigned to receive or not to receive (control group) the Marik protocol for four days or until discharge from the ICU (hydrocortisone was given for seven days or until ICU discharge, followed by a taper over 3 days). In-hospital mortality was nonsignificantly lower in the intervention group than in the control group (24% vs. 28%; 14.3% reduction; p = 0.82). Thirty-day mortality was 40% in the intervention group and 42% in the control group. Mean duration of vasopressor use was significantly lower in the intervention group than in the control group (75.7 vs. 96.1 hours; p = 0.01). The intervention group may have been sicker than the control group at baseline, since the admission diagnosis was septic shock in 92% of patients in the intervention group and 76% of those in the control group.⁷

Two hundred eleven patients at 10 intensive care units in Australia, New Zealand, and Brazil with septic shock were randomly assigned to receive the Marik protocol or hydrocortisone alone (control group) until cessation of vasopressor administration or until any of the other criteria for stopping the intervention were met, or for a maximum of 10 days. Patients were enrolled a maximum of 24 hours after they fulfilled the diagnostic criteria for septic shock. The median time from meeting eligibility criteria until the first dose of vitamin C was 12.1 hours. Patients in the control group were allowed to receive thiamine at the discretion of the treating physician. There was no significant difference between groups in the primary outcome measure, which was the time alive and free of vasopressor administration up to day 7. Ninety-day mortality was 28.6% in the intervention group and 24.5% in the control group.⁸

Possible Explanations for the Less-Than-Impressive Results

Although there was some evidence of a modest benefit from the randomized trials reviewed above, the results were far less impressive than those reported by Marik and by some other ICUs units around the world that adopted his protocol. Marik and associates have argued that the failure of the randomized trials to confirm their results may have been due to two factors.^9,10 First, the researchers may have waited too long to begin the treatment. In Marik’s series of patients, almost all were given vitamin C, hydrocortisone, and thiamine within six hours of diagnosis. In contrast, among the studies that provided such data, many or most patients did not receive the Marik protocol until 12 or more hours after severe sepsis or septic shock was diagnosed. A retrospective cohort study of 206 patients treated for septic shock at a Wisconsin tertiary academic center found that ICU mortality was markedly lower among patients who received the Marik protocol within six hours of diagnosis than among those whose treatment was started six to 24 hours after diagnosis. Second, Marik’s group argued that excessive amounts of fluid may have been administered in the randomized trials, and that giving too much fluid can greatly interfere with the beneficial effects of vitamin C, hydrocortisone, and thiamine.

I am not an expert in the treatment of septic shock, so I am not in a position to evaluate the arguments made by Marik and colleagues. However, I am aware of many apparently effective treatments that yielded negative results in randomized controlled trials because the investigators deviated from the protocol used in the successful trials. Dr. Marik is a renowned authority on critical care, having written over 400 peer-reviewed journal articles, 50 book chapters, and four books on critical care. In recognition of his expertise and considering that countless lives could be saved if his observations turn out to be correct, researchers should conduct a large, randomized trial that adheres exactly to the methods Marik and associates found to be so successful.