The PROWESS trial and subsequent registries in different nations have documented a consistent reduction in mortality in patients with severe sepsis when treated with DrotAA (Table 1). Following publication of the PROWESS trial findings 9, the open-label Extended Evaluation of Recombinant Human Activated Protein C (ENHANCE) trial 19 showed a mortality rate similar to that in PROWESS, adding further support to the findings of the original controlled trial. Descriptive analyses of national registries have also confirmed that the results of the PROWESS trial are robust.

According to the findings of the global Promoting Global Research Excellence in Severe Sepsis database of 12,492 patients with severe sepsis from 37 countries 20, the adjusted odds ratio (OR) for hospital mortality associated with DrotAA was 0.75 (95% CI 0.63 to 0.90; P = 0.002), which is similar to that observed in the PROWESS trial (OR 0.8, 95% CI 0.69 to 0.94) 9. Similarly, in the UK, Intensive Care National Audit and Research Centre data indicate that the relative risk for death associated with DrotAA was between 0.75 (95% CI 0.68 to 0.83) and 0.85 (95% CI 0.78 to 0.93) 21. These studies also demonstrate a greater treatment effect for patients with three or more organs in failure, with an ARR for dying of up to 17% 2021.

These data are consistent with the clinical experiences of other national registries. Data from a Polish registry showed that patients treated with DrotAA with a mean of 3.7 organs in failure had an ARR for death of 17.3% (38.9% versus 56.2%), giving an NNT of 6 22. In the group of patients with APACHE II scores of 25 or more, the absolute mortality risk was reduced by 13% 22. Likewise, a Belgian registry suggested that DrotAA reduced the odds of death by 39% (OR 0.61, 95% CI 0.40 to 0.92), with an adjusted ARR of 12.8% (expected versus observed mortality of 63.5% versus 50.7%) 2324. Data from an Italian pharmaco-surveillance registry showed an 8.4% ARR in patients treated with DrotAA (46.4% versus 54.9%; P = 0.0004), but these data must be interpreted with caution because the untreated group included older patients and a greater percentage of patients in septic shock (66.8% versus 77.1%; P < 0.0001) who were therefore at greater risk for death 25.

Also favourable are the long-term outcome data. Follow-up data from the PROWESS study showed that hospital survival is greater with DrotAA (70.3% versus 65.1% for placebo) 15, with no extra ICU resource use apart from the DrotAA acquisition cost 26. Patients with more severe disease and an APACHE II score of 25 or more had a longer median survival time (379 days), and an additional 11% of patients in the DrotAA group were alive at 1 year 15.

Faced with the decision of whom to treat with DrotAA, intensivists should consider in which patients is treatment with DrotAA indicated, and in which patients is the greatest benefit in terms of both mortality and cost-effectiveness likely to be realized.

Most of the trials have used inclusion and exclusion criteria similar to those in the PROWESS trial 9. In our experience, in which we have treated more than 300 patients with DrotAA, the exclusion criteria most frequently encountered are severe chronic illness, high bleeding risk, advanced cancer, excluded concomitant medication, expected survival under 24 hours, and severe thrombocytopenia (platelet count <30,000/mm³) 27.

It is clear that the key to appropriate use of DrotAA lies in the assessment of risk for death. However, there is still no consensus on a good operational definition of 'high risk' in patients with severe sepsis. In the USA DrotAA is recommended in patients 'at high risk of death, for example, those with an APACHE II score of =25', whereas in the European Union DrotAA is indicated in the presence of multiple organ dysfunction, irrespective of pretreatment APACHE II score. The recommendations of the SSC guidelines are based on APACHE II score, organ dysfunction, the presence of septic shock and sepsis-induced acute respiratory distress syndrome, but the SSC also acknowledges the lack of a clear definition of high risk and emphasizes the importance of clinical judgement in guiding timely treatment with DrotAA 4.

Using the APACHE II score to select patients for treatment with DrotAA has several limitations. Initially, this score was developed for use in populations, not to inform decisions in individual patients. Moreover, the APACHE II score in the PROWESS trial was calculated based on data obtained within the 24 hours before study randomization, not during the first 24 hours in the ICU. This renders the APACHE score problematic in assisting patient selection because physiological parameters change continuously and are modified by treatment. The APACHE II score may change but not necessarily the severity of the underlying process 7. In spite of this, an APACHE score above 25 did appear to identify a group of patients who were highly likely to experience a long-term survival benefit 15.

Results from the ADDRESS trial 16, which was conducted in patients at 'lower risk of death', confirmed that the original decisions of the FDA and EMEA may have been correct in terms of the wording of their licences. In this trial, enrolment was terminated early because interim analysis revealed a low likelihood that DrotAA would significantly reduce the 28-day mortality rate. Of the 2,613 patients, including 1,297 in the placebo group and 1,316 in the DrotAA group, both groups had similar 28-day mortality (17.0% with placebo versus 18.5% with DrotAA; P = 0.34) and in-hospital mortality (20.5% versus 20.6%; P = 0.98).

In patients with severe sepsis, the number of organs in failure is frequently used for prognostication at the bedside. In addition, multiple scoring systems (for instance, the Sequential Organ Failure Assessment [SOFA] score and the Multiple Organ Dysfunction Score) were developed to help clinicians to summarize organ dysfunction and to predict risk for death 2829. Although the degree of organ dysfunction at baseline, calculated for example using a SOFA score, is highly predictive of 28-day mortality, it appears probable that a single assessment in time is inferior to assessment of dynamic changes (for instance, delta SOFA) 3031. Using such a dynamic measurement over time more closely reflects the patient's response to therapeutic interventions 32, in that improvement in cardiovascular, renal, or respiratory function from baseline to day 1 is significantly related to improved survival. For example, based on static baseline measurements, the mortality rate in patients not receiving vasopressors decreased by 6% if they remained on no vasopressors; it increased by 11% and 32% if patients required a low or high dose of vasopressors, respectively, at day 1 32. The effects of DrotAA on dynamic changes in vasopressor use and SOFA score during the first 24 hours of treatment with DrotAA are currently being tested in a phase IV open label trial, which is expected to complete recruitment in December 2007 http://clinicaltrials.gov/ct2/show/NCT00279214?term=drotrecogin&rank=4.

As well as using baseline organ dysfunction to predict mortality, dynamic assessment of the plasma level of PC concentration has been shown to predict mortality in severe sepsis 33. In particular, having low PC levels (<40% of normal) carries an OR for death of 2.75. The OR ratio improved to 0.43 if levels rose above 40% of normal on day 1. Conversely, for patients with PC above 40%, a decrease by more than 10% increased the risk for death by 1.78 33. Perhaps these changes could be used not only prognostically but also to guide treatment in terms of the dose and duration of DrotAA therapy. A phase II clinical trial (RESPOND trial [Research Evaluating Serial Protein C levels in severe sepsis patients ON Drotrecogin alfa (activated)]; NCT00386425; http://clinicaltrials.gov/ct2/show/NCT00386425?term=drotrecogin&rank=6 is evaluating the impact of different doses and duration of treatment with DrotAA on the change of plasma levels of PC using serial measurements of PC in patients with severe sepsis.

Identifying the period of time in which to treat patients with severe sepsis is important in terms of maximizing cost-effectiveness and decreasing morbidity and mortality. Early identification and treatment of patients with severe sepsis using standard supportive care significantly improves outcomes 44. The ENHANCE trial suggested greater benefit in patients treated earlier (≤24 hours) rather than later (>24 hours from first documented sepsis-induced organ dysfunction to treatment) with DrotAA 19.

The effect of timing of DrotAA treatment on the outcome of severe sepsis was also recently evaluated in patients receiving either DrotAA or placebo who were enrolled in five severe sepsis trials with similar entry criteria, using the INDEPTH database 45. The study demonstrated that, compared with placebo, DrotAA treatment conferred a potential survival benefit, regardless of time to treatment (for instance, even when given late). However, the greatest reduction in mortality was observed in patients who were treated within 24 hours of developing the first organ failure 45, with no differences found in the source of infection according to timing of intervention 46.

The majority of the risk-adjusted survival benefit observed in patients receiving early DrotAA treatment was accounted for by a reduction in the number of deaths due to sepsis-induced multiple organ failure 45. These data were again confirmed by a retrospective analysis conducted among 274 patients with severe sepsis who received DrotAA in five teaching hospitals 47. Hospital survival was higher for patients with prompt initiation of DrotAA (same day 67.2%, next day 59.6%, later 48.4%), with an adjusted OR (95% CI) after controlling for age, vasopressors, mechanical ventilation, and other organ dysfunctions of 0.52 (0.45 to 0.60). Among those patients who received DrotAA within 24 hours, mortality rates were similar to those in DrotAA patients included in PROWESS. More recent data from a Canadian multicentre observational study 48 indicate a substantial mortality reduction (OR 0.51, 95% CI 0.28 to 0.92; P = 0.024) in those patients who received DrotAA within 12 hours of diagnosis of severe sepsis. Multivariate analysis in this study identified time to treatment as an important independent but modifiable risk for death in this population.

It is acceptable, however, to make a distinction between conditions that require early treatment with DrotAA and others in which it is reasonable to allow some time to assess response to standard treatments. A group of diseases that have been shown to benefit from early treatment (within 3 to 6 hours) with DrotAA include purpura fulminans, toxic shock syndrome, meningitis with multiple organ failure, and severe CAP (for instance, Streptococcus pneumoniae infection). In various other conditions, infection source control and response to organ support may lead to clinical improvement within 6 to 12 hours. These situations include ascending cholangitis or pyelonephritis, secondary to obstruction, catheter-related sepsis, and intra-abdominal collections, or abscesses drained surgically or percutaneously 13. If there is deterioration or lack of response to source control, DrotAA should be initiated in the absence of contraindications.

An important issue that deserves careful attention when considering the efficacy of DrotAA in clinical trials is the influence of the learning curve and the level of familiarity with the treatment protocol 49. These factors strongly influence the survival benefit with DrotAA, as reflected by the observation that there was either no treatment effect or even higher mortality in patients treated with DrotAA in the presence of one or more protocol violations pertaining, for instance, to the timing of drug administration, the enrolment of patients receiving an excluded medication, and/or failure to administer study medication according to the protocol.

These violations appear to occur during the learning curve and are less common for subsequent patients, particularly in centres treating a large number of patients 49. This suggests that experience in implementing the protocol may contribute to the greater observed treatment effect at high-enrolling sites. A similar effect might occur in routine clinical practice when physicians become more familiar with the indications for and optimal timing of drug administration, which was confirmed by data from the Polish registry 22 and from a recent UK single-centre study 50. For these reasons, the SSC guidelines strongly encourage the use of a standardized policy in the ICU for the administration of DrotAA 4.

DrotAA is cost-effective in the treatment of severe sepsis with multiple organ failure when it is added to best standard care 5859. An analysis of a UK group of patients with severe sepsis and multiple organ dysfunction estimated an additional mean cost per patient treated of £6,661, with a base-case cost per quality-adjusted life year (QALY) of £8,228 in patients with severe sepsis and multiple organ failure (based on 28-day survival data) 58. Simulation results indicate DrotAA is a cost-effective use of resources in 98.7% of cases. In the UK the cost for patients with severe sepsis and multiple organ dysfunction is £6,637 pounds per QALY based on 28-day effectiveness data and £10,937 per QALY based on longer term follow-up data 5859. This cost compares favourably with the cost per QALY of other drugs (for instance infliximab [£23,936] and etanercept [£16,330]) or 6 months of ribavirin and interferon (£7,000).

Conservatively, in the UK for example, DrotAA could potentially save 500 lives annually for a cost of approximately £25 million (the UK cost for interferon-β in multiple sclerosis is approximately £50 million). Based on the NNT to save a life, DrotAA compares very favourably with other agents 7, and although the cost for the initial 4 days of treatment is high, DrotAA is less expensive if evaluated on the basis of cost per life saved.