Statistics relative risk reduction

Recent publications on major clinical trials whose implications will involve a recommendation to change clinical practice have included summary statistics that quantify the risk of benefit or harm that may occur if the results of a given trial are strictly applied to an individual patient or to a representative cohort. Four simple calculations will enable the non-statistician to answer the simple question How much better would my chances be (in terms of a particular outcome) if I took this new medicine, than if I did not take it . These calculations are the relative risk reduction, the absolute risk reduction, the number needed to treat, and the odds ratio (see Box 6.3). 

Several direct thrombin inhibitors have been studied in NSTEMI and STEM I patients and were compared to unfractionated heparin. In the GUSTO lib- and OASIS-2 trial (42,43), hirudin was studied versus heparin in patients with ACS. Despite early benefits, no statistical significance could be demonstrated at 30 days. Together with the OASIS-1 data, a combined analysis indicated a 22% relative risk reduction in cardiovascular death or Ml at 72 hours, 17% at 7 days, and 10% at 35 days (42). 

In MADIT, the relative risk reduction in total mortality with the ICD was 54% at 27 months compared to conventional therapy, (p = 0.009, hazard ratio of 0.46, 95% confidence limits 0.26-0.82). The fact that 50% of the ICD patients received an ICD shock made investigators suspect that the device was protecting the patients from death. The statistical methods used in MADIT allowed a relatively small number of patients to show a statistically significant result however, these methods did not allow for multivariate analyses or subanalyses. Most of the benefit of the ICD occurred early after implantation. 

The two other available double-blind trials on Echinaceae purpureae herba [14] and on a combination of Echinacea angustifolia with Eupatorium perfoliatum, Bap-tisia tinctoria and Arnica montana [15] yielded very similar rates of relative risk reduction 12% (95% Cl -13 to 32%), and 14% (95% Cl -2 to 27%) respectively. In the second trial [15] which included more than 600 volunteers the difference just missed statistical significance. 

Although all available trials show at least some reduction in the risk of getting an infection, the evidence for a preventive effect of preparations containing extracts of Echinacea is insufficient to allow for reliable conclusions. If the size of the relative risk reduction is - as the newer, more reliable trials suggest - in fact in the range of 12 to 20%, the statistical power of the available trials is insufficient (the number of patients included in the trials has been too small to confirm statistically a potential effect of this size). 

While the relative risk is a standard statistic that can be used to compare treatments, it can be difficult to understand and to relate to practice. For example, although the relative risk of 3.7 that was calculated above indicates that Drug A is associated with nearly four times the risk of cure compared with Drug B, this gives no indication of the practical implications. For this reason, effects are often quoted as the Number Needed to Treat (NNT). The NNT is calculated as the reciprocal of the absolute risk reduction (ARR). In the example in Table 3, the NNT refers to the number of patients who need to receive Drug A before an additional cure is likely to occur. 

Also the secondary endpoints of microvascular comphcations were markedly altered 61% (Cl 13-83%) lower relative risk of nephropathy, 58% (Cl 14-79%) lower risk of retinopathy, and 63% (Cl 21-82%) lower risk of autonomic neuropathy. For the primary CVD endpoint risk reductions were seen for all the different components except for mortality. It may be added that this trial was not statistically powered to evaluate the interventional impact on mortality. 

The PPP (Primary Prevention Project) study [15] was a randomized, controlled (no placebo), open trial investigating low-dose aspirin (100 mg daily) in 4,495 subjects (42.5% males) with one or more cardiovascular risk factors. Follow-up was 3.6 years. Aspirin was found to lower the frequency of all the end points, being statistically significant for cardiovascular death (relative risk 0.56), and total cardiovascular events (relative risk 0.77), with non-significant reduction in myocardial infarction (relative risk 0.69) and stroke (risk reduction 0.67). Seven hundred and forty-two diabetics (17%) were included in the study but no separate information on this subgroup is provided in the original publication. 

Adjuvant therapy is usually systemic therapy that is administered to treat any existing micrometastases remaining after treatment of localized disease. Because adjuvant therapy is given to patients with no clinical evidence of cancer, the benefit of the treatment cannot be proven for an individual patient, but only for patient populations. Treatment decisions are based largely on an assessment of the presence of risk factors in an individual patient and the patient s estimated risk for cancer recurrence. The effectiveness of adjuvant therapies is measured statistically, by the relative and absolute reduction in the risk of recurrence. In con-... 

---