A drug (or any medical treatment) should of course only be given if it uses the patient. In evaluating the merits of both the property of the drug, the desired effect (effectiveness) is to achieve, as well as the type and the probability of occurrence of adverse effects (safety) a. Costs and benefits are normally weighed against each other (economic analysis as part of the clinical decision-making). Potency and efficacy efficacy (potency) is the ability to achieve a therapeutic effect (eg., Blood pressure lowering). (D. E. When the patient be selected by appropriate criteria and exactly to the dosage regimens hold) the potency can be accurately judged only under ideal conditions (effectiveness under certain controlled conditions). Thus, the potency is measured under professional supervision in a patient population that reacts most likely due to a drug such. B. in a controlled clinical trial. The effectiveness (effectiveness under everyday conditions), in contrast to the potency of how well a drug in clinical practice is; often is a drug that is effective in clinical trials, not very effective in routine use. For example, a drug may reduce blood pressure very effective, but in everyday life could be less suitable (effective) because of the numerous side effects, patients taking the medicine. Although doctors accidentally prescribe unsuitable drug, the effectiveness (daily activity) may be less than the potency of (e. B. upon administration of a fibrinolytic drug in the not recognized a hemorrhage in the CT in a patient suspected of having ischemic stroke, was). Thus, the efficiency in everyday life is most often be less than the effectiveness under controlled conditions (potency). clinical endpoints should be used to assess the potency of the effectiveness and instead of surrogate endpoints or intermediate. Clinical endpoints Clinical endpoints are those that affect the well-being of the patient. These include: extension of life improvement of the functions (eg prevention of disability.) Relieving symptoms surrogate endpoints or intermediate surrogate endpoints include properties not directly related to the condition of the patient. These are often variables such as physiological parameters (eg., Blood pressure) or measured values ??(z. B. glucose or cholesterol levels, tumor size on a CT scan), the actual clinical endpoints are to predict. For example, physicians will typically assume that lowering blood pressure prevents the clinical endpoint of uncontrolled hypertension (z. B. death as a result of myocardial infarction or stroke). It may be, however, that while a drug lowers blood pressure, but not the death rate because it has fatal side effects under certain circumstances. If the surrogate is rather a marker (eg. As HbA1c) as a cause of disease (eg. As high blood pressure), an intervention could reduce the marker while, but this has no effect on the underlying disease. Thus, surrogate endpoints are less suitable criteria for evaluating the effectiveness of clinical endpoints. In contrast, surrogate endpoints can be much more practicable, for instance, if clinical endpoints very rarely or only over a long period occur (eg. As kidney failure due to uncontrolled hypertension). In these cases, clinical trials would have very extensive and carried out over a long period, unless it is a surrogate (z. B. lowering blood pressure) is used. In addition, the most important clinical endpoints, death and disability, dichotomous (d. H. Yes / no), while surrogate endpoints mostly continuous and numerical variables (eg., Blood pressure, blood sugar) are. Unlike dichotomous events back pass numeric variables an indication of the size of an effect. Thus, surrogate end points often provide significantly more data for the evaluation of clinical endpoints, so that clinical trials can be carried out with a lower patient population. However, should ideally correlate with the clinical endpoints surrogate endpoints. There are numerous studies in which such a correlation is apparently given, but was not really there. For example, the treatment results with estrogen and progesterone in some postmenopausal women at a better lipid profile but is not accompanied by a corresponding reduction in the risk of myocardial infarction or cardiac death. Similarly, a decrease in blood glucose at about normal levels leads in diabetic patients in the ICU to a higher mortality and morbidity (possibly through occurring hypoglycemia) than in patients in whom the blood sugar was not quite as deep lowered. Some oral agents do not reduce blood sugar levels, including the level of HbA1c, but reduce the risk of cardiac events. Although some antihypertensive drugs not reduce reduce blood pressure, but the risk of stroke. Side effects Clinically relevant side effects also represent clinical endpoints; for example: Death Disability – neurological status ailments, the surrogate side effects (. eg change in the concentrations of serum markers) is often used, but as with the surrogate efficacy (potency) should ideally correlate it with the clinical side effects. Clinical studies carefully designed to demonstrate efficacy, can always recognize side effects difficult if the side effect occurs very rarely or the time to occurrence of side effects is longer than the evidence of effectiveness. For example, since cyclooxygenase-2 inhibitors (COX-2 inhibitors) reduce pain quickly, their efficacy can be detected in a relatively short study. However, the increased in some COX-2 inhibitors incidence of myocardial infarction shows over a longer period of time and is not apparent in shorter, smaller studies. For this reason, and because clinical trials can exclude certain groups of patients and high-risk patients, adverse effects are not fully known at launch, but may in the broad clinical application only after years show (drug development). Many adverse drug reactions are dose-dependent. Consideration is indicated by the benefits and side effects Whether a drug depends on the benefit-risk assessment. With these ratings in doctors often incorporated subjective factors such as personal experiences, stories, experiences of colleagues and experts opinions. The number needed to treat (number needed to treat, NNT) is a more objective factor in the likely benefits of a drug (or any other intervention). The number needed to treat the number of patients who need to be treated to a patient benefited from the treatment. For example, if a drug mortality of disease of 10% reduced to 5%, there is an absolute risk reduction of 5% (1 of 20). This means that out of 100 patients 90 survive without treatment and thus do not benefit from the drug would. And 5 out of 100 patients will die, although they take the drug. But only 5 of 100 patients (1 of 20) benefit from taking the drug; that is, it must be 20 patients treated for one patient a benefit from treatment, the NNT is 20. The number needed to treat can be simply calculated as the inverse of the absolute risk reduction; z. For example, the absolute risk reduction is 5% (0.05), NNT = 1 / 0.05 = 20. The number needed to treat can also be used for the calculation of side effects which then spoken by the number needed to harm (NNH) is, that the number of patients to be treated, which are necessary to obtain an unwanted side effect. is the benefit increase (Benefit Increase) / number of necessary BHDL of Multicalc (number needed to treat Multicalc) Importantly, the number needed to treat on the absolute risk change is based, it can not be calculated from a change in the relative risk: clinical computer.. The relative risk is the proportional difference between two levels of risk. For example, a drug that reduces mortality from 10% to 5%, but an absolute reduction in mortality by 5%, the relative mortality increases by 50% from (ie, with a mortality of 5% occur 50% fewer deaths than in a mortality of 10%). In most cases, the data refer to the literature on the relative risk reduction, as this is a drug appear more effective than in the absolute risk reduction (in the example a reduction in mortality by 50% sounds much better than a reduction of 5%). In contrast, side effects are often presented as absolute risk increase, as they have a drug appear safer. For example, if a drug increases, the incidence of bleeding from 0.1% to 1% is reported rather by an increase of 0.9% as 1000%. Tips and risks The number needed to treat (NNT) should be calculated based on the absolute and relative risk based change. Be NNT and NNH placed opposite, it is important to consider the scope of the specific benefits and potential harm. For example, it may be useful, a drug that has significantly more side effects than benefits, but to prescribe when these harmful effects are of little importance (eg. As reversible, weak), whereas the benefits but crucial (z. B. . prevention of mortality and morbidity). In any case, the clinical endpoints are best suited. Patient groups using genetic profiling are identified increasingly that respond better to certain drugs with regard to benefits and side effects. For example, breast cancer patients on the genetic marker HER2 can be tested, which allows a prediction about the effect of certain chemotherapy drugs. Patients with HIV / AIDS can 57 on the allele HLA-B *: 01 are tested, which provides information on the possibility of hypersensitivity to abacavir; This reduces the incidence of hypersensitivity reactions and increases the NNH. Genetic variants in drug metabolizing enzymes allow a prediction of how patients respond to drugs (pharmacogenetics); they often affect the likelihood of benefit, damaging effect, or both. A therapeutic index target in drug development is to have caused the greatest possible distance between the therapeutic dose and the dose side effects. A large distance is referred to as great therapeutic breadth, therapeutic ratio or therapeutic window. When the therapeutic index is low (eg. As <2) factors that are usually clinically insignificant (eg. As food-drug interactions, drug-drug interactions, minor errors in dosage) have adverse clinical outcomes are. For example, warfarin has a narrow therapeutic index and interacts with numerous drugs and foods. An inadequate anticoagulation increases the risk of complications of a disease that is treated with anticoagulants (eg. As increased risk of stroke in atrial fibrillation), while an excessive anticoagulation increases the risk of bleeding.