remember to round it to whole numbers Number Need to Treat In therapy studies, treat ment effects from binary outcomes, such as “positive” or “negative”, can be present in various ways (e.g., relative risk reduction, absolute risk reduction, number needed to treat) [1 ]. LAUPACIS et al introduced the concept of number needed to treat as an alternative approach to summarizing the effect of treatment [2 ]. The number needed to treat is defining as “the number of patients who need to be treat ed to prevent one additional adverse even” [3 , 4 ]. When the experimental treatment increases the risk of an undesirable outcome/event the number needed to harm can be compute. The number needed to harm has the same mathematical expression as number needed to treat [1, 5 ] remember to round it to whole numbers

Clinical questions Matching the strongest design to clinical question Abnormality Is the patient sick or well? Diagnosis How accurate are tests used to diagnose disease? Frequency How often does a disease occur? Risk What factors are associated with an increased risk of disease? Prognosis What are the consequences of having a disease? Treatment How does treatment change the course of disease? Prevention Does early detection and treatment improve the course of disease? Cause What cause the disease? Cost How much will care for an illness cost? Prevalence Prevalence Cohort study, case control study Cohort study Clinical trial Clinical trial What is the difference between risk and Hazard? A hazard is a substance, agent, or physical situation with a potential for harm in terms of injury or ill health, damage to property, damage to the environment, or a combination of these. It is a noun. Risk relates to the likelihood of the harm or undesired event occuring, and the consequence of its occurrence. It is a function of both exposure to the harzard and the likelihood of harm from the hazard. It is a probability. Oxford textbook of public health p1047 Cohort study, case control study

The Six Ds Outcome of Disease Death Disease/Illness Discomfort Disability Dissatisfaction Destitution

Study & Bias

All patients with the condition of interest Internal validity Sampling Sample Sample Selection bias Measurement, confounding bias ??? External validity Conclusion

Efficacy vs. Effectiveness

The story of selling bug-killer Well, if this does not kill every of the bug, I will give you $1000000000000000 in return!

I have try it, but it does not work!

Oh! I have forgotten to told you, you have to put the drug into every bug’s mouth! If it does not kill it, don’t hesistate to come back!

Efficacious treatment One that has the desired effects among those who receive it Effective treatment One does more good than harm in those to whom it is offered

Intention to Treat analysis analyze according to treatment assigned Drop out analyze according to treatment assigned i.e. effectivness Population of patients with the condition sample Cross over Drop out Drop out analyze according to treatment recieved i.e. efficacy Population of patients with the condition Cross over sample Per treatment analysis Drop out

Efficacy Trial Internal validity Generalizability Effectiveness Trial Noncompliance Less selected patient Less selected clinician Costs Impracticality Other Internal validity Generalizability Effectiveness Trial Fletcher 3rd

Examination paper 2000/5/30 Drug A Drug B Placebo % patients with reduction in number of complications 12% 19% 5% % patients with side effects 4% 6% 2% A randomised clinical trial was carried out with sample size of 500 patients to determine whether drug A or B is better at reducing the number of complications in a particular disease. Both drugs were compared to placebo.

B1a: Calculate the NNT ARR (treatment effect) NNT ARI (side effect) NNH LHH=NNH /NNT Drug A vs. Placebo 12%- 5%=7% 14.29 2% 50 3.5 Drug B vs. Placebo 19%- 5%=14% 7.14 4% 25 Drug B vs. Drug A 19%- 12%=7% 14.28 ARR = ABI = Absolute benefit increase ARI = Absolute risk increase LHH: Likelihood to be helped versus harmed

B1a: Calculate the NNT ARR (treatment effect) NNT ARI (side effect) NNH LHH = NNH/NNT Drug A vs. Placebo 7% 14.29 2% 50 3.5 Drug B vs. Placebo 14% 7.14 4% 25 Drug B vs. Drug A 19%- 12%=7% 14.28 Both drug A and B are more effective than placebo, though drug B is more efficacious than drug A

B1a: Calculate the NNT ARR (treatment effect) NNT ARI (side effect) NNH LHH Drug A vs. Placebo 7% 14.29 2% 50 3.5 Drug B vs. Placebo 14% 7.14 4% 25 Drug B vs. Drug A 14.28 Drug B is also more effective than drug B as patient more likely to be helped than harmed when they use drug B rather than drug A We have to treat more people to show the benefit of drug B over drug A than that of placebo

B1c: Potential benefits for using NNT Impact of treatment is related to the baseline risk of the adverse outcome it is designed to prevent easier to derive risk-benefit analysis before making clinical decision. easier to interpret than relative risk, odds ration and relative risk reduction and thus easier to communicate with colleagues, nurses and general public. it is a measure of therapuetic effort required to obtain one beneficial event. Consequently, it is useful in comparing utility of treatments when we have comparable trials with comparable outcomes in comparable patients. it is a measure of therapeutic effort required to obtain one beneficial event.

Study