Incidence= [Number of new cases in a time period/Population] x 100,000
Prevalence= [Number of existing cases in a time period/population] x 100,000 Bradford Hill Criteria of causation Strengt of !ssociation "onsistenc# Specificit# in te cause $emporalit# %ose&response relationsip Plausibilit# "oerence 'xperimental e(idence !nalog# Cross-sectional study) assesses bot te ealt status and te exposure le(els of indi(iduals witin a population at one point in time* Case-control study) a retrospecti(e stud# tat initiall# identifies two groups of sub+ects* !ll indi(iduals in one group a(e te particular disease or condition under in(estigation ,te cases-, wereas e(er#bod# in te oter group is free from te disease ,te controls-* Odds ratio) te ris. of te odds of de(eloping te disease in te exposed group di(ided b# te odds of de(eloping te disease in te unexposed group* "ases "ontrols 'xposed a b Not exposed c d /0= ,a/c-/,b/d-= ad/bc 1f exposure if armful, /021 1f exposure is protecti(e, /031 1f /0=1, no ris. can be attributed to te disease Cohort study) 4ollows a group of sub+ects forward in time and compares teir outcomes after one group is exposed to some .nown of suspected cause of disease wile te oter group is not exposed* Relative risk) compares te ris. of some ealt&related e(ent occurring in two groups tat are included in a prospecti(e stud#* 1t is te probabilit# of disease occurring in te exposed group di(ided b# te probabilit# of disease in te unexposed group* %isease present %isease absent 'xposed a b Not exposed c d 00= [a/,a5b-]/[c/,c5d-] 1f 0021, association is positi(e 1f 0031, association is negati(e ,or protecti(e- 1f 00=1, not associated wit ris. of disease Attributable risk) te probabilit# of disease in te exposed group minus te probabilit# of disease in te unexposed group* !0= [a/,a5b-] 6 [c/,c5d-] Absolute risk reduction) te difference in te probabilit# of disease between te treatment and control groups* 1t is calculated te same as !0, but it tells #ou te ow muc of te difference in reduction of disease incidence between te groups is due to te treatment* !00= [a/,a5b-]& [c/,c5d-] Relative risk reduction) te comparati(e reduction in rates of bad outcomes between te experimental and control groups in an 0"$ or coort stud#* 000= absolute ris. reduction/probabilit# of disease in unexposed group, tus 000= 7a/,a5b-] 6 [c/,c5d-]8/[c/,c5d-] Number needed to treat) number of patients wo would need to be treated in order to pre(ent one additional bad outcome* NN$= 1/!bsolute ris. reduction= 1/[a/,a5b-] 6 ,c/,c5d-]