Scene 1

Dr Mayaka (email: Rachel.Mayaka@howard.edu)

Scene 2

OUTCOME + NO  OUTCOME - EXPOSURE + a b a+b NO EXPOSURE - c d c+d a+c b+d a+b+c+d

Scene 3

Absolute risk Absolute risk reduction Absolute risk increase Number needed to Treat (NNT) Number needed to Harm (NNH) Relative risk Odds ratio

Scene 4

Comparative measures: measure the association between the exposure and the outcome. measure the  strength  of the association between the exposure and the outcome. Absolute measures: Measure the effects of interventions in increasing the likelihood of positive health outcomes or reducing the risk of developing negative outcomes, such as disease or death.

Scene 5

Determine the absolute risk in the control group Determine the number of times the event (outcome) occurs in the control group  Determine the total population in the control group  To determine the risk, divide the number of times the outcome occurs by the number of individuals in the group Determine the absolute risk in the experimental group Determine the number of times the event (outcome) occurs in the experimental group  Determine the total population in the experimental group  To determine the risk, divide the number of times the outcome occurs by the number of individuals in the experimental group

Scene 6

Absolute Risk (Incidence):  This is the absolute risk of contracting the disease  Can be determined for exposed or unexposed group

number of  new cases in exposed or unexposed group      number of people in that group

Incidence in exposed     = 6/120                                     =0.05 or 5% Incidence in Unexposed = 20/125                                     = 0.16 0r 16%

Drug A was studied for the prevention of CINV. The trial included 245 patients.120 pts were randomized to Drug A and 125 to placebo. 20 pts in the placebo group vs 6 in the treatment group developed CINV. Determine the absolute risk in both groups

CINV NO  CINV Drug A 6 114 120 Placebo 20 105 125 Total 26 219 245

Scene 7

Absolute risk reduction (ARR) Also known as “risk difference”  Is the measure used when the rate of the event in the exposed group  is less  than the rate  in the unexposed group It is an absolute value                              0.16        -         0.05         = 0.11 or 11%    i.e  for every 100 patients treated with drug A , 11 fewer patients experience CINV

Scene 8

The number needed to treat is the reciprocal of ARR (1/ARR) It tells us how many patients we need to treat before we see one fewer patient have the outcome If benefit of clinical trial is shown by fewer events in the treatment exposure of interest, then NNT is reported. If 5% of the  treatment  group and 16% of the  control  group had CINV,  the absolute difference is 11% so NNT is   (NNT=1/0.11 = 9)   This means that 9 people need to be treated with drug A to prevent one patient from experiencing CINV

Scene 9

Measures to use when we are exploring the increase in side effects of drug therapy

With a medication we expect fewer cases of disease to occur in relation to the condition the drug is treating Absolute risk increase Number needed to harm (NNH)

Scene 10

Absolute risk increase Also known as “attributable risk” Is the measure used when the rate of the event in the exposed group  is greater  than the rate  in the unexposed group It is an absolute value

Scene 11

Incidence among exposed:  (300/1000)*100=30% Incidence among non-exposed: (50/1000)*100=5% Attributable risk: 30%-5%=25%

What is the absolute risk increase due to HRT ?

Breast cancer (Yes) Breast cancer (No) Total  HRT therapy 300 700 1,000 No HRT therapy 50 950 1,000 Total  350 1,650 2,000

Scene 12

The number needed to harm is the reciprocal of ARI (1/ARI) It tells us how many patients we need to treat before we see one  additional  patient have the outcome 1/0.25= 4 For every 4 patients we treat with HRT, we expect one additional patient to develop breast cancer

Scene 13

Relative Risk (RR)   used   only   for   cohort,   RCT    studies Odds Ratio (OR)   used   for   cross-sectional    and   case  control   studies Cohort   studies   sometimes   use    OR   but   rarely!

Scene 14

Also  called   risk   ratio   or   rate   ratio A  relative   risk   ranges   from   0   to  +   infinity Stronger  measure  of association  compared    to  OR

Yes  No  Yes  a b a+b No  c d c+d a+c b+d

Scene 15

When   a RR>1 is  obtained   the  exposure  in    question   is   considered   to   have   an   adverse/risk    effect because   individuals   exposed   to   it   have   increased    risk  of  experiencing   the   outcome   compared   to    those  not   exposed A   calculated  RR is  considered  a  point  estimate    and  just  like  other  statistical  measures  it  has  a    95%CI  interval 95% CI of  RR  that  include 1  are  not  statistically    significant

RR=1  means   no   association   between   the  two    variables   of   interest When a RR<1 is  obtained  the  exposure  in    question   is   considered   to   have   a   protective    effect –   because  individuals   exposed   to  it   have   decreased    risk  of  experiencing   the   outcome   compared   to    those  not   exposed

Scene 16

RR   =   a   /   (a+b)   =   25/35   =   3.3

HD+ HD- Total Dg   A+ 25 10 35 Dg   A- 14 51 65 Total 39 61 100

Those   taking   drug   A   have    3.3  times the risk   for    developing heart disease    compared  to  those  not    taking   drug   A

OR Those  taking drug  A  have    greater  risk   for  developing    heart   disease   compared   to    those   not  taking   drug   A

Scene 17

Ratio of the odds of exposure among diseased to the odds of exposure among non-diseased

Scene 18

OR=1   means   equal   odds   of  outcome   for    those   exposed   compared   to   unexposed   .   No    association! When   a OR<1 is  obtained   the   exposure  in    question   is   considered   to   have   a   protective    effect –   because  individuals   exposed   to  it   have   decreased    odds   of  experiencing   the  outcome   compared   to    those  not   exposed

When a OR>1 is  obtained  the  exposure  in    question  is  considered  to  have  an  adverse    effect because   individuals   exposed   to   it   have   increased    odds   of  experiencing   the  outcome   compared   to    those  not   exposed A   calculated  OR is  considered  a  point    estimate  and  just  like  other  statistical    measures   it   has   a  95%CI   interval 95% CI of OR  that  include 1  are  not  statistically    significant

Scene 19

Pa t ients	 t a k ing   drug   A   ha v e   9.1  times  the odds of    developing hip Fracture    compared   to   those   not   taking    drug   A OR P at i ents   taking   drug   A   ha v e   a   greater	odds  for developing    hip   fracture   compared   to   those    not   taking   drug   A

Hip   Fracture + Hip   Fracture - Total Dg   A+ 25 10 35 Dg   A- 14 51 65 Total 39 61 100

Scene 20

Gives the percentage reduction that a specific drug or treatment will produce in the probability of an event Formulas: 1- (RR) (% risk in control - % risk in experimental)                  % risk in control group ARR/% risk in control group

With the control drug (drug A) there is a 20% risk of death in sepsis patients With the experimental drug (drug B) there is a 8% risk of death in sepsis patients What is the relative risk reduction?                                           OR

0.08 ÷ 0.20 = 0.4 RRR = 1- 0.4         = 0.6

Scene 21

Screening  : “The presumptive identification of   unrecognized  disease or defect by the application of tests, examinations, or other procedures that can be applied rapidly”

Scene 22

Gold standard test :  The best single test that is considered the current preferred method of diagnosing a particular disease (X).* Breast cancer  → Biopsy Tuberculosis → Chest X-ray and Blood Tb Test

Scene 23

Screening results accuracy using 2x2 table ► A:  true positive 		► B:  false positive    ► C:  false negative 		► D:  true negative

Gold standard test results   (Disease +) Positive     Negative  Screening test results Positive  A B Negative   C D

Scene 24

True positive (cell a):  The current screening test indicates they  have the disease and the gold standard test confirms this   positive with  tst  and sputum smears and cultures) False positive (cell b):  The current screening test indicates they have the disease however, according to the gold standard test, they do  not  have the disease positive with  tst  but no positive sputum smears and cultures and no clinical symptoms

False negative (cell c):  The current screening test indicates they do not have the disease, however according to the gold standard test, they do have the disease negative  with  tst  and positive with sputum smears and cultures True negative (cell d):  The current screening test indicates they do not have the disease, and this result is validated with the gold standard test negative  with  tst  and negative with sputum smears and cultures

Scene 25

Screening Test Gold Standard Test (or actual presence of  disesase ) Cell A Positive Positive True Positive Cell B Positive Negative False Positive Cell C Negative Positive False  Negative Cell D Negative Negative True Negative

Adding cells A and C gives us the number of people who truly have disease Adding cells A and B gives us the number who test positive with the screening test Adding cells C and D gives us the number who test negative with the screening test

Scene 26

Assessment tools overview Validity: Sensitivity  the ability of a test to correctly classify an individual as ′ diseased′ Proportion of persons with disease who have a positive test Specificity  The ability of a test to correctly classify an individual as  disease- free. Proportion of persons without disease who have a negative test Predictive value: Positive predictive value (PPV) -  Proportion of persons with a  positive  test who truly  have  the disease Negative predictive value (NPV) -  Proportion of persons with a  negative  test who truly  do not have  the disease

Scene 27

Gold standard test results   Positive     Negative  Screening test results Positive  A B Negative   C D

► Positive P.V:  a / ( a+b )   Proportion of persons with a  positive  test who truly have the disease ► Negative P.V:  d / (c+d) Proportion of persons with a  negative  test who truly do not have the disease

► sensitivity:  a / ( a+c )  or TP/(TP + FN) Proportion of persons  with disease  who have a  positive  test ► specificity:  d / ( b+d )  or TN/(TN + FP) Proportion of persons  without disease  who have a  negative  test

Scene 28

Case #1  : A mammogram test was used to screen for breast cancer in 1,000 women. Of those, 250 women have biopsy-proven adenocarcinoma of the breast.  400 women with the mammogram tested positive,  but only 100 of these patients had biopsy-proven cancer. Sensitivity ? Specificity ? Positive P.V ? Negative P.V ?

A picture containing text

Description automatically generated

Scene 29

Yes  No  Yes  A 100 B 300 a +b 400 No  C 150 D 450 c+d 600 a +c 250 b+d 750 1,000

Sensitivity = 100/250  = 0.4   or 40% Specificity = 450/750  = 0.6   or 60% PPV          = 100/400  = 0.25 or 25% NPV         = 450/600  = 0.75 or 75%

Scene 30

Case #1 Results Interpretations  : Sensitivity =    40%  →    40% of people who actually have breast cancer were screened positive in mammogram.   Specificity =    60%  →    60% of people who don’t have breast cancer were screened negative in mammogram. Positive P.V = 25%  →   out of 100 people screened positive, only 25 will truly have breast cancer.  Negative P.V = 75%  →  out of 100 people screened negative, about 75 will not be diagnosed with breast cancer.

Measures of disease occurrence

Page 1 (0s)

Page 2 (38s)

Page 3 (1m 23s)

Page 4 (1m 29s)

Page 5 (1m 53s)

Page 6 (2m 37s)

Page 7 (4m 18s)

Page 8 (5m 37s)

Page 9 (6m 37s)

Page 10 (7m 4s)

Page 11 (7m 41s)

Page 12 (9m 5s)

Page 13 (9m 53s)

Page 14 (10m 16s)

Page 15 (10m 59s)

Page 16 (12m 8s)

Page 17 (13m 57s)

Page 18 (14m 33s)

Page 19 (15m 24s)

Page 20 (16m 54s)

Page 21 (18m 35s)

Page 22 (19m 7s)

Page 23 (19m 56s)

Page 24 (20m 29s)

Page 25 (22m 13s)

Page 26 (22m 53s)

Page 27 (23m 47s)

Page 28 (24m 57s)

Page 29 (26m 28s)

Page 30 (27m 14s)