EPIDEMIOLOGY AND BIOSTATISTICS Flashcards
INCIDENCE PREVALENCE SENSITIVITY SPECIFICITY NPV PPV
SENSITIVITY
DETECTION RATE ABILITY OF THE TESTS TO FIND THOSE WITH THE DISEASE TRUE POSITIVE IMPLIES THAT AN INDIVIDUAL WITH THE DISEASE WILL TEST POSITIVE FOR IT
DZ NO DZ
TEST POSITIVE A B
TEST NEGATIVE C D
SPECIFICITY
ABILITY OF A TEST TO IDENTIFY THOSE WITH THE DISEASE TRUE NEGATIVE IMPLIES THAT A PERSON WITHOUT THE DISEASE WILL HAVE NEGATIVE TEST
SENSITIVITY
SENSOUT=TP/TP+FN
TN = 1 - SENSITIVITY
SENSITIVITY= A/A+C
SPECIFICITY
SPECIFICITY= D/D+B
TP = 1 - SPECIFICITY
SPIN= TN/TN+FP
NPV
PPV
DZ NO DZ
TEST POSITIVE A B
TEST NEGATIVE C D
NPV= TN/TN+FN
NPV = D / D+C
DECREASES AS PREVALENCE DECREASES
PPV=TP/TP+FP
PPV= A/A+B
INCREASES AS PREVALENCE INCREASES
RELATIVE RISK
COHORT STUDIES
PROSPECTIVE FORWARD »_space;»»»»»>
DZ NO DZ
RISK FACTOR
PRESENT
EXPOSED A B
RF ABSENT
UN EXPOSED C D
RR = EXP / UNEXPOSED
RIESGO EXP A/A+B
RIESGO UNEXPOSED= C C+D
RR = A / A+B / C/ C+D
ODDS RATIO
CASE CONTROLS STUDIES
RETROSPECTIVE ««««««««<
DZ CONTROLS
CASES NO DZ
RISK FACTOR
PRESENT
EXPOSED A B
RF ABSENT
UN EXPOSED C D
OR= DZ / NO DZ = CASE/CONTROLS
ODD CASES= A/C
ODD CONTROLS= B/D
OR A/C / B/D = AD/CB
INCIDENCE
DZ NO DZ
TEST POSITIVE A B
TEST NEGATIVE C D
RESTAR DEL DENOMINADOR LOS CASOS QUE YA ESTABAN ENFERMOS RECORDAR QUE SE ANALIZA UN INTERVALO DE TIEMPO HAY QUE DVIDIR PARA EXPRESAR LA INCIDENCIA EN EL ANNO EL SEMESTRE EL MES ETC.
INCIDENCIA = CASOS NUEVOS ENFERMOS / POBLACION SUCEPTIBLES DE ENFERMAR X TIEMPO DE ANALISIS
HAY QUE DIVIDIR EL RESULATDO SI EL ANALISIS SE HACE PARA VARIOS ANNOS ENTRE EL NUMERO DE ANNOS
DEL DENOMINADOR RESTAR LOS FALLECIDOS EN EL PERIODO DE ANALISIS AQUELOS QUE YA ESTABAN ENFERMOS ANTES DE COMENZAR EL ANALISIS ESTOS NO SON POBLACION SUCEPTIBLE DE ENFERMAR Y NO SON DENOMINADOR ERAN YA ENFERMOS
EJEMPLO EN UNA POBLACION LOS SICKLEMICOS NO SON SUCEPTIBLES DE MALARIA LOS QUE PADECIERON DE HEPATITIS A NO SON SUCEPTIBLES DE ENFERMAR LOS QUE YA TIENEN HEPATITIS B NO PUEDEN SER CASOS NUEVOS Y ESTO HAY QUE RESTARLO DEL TOTAL DE POBLACION SUCEPTIBLE DENOMINADOR
PREVALENCE
DZ NO DZ
TEST POSITIVE A B
TEST NEGATIVE C D
A+C/ A+B+C+D
IN A NORMAL OR GAUSSIAN CURVE
CENTRAL TENDENCY MEASSURE ANALYSIS
IS AS FOLLOWS:
MEAN AVERAGE
MEDIA DATA POINT IN THE MIDDLE
MODE MOST FREQUENT MEASUREMENT
NEMO: SALIR CON MODALES A LA CALLE
MODE IS ALWAYS THE OUTER VALUE, MEAN THE INNER ONE
NORMAL DISTRIBUTION
SYMMETRY 50%
POSITIVE SKEWED DISTRIBUTION RIGHT SKEWED
(1, 2, 3, 4, 8, 8, 8, 20, 100)
«««««
NEGATIVE SKEWED DISTRIBUTION
LEFT SIDED SKEWED
»_space;»»»»>
(1, 2, 2, 2 , 3, 3 ,4 ,6,10)
Example
Screening Test
Diastolic Hypertension
Yes No Total
Positive 36 25 61
Negative 9 230 239
45 255 300
SENSITIVITY 36/36+9 > 36/45 = 0.8 = 80%
SPECIFICITY 230/230+25 > 230/255 = 0.9 = 90%
Example
Screening Test
Yes No Total
Positive 36 25 61
Negative 9 230 239
45 255 300
PPV= 36/36+21 = 36/61 0.59 = 59%
NPV= 230/239 = 0.96 = 96%
Example
Screening Test
Diastolic Hypertension
Yes No Total
Positive 36 25 61
Negative 9 230 239
45 255 300
PREVALENCE
45/300 = 15%
FALSE POSITIVE RATE B/B+D 25/255 = 0.098 = 9.8%
FALSE NEGATIVE RATE C/A+C 9/45 = 0.2 = 20%
IT IS ESTIMATED THAT THE PREVALENCE OF SMOKING IN A POPULATION IS 50 %. A COHORT STUDY CONDUCTED USING A RANDOM SAMPLE FROM THIS POPULATION SHOWED THAT THE FIVE YEAR RISK OF ISCHEMIC STROKE IS 1:1000 IN SMOKERS AND 0.5:1000 IN NON SMOKERS
WHAT IS THE RELATIVE RISK (RR) OF ISCHEMIC STROKES FOR SMOKERS COMPARED TO NON SMOKERS?
RELATIVE RISK
Relative risk represents a measure of outcome in follow-up studies. It is the risk ratio which compares the risk among the exposed to the risk among the unexposed. In this case, the five-year risk of the exposed (smokers) is 0.1% (1:1000), and the risk of the unexposed is 0.05 (0.5:1000); therefore, the relative risk is 0.1/0.05 = 2.0.
IT IS ESTIMATED THAT THE PREVALENCE OF SMOKING IN A POPULATION IS 50 %. A COHORT STUDY CONDUCTED USING A RANDOM SAMPLE FROM THIS POPULATION SHOWED THAT THE FIVE YEAR RISK OF ISCHEMIC STROKE IS 1:1000 IN SMOKERS AND 0.5:1000 IN NON SMOKERS
What percentage of the strokes observed in smokers is attributed to their smoking status?
ATRIBUTABLE RISK PERCENT
In this scenario, you should calculate the attributable risk percent (ARP), which is also called etiologic fraction. ARP is a measure of excess risk. It estimates the proportion of the disease in exposed subjects that is attributed to exposure status. Two approaches can be used to calculate ARP. The first approach uses the following formula: ARP = (Risk in exposed – Risk in unexposed)/Risk in exposed = (0.1 – 0.05)/0.1 = 0.5 (50%). The other approach uses relative risk (RR) to calculate ARP:
ARP = (RR – 1)/RR = (2 – 1)/2 = 0.5 (50%). The following interpretation is valid in this case: 50% of the ischemic strokes observed in smokers can be attributed to their smoking status, and could therefore have been eliminated if they had not smoked. Educational Objective: ARP is a measure of excess risk. It estimates the proportion of the disease in exposed subjects that is attributed to exposure status.
IT IS ESTIMATED THAT THE PREVALENCE OF SMOKING IN A POPULATION IS 50 %. A COHORT STUDY CONDUCTED USING A RANDOM SAMPLE FROM THIS POPULATION SHOWED THAT THE FIVE YEAR RISK OF ISCHEMIC STROKE IS 1:1000 IN SMOKERS AND 0.5 IN NON SMOKERS
What percentage of the strokes observed in the population is attributed to smoking?
POPULATION ATRIBUTABLE RISK PERCENT
PARP estimates the proportion of the disease in the population that is attributed to the exposure. Unlike attributable risk percent,
PARP is the measure of excess risk in the total population, not only in exposed subjects.
PARP can be calculated using the following approach: PARP = (Risk in the total population – Risk in unexposed)/Risk in the total population
Knowing the risk of stroke in exposed (0.1%), the risk in unexposed (0.05%), and the prevalence of exposure in the population (0.5, or 50%), it is possible to calculate the risk in the total population: Risk in the total population = 0.1 x 0.5 + 0.05 x 0.5 = 0.075 The risk in the total population is 0.075%. Now we can calculate PARP: PARP = (0.075 – 0.05)/0.075 = 0.33 (33%) The following interpretation is valid in this case: 33% of the ischemic strokes observed in the population can be attributed to smoking. Educational Objective: PARP estimates the proportion of the disease in the population that is attributed to the exposure. Unlike attributable risk percent, PARP is the measure of excess risk in the total population, not only in the exposed subjects.
PRECISION
ACCURACY
RELIABILITY
AROUND ONE SAME POINT
SENSITIVITY+ SPECIFICITY = VALIDITY = TRUE AROUND THE CENTER POINT
CAN BE REPRODUCED
40, 000 STUDENTS TAKE USMLE STEP 3 EACH YR
X SCORE 222 SD 16 POINTS
HOW MANY STUDENTS SCORE ABOVE 254
STANDARD DEVIATION
1
2
3
68
95
99.7
2,5 13.5 34 X 34 13.5 2.5
SEM
Z SCORE
HOW FAR IS YOUR RESULT FROM THE MEAN
PRECISE SEM = SD / \/
n
1 SD Z CORE
2 SD Z SCORE
3 SD Z SCORE
EX
RR 0.7
NEW DRUG FOR A. FIB STROKE REDUCTION CI 0.5 - 1.5
THE ONE WITH THE NARROWER CI THE MORE PRECISE SEM = SD / \/```n``` THE MORE (SQUARE) SAMPLE SIZE EX: CI FROM 4-6 X=6 TO 5-7 X=6 YOU HAVE TO INCREASE 4 TIMES n
WHEN AN OUTCOME HAS A CI THAT CROSSES 1 THE RESULT IS NOT PRECISE ENOUGH TO BE USEFUL HAS NO VALIDITY
CORRELATION COEFICIENT r
/ +1
\ -1
_ 0
ARR AND NTT
DECREASE IN PERCENT OF DEATH OR DISEASE WITH THE TREATMENT COMPARED TO 100 % OF THE POPULATION.
THE MORTALITY OF ANTERIOR MI WITH NO TREATMENT IS 40 % WITH ANGIOPLASTY WITHIN 90 MIN OF ARRIVING TO THE UNIT ONLY 20 OF 100 WILL DIE WITHIN THE YEAR AFTER THE MI
THE NUMBER NEEDED TO TREAT
HIGH YIELD*
5 ANGIOPLASTIES TO SAVE 1 LIVE
NNT= 1 / ARR
ARR 40/100 -20/100 = 20/100 = 0.2
1/ 0.2 = 1/ 1/5 = 5
UNA Y OTRA VEZ HASTA EL CANSANCIO ULTRA HIGH YIELD*
VAMOS A JUGAR EL JUEGO DE LOS NOMBRES ES MUY DIVERTIDO PARA CHICOS PARA GRANDES
BUSCA UN NOMBRE CORTO REPITELO CON PA Y DI BANANA BANA Y LUEGO FI FA Y FO
ANA ANA BOBA LA BANANA NANA BOBA ANA FI FA Y FO ANA - ANA
AHORA CON STATINS AND LDL
IN PATIENTS WITH CAD EQUIVALENTS THE USE OF STATINS MEDICATIONS MAY REDUCE MORTALITY
IN GENERALLY HEALTHY PERSONS WHOSE ONLY ABNORMALITY IS AN ELEVATED LDL.
THERE IS ABOUT 3 PERCENT MORTALITY RATE FROM CARDIOVASCULAR DISEASE OVER FIVE YEARS WITH THE USE OF STATINS THIS REDUCES TO 2 % A DIFFERENCE OF 1 % THIS ARR 1% IN OTHER WORDS YOU MUST TREAT 100 PERSONS FOR 5 YEARS TO SAVE 1 LIFE
HOWEVER THE RRR IN THIS EXAMPLE LOOKS MUCH MORE IMPRESSIVE GOING FROM 3 % MORTALITY TO 2% MORTALITY
YES THERE IS A 33 % RRR IN MORTALITY BUT ONLY 1 % IN ARR
MEANWHILE THE RISK OF SERIOUS LIVER TOXICITY IS AT LEAST 3% SO THE NUMBER NEEDED TO HARM SOMEONE FROM STATINS IS 33 THE NUMBER YOU NEED TO TREAT TO HELP 1 PERSON IS 100
IN OTHER WORDS FOR EVERY PERSON YOU HELP WITH STATINS IN A GENERALLY HYPERLIPIDEMIC POPULATION YOU HARM 3 PEOPLE
FOR EVERY 100 ANGIOPLASTIES YOU DO 1 PATIENT HAS MAJOR BLEEDING LEADING TO DEATH THE RATE OR ATTRIBUTABLE RISK OF FATAL COMPLICATIONS OF ANGIOPLASTY IS 1 PERCENT OR 0.01 THE NUMBER NEEDED TO HARM IS OR .
NNH = 1 /AR
BIAS
SELECTION
BERSON
HAWTHORNE PLACEBO AND DOUBLE BLINDING
LEAD TIME EARLY DETECTION WRONGFUL INCREASED SURVIVAL
LENGHT TIME
CONFOUNDING AND EFFECT MODIFICATION
