Bios Tat

ก  
  
 2 ( .300) !"# Quantitative Method I .   

  

  . ----------------------------------------------------------------------------------------------------------------------------------


$%& #'("#!"# ()" 1 "+,
'-   ก!
ก,ก
-'
.!#/#0#%(1

()"2 & %ก'/$# 2 3 ก& . 4ก ,/"%(1

()"3 "+,
'-   ก!
ก,5&,,ก6ก#
ก ,()"4 "+,
'-   ก!
ก,ก #..  $#" %. 1. DISTRIBUTION • Normal distribution • • •

#ก#'
.$! Histogram ก8/#-#'
.$! ) 9')#'
.$!

• Binomial distribution • Poisson distribution 2. CENTRAL TENDENCY •

Arithmetic Mean : '/:!('2

•

Geometric Mean : '/:!('2

•

Harmonic Mean : '/:!;9. '

• •

Median Mode


"/#()"", (THAMLE :Thailand Medical Licensing Examination 2007 ) กก

 Hct % %  60  ( )(* + (,ก ก ++ normal distribution 89 %: 8; ;< ก:=9=< a. median S.D. b. mean S.D. c. median range d. mean range e. mode range

3. VARIATION The variance: &B '/!,/#,"ก$6#กก ("#()".5
/()3ก)C"C/#ก'/:!C"'/ก#("#

()".5("#& ก (δ2) Standard deviation : &B '/!,/#,"ก$6#กก ("#()".5("#ก%/.
"/# (S2) The coefficient of variation Probability ; risk , rate…etc.

4. INCIDENCE CUMULAIVE INCIDENCE (CI) number of new events at specified time ×k number of risk group at specified time

K= 10n Meaning: average risk in each event Unit : rate per population Example : new cases of DHF = 100 cases persons During three months

Risk group or population at beginning of time = 10,000

100 CI=

×1000 = 10 per 1,000 population 10,000

Incidence density (ID) Number of new events at specified time ×k person – time K = 10n

Meaning : risk hazard or hazard rate Unit: cases per person per time Person time (p-t): number of persons at each point of time; 10 persons × 10 days = 100 p-d 5 persons × 5 days = 25 p-d Total of p-t = 125 p-d Or number of persons × length of observation

Example: news cases of DHF 100 cases Person time = 125 p-d

During three months

100 cases ID =

×100 125 p-d

= 80 cases per 100 person per day


"/#()"", (THAMLE :Thailand Medical Licensing Examination 2007 )( ก =BC+9%ก 60 (*  ) 1,000  + (, DM 200  ;< ก (:H(: I(BJH (* +C9 (, DM ;H%ก 50  8 incidence (, IB a. 5% b. 6.25% c. 20% d. 25% e.50%

5. PREVALENCE Number of existing cases of disease at point of time ×k Total number of population at the same point of time

K= 10n Meaning : magnitude of problems Unit : rate per population

Example : total cases of DHF 200 cases Total population = 10,000 persons

200 P

=

× 1,000 = 20 per 1,000 population 10,000

6. OTHERS 6.1 Fatality rate

Number of death from each disease ×k Total number of cases

K = 10n Meaning : severity of disease Unit : rate per population

Example : number of 10 death from DHF Total cases of DHF = 100 cases

10 Fatality rate =

× 100 = 10 % 100

10 er 100 population

6.2 SURVIVAL RATE Survival rate is the rate to determined survival time of the patient in each disease by using KaplanMeier procedure.


"/#()"", (THAMLE : Thailand Medical Licensing Examination 2007 ) กก JกLก ก;< leptospirosis = I(* 2548 %

%+ ;ก  = 0.4 : %

ก  = 0.02 %( ก  case fatality rate  IB a. 0.0002 b. 0.004 c. 0.4 d. 0.02 %( ก  e. 0.4 %( ก 

6.3 HEALTH IMPACT 6.3.1 Risk difference or attributable risk: Incidence of exposed (I e) – Incidence of unexposed (I o) Meaning : risk difference between exposed group and unexposed group Unit : the same as incidence in the formula Exposed persons who get sick R (exposed) = All exposed persons

Unexposed persons who get sick R (unexposed)= All unexposed persons

6.3.2

Risk ratio: R (exposed) RR= R (unexposed)

Meaning : relative risk between exposed group and unexposed group. 6.3.3

Odds of case exposure

Exposed cases All cases

6.3.4

Unexposed cases All cases

Odds of control exposure

Exposed controls All controls

Unexposed controls All controls

6.3.5

Odds ratio: Odds of case exposure OR= Odds of control exposure

7. EFFICACY OF DRUG THERAPY 1. Risk reduction (ARR) : EER-CER EER = experimental event rate: number of success in experimental group divided by total samples of experiment. CER = control event rate: number of success in control group divided by total samples of control. Using in phase III of drug trial 2. EFFICACY OF DRUG THERAPY Number needed to treat (NNT) is the number of patient that must be treated with the proposed therapy in order to have one additional successful result. ;H ก;H I< NNT = 1/ ARR 3. RISK OF DRUG THERAPY Number needed to harm (NNH) tell us how many people need to exposed before harm will be fall one additional person or one additional harmful outcome event will be occur. ;H ก;H < NNH = 1/ AAR AAR= absolute risk difference. Using in phase IV of drug trial.

8. ASSOCATIONS Correlation : B ;  B % ( B S%กก Covariance: ( (   กH ; (, B ( H)=B9 ก;<( ( =ก ;  B

9%C:

9. ESTIMATION 9.1 Point estimation: (,ก (  ; % % ( ก <9 : =< BJ H T<=9%9 C:

% 9.2 Interval estimation: (,ก (  ; % % ( ก %C= =< BJ H T<=9

9%C: % : % ก (  +%ก8J H H) <:HC <%  ; %  10. Level of confidence: <+ SH%HB8J T%กH ; % % ( ก %C= % H

( I<9  P(L<μ
-

10.1 Confidence intervals (CI) The limits of the 95% CI show how precise the result are. If the CI is very wide, the result are not very precise. If the CI is very narrow, the result are very precise 10.2 TYPE I ERROR This occurs when the null hypothesis is rejected event though it is really true. This is a false positive study result. Alpha (α) known as the level of significance, is defined as the maximum probability of making a Type I error that we are willing to except. 10.3 TYPE II ERROR This occurs when the null hypothesis is not rejected (Ho is accepted and no difference is found) even though it is really false (and the groups are different). Beta (β) as the maximum probability of making a Type II error or failing to reject the null hypothesis when it is actually false. Power is the ability to detect a statistically significant difference when it actually exists.(1-β) A summary of the types of conclusions and errors that may be drawn from a statistical test of a drug vs. placebo


  (THAMLE :Thailand Medical Licensing Examination 2006) )ก JกL( ;;V%  =B +ก+  กT<JกLกก: % % :50 ก:;ก:BJH +:ก <:% I + ก  % ก:<:%  :ก ;%SH(%SHX)+ ก  8)% I J :< type II error I<9 a. ;Hก:<:% b.:<ก:<:% c.:< p< 0.05 (, p< 0.01 d. ;H p < 0.05 (, p< 0.1 e.:< ก  B ก:<:%

(THAMLE :Thailand Medical Licensing Examination 2006) ก JกL++ case-control B ก =9 aspirin ก+ก (,T B=< :S%< + Odds ratio = 0.7  confident interval 95% ก+ 0.31.3 89 ;H) (, 2   confident interval 95% IB a. 0.3-1.3 b. 0.3-0.7 c. 0.5-0.9 d. 0.1-1.1 e. 1.3-2.9 (THAMLE :Thailand Medical Licensing Examination 2006) +%+ ;ก % ก ก;
--------------------------------------------------------------------------------------------------------------------------------------

C #" D3C)"/  !.. 1. Daniel W. W. Biostatistics: A foundation for analysis in the health sciences. Eight edition. John Wiley & Sons, Inc. 2005. 2. ก+,-. /.012-34+52. ก.6/1786.9:3;<1=1: ;<1=1;>.:6+4ก.6461:.6?,9/1@+- A6BC1DC3@EF.,Bก6G3D:./1H-.,+2546. 3. Dan Mayer. Essential evidence based medicine. Cambridge University press 2004. 4. David L. Katz. Clinical epidemiology & evidence-based medicine: fundamental principles of clinical reasoning & research. Sage publications. 2001.


ก  
  
 2 ( .300) !"#

SCREENING

.   

  

  . -------------------------------------------------------------------------------------------------------------------------------------------SCREENING Scope: The procedures which can be applied rapidly to sort out apparently well persons who probably have a disease from those who probably do not. 1. DIAGNOSTIC • It is the procedure to obtained clinical information from history, physical examination, laboratory and imaging procedures. 2. CONCEPTS OF SCREENING • Person with a disease can be identified by use of a screening test before the time of routine diagnosis. • Treatment at the time of detection by screening, as opposed to the time of routine diagnosis, results in an improved chance of survival. 3. TYPES OF SCREENING

• Mass screening: whole population • Multiphasic screening • Targeted screening • Opportunistic screening

4. CRITERIA FOR A SUCCESSFUL SCREENING • -Morbidity or mortality of the disease must be a sufficient concern to public health. • -Effective early treatment must be known to reduce morbidity or mortaliy. • -A high risk population must be exist. CRITERIA • The test should be sensitive and specific. • The test must be acceptable to the target population. • Minimal risk should be associated with the test. 5. VALIDITY • Sensitivity: The proportion of people with the disease who have a positive test for the disease. • Specificity: The proportion of people without the disease who have a negative test. • Positive predictive value: The proportion of persons with positive test results who actually have the disease of interest. • Negative predictive value: The proportion of persons with negative test results who actually do not have the disease of interest.

Sensitivity

True positive TP

+ FN

Specificity

True negative TN + FP

Negative predictive value

Positive predictive value

True positive TP + FP

Accuracy

True negative

TP + TN

TN + FN

Grand total


  (THAMLE :Thailand Medical Licensing Examination 2005) (d=9%=<H)=B9ก JกL clinical trial  H

( validity ) กH< A. T HJกL%+ ;ก C B. ก:<:% :ก:+ %+:ก% C. ก:<:% :ก:+) ก D. =9ก  ( randomization ) =ก  % E. ;< ก: % I<9 100%

(THAMLE :Thailand Medical Licensing Examination 2005) )++<%+=B;
I (,T 

: positive

60

40

: negative

20

80

89BJH <%+I<9: positive jT%ก (,T  I A. 0.3 B. 0.33 C. 0.6 D. 0.67 E. 0.75 (THAMLE :Thailand Medical Licensing Examination 2006) =ก

B Sj%I9B<ก=Iก 1,000  T<=9j);<=B +ก+ก <%+

kI<9:< j ก <%+

k

specificity % j);<=B;< (, 9%: IB a.60 b.86 c.71 d.92 e.83

(THAMLE :Thailand Medical Licensing Examination 2007)%

ก % C9=Bl% 70 (* (, 0.1/(* %

ก % C9(mT B= (, 0.08/(* %

ก % C9(mT   :)I9 (, 0.02/(* 8 life expectation% C9=Bl% 70 (* H (,T B=:  :)I9 a. 1 b. 2 c. 5 d. 8 e. 10 6. Cutoff point • It is the point on the continuum between normal and abnormal. • Receiver operator characteristics curve (ROC)
 

(THAMLE :Thailand Medical Licensing Examination 2007) กก <:% B =  กL SH% =9<ก % T H  8J ; 89Bก กL j 

 ;H 9 8B
(THAMLE :Thailand Medical Licensing Examination 2007) กก <:% B =  กL SH% =9<ก % T H  8J ; 89Bก กL j 

 ;H 9  8B
7. Ideality test • Inexpensive, convenient and painless, safe and 100% accuracy. • Gold standard is a jargon term, used to describe a method, procedure, or measurement that is widely accepted as being the best available. 8. Bayes> theorem

Bayes’ theorem Positive predictive value; (Se) (P) (Se)(P)+(1-Sp)(1-P)





Se = Sensitivity
Sp = Specificity
P = Prevalence


  (@ABCD@
BD)

(THAMLE :Thailand Medical Licensing Examination 2005) ก ;k Bl; % 60-70 (* %BVC;ก :ก: Sก+j)Bก ++ linear relation 9%=< %I(j< (, null hypothesis A. %ก+%BVC;ก :ก: S B. j)Bก Iก+%BVC;ก :ก: S C. j)  inverse relation ก+%BVC;ก :ก: S D. < B j)Bก ก+%BVC;ก :ก: S E.  B j)Bก ก+%BVC;ก :ก: S < 0.05 (THAMLE :Thailand Medical Licensing Examination 2005) =ก JกL(dH: % cerebrovascular disease I<9< j ODD ratio

95%Cl

ก C++B H

1.8

0.99-2.20

<SH 

0.4

0.35-0.45

C9Bl;

0.9

0.75-2.10

%กก 50(* 2.3

0,96-3.20

<T:B; C 1.55

0.44-2.85

9%=< (, protective factor A. B. C. D. E.

ก C++B H <SH  C9Bl; %กก 50 (* <T:B; C

(THAMLE :Thailand Medical Licensing Examination 2006)  =9%I =ก ) %  % %ก infarction ก+p ;ก

H (:HI( a. Ki-square b. Pair T test

(THAMLE :Thailand Medical Licensing Examination 2006) ก JกL ; 

 กHก+ก =9 (s;=ก

กLT %   =+ ; BJH C9)ก ;I<9 ก+9%C:I9 กHก+% 

<+ก JกL ก ;;กT :;<(s;HH ( =< (,++ Ratio a.% b.  c. <+ก JกL d.ก ;;uกT  e.;<% (s; (THAMLE :Thailand Medical Licensing Examination 2006) ) %==ก =9+ ;ก % .  C9=9+ ;ก ( : 1,000   :S%ก 100  T< :S%ก j : 10  (,ก ++=< a. cluster sampling b. simple random sampling c. systematic random sampling (THAMLE :Thailand Medical Licensing Examination 2007). ก )  CA breast stage 1,2,3,4 + C9(m 9%: 20,40,60,80 ( ก JกLj (,++=< (9%j 4  :S%ก) a. Ratio b. Interval c. Nominal d. Ordinal

(THAMLE :Thailand Medical Licensing Examination 2007) C  B BJH JกL กHก+

 %   9  4 ;< (%  w  < j 20,35,35,10 ก JกLj:กL9%C:++=< 1. Interval 2. Ordinal 3. Ratio 4. Nominal

-------------------------------------------------------------------------------------------------------------------------------------------Reference 1. B. Burt Gerstman. (1998). Epidemiology kept simple: An introduction to classic and modern epidemiology. New York: Wiley-Liss. 2. John M. Last.(1988) A dictionary of epidemiology. 2nd edition. New York: Oxford university press. 3. Raymond S. Greenberg., et al. (2001). Medical epidemiology. 3rd edition. New York: Lange/McGrawHill. 4. Leon Gordis. (2000). Epidemiology. 2nd edition. Philadelphia : W.B. Saunders company.

76b-476b-BAc- imork (Mc) 6.36 P.M. 27/10/2551