dreamer
Forum Elite
Topics: 52
Posts: 152
|
hi guys, i am reading 2004 kaplan bs notes page 7 and i get some confused qs .why increased incidence doesnt affect the positive predictive value while increase prevalence can ?? could someone please explain to me???
___________________
you may say i am a dreamer,but i am not the only one...........
|
| dreamer
Forum Elite
Topics: 52
Posts: 152
|
may i add one more q? 2004 kaplan page 12,why does cross sectional studies can tell the prevalence while case control and cohort cant ? i think in the case control studies, since it is retrospective, you have already known who are sick pp who are healthy ones, so you can calculate the prevalence??
somehow i get this feeling that i mess stuff up ( plz help
___________________
you may say i am a dreamer,but i am not the only one...........
|
| Leopard
Forum Guru
Topics: 30
Posts: 401
|
Agreed incidence directly does not affect the PPV but indirectly does effect because
Prevalence = Incidence x duration
If incidence changes ,prevalence changes which in turn does affect the PPV.
As far your 2nd question is concerned purpose and way of studying of of Cross sectional study from Case - control and Cohort study.
Cross sectional study = It is a "snap shot" at some particular time about the cases present.It does not tell you about the relationship of disease and risk factors
Case - control study = As its name implies ,it is comparsion of Cases (patients) and controles (normals) with regard to exposure and non exposure to risk factor.
Cohort study = Again this has nothing to do with number of cases .It is the study of people who are exposed to a risk factor.These are followed up for some time to know the relationship of that risk factor with occurance of disease.
Hope it helps.
Best of luck
___________________
FA is just a good revision book.It is not a "real" learning tool.
|
| AutumW
Forum Newbie
Topics: 0
Posts: 13
|
Ah, you know what review program has a really good explanation of this is Score95.com. I just purchased a membership and I was checking in their notes about what guys are talking about. I wanted to post what their explanation, but I guess its against the rules of the forum.
Cheers!!!
|
| Slacker
Forum Elite
Topics: 29
Posts: 141
|
Leapard, correct me if i'm wrong, but wat i learned is that the Prevalence & Incidence relationship doesn't work in reverse...change in incidence can change prevalence, but change in prevalence CAN'T change incidence. This is not simply a mathematical relationship. this is medicine... thinking in reverse is illogical. the equation is simply to show the effect of incidence and duration of prevalence.
Prevalence = Incidence x duration
|
| Leopard
Forum Guru
Topics: 30
Posts: 401
|
BTW ,did I say that that change in prevalence changes incidence ??
Another question is for you.Let us suppose in one community presvalenc of disease is 1000 cases /yr and in another community prevalence of disease is 2000 cases /yr ,what do you think incidence is equal or different and how will you calculate it.
Best of luck
___________________
FA is just a good revision book.It is not a "real" learning tool.
|
| dreamer
Forum Elite
Topics: 52
Posts: 152
|
i guess you cant tell the incidence or calculated it from this data directly? coz prevalence include not only the pp who are catching the disease in that year but also who have already had it before that year (assume it chronic disease so it would be more obvious) .so my answer would be you cant tell the incidence dirrectly from the number of prevalance. not sure if i get this right .plz correct me ...
btw, i am still confused how you get this conclusion : the higher the prevalence of a disease,the higher ppv of the test. is it just logical thing or it is about calculation??? could someone explain to me?thanks
___________________
you may say i am a dreamer,but i am not the only one...........
|
| Leopard
Forum Guru
Topics: 30
Posts: 401
|
How prevalence affects the PPV.You take to hypothetical situation for population of 200 for a disease for which the test sensitivity is 90% and specificity is also 90% and make 2x2 figure.
1st situation
No of disease people = 100
No of healthy people = 100
prevalence = 100/200 = 50%
True positive = 100 x 0.9 = 90
False negative = 100 -90 = 10
True negative = 100x 0.9 = 90
False positive = 100-90 = 10
PPV = 90 /90+10 = 90%
2nd situation (prevalence is changed)
No of diseased persons = 150
No of healthy persons = 50
True positive = 135
False negative = 15
True negative = 45
False positive = 5
PPV = 135/135+5 = 135/140 =93%
No you might have realized that by changing prevalence the PPV changes
Best of luck
___________________
FA is just a good revision book.It is not a "real" learning tool.
|
| dreamer
Forum Elite
Topics: 52
Posts: 152
|
thanks for the nice explaination
___________________
you may say i am a dreamer,but i am not the only one...........
|
|
| |
| | | | | | | | | |
