Biostat question

[AndyRSC]

---

Members don't see this ad.

I recall being puzzled, as I studied biostats for step I, why in the formula

Incidence x Duration = Prevalence

the prevalence does not inform / feed-back to incidence in such examples as HIV. I recall book tables and questions where prevalence is shrunk (cure) or increased (better treatment > longer lifespan), yet as numbers of HIV-positive individuals shrinks or grows, incidence remains the same. This would certainly apply to epidemiology of non-infectious diseases, but not diseases that can only originate from prevalence.

Now that I'm reviewing biostats for step II, this irked me all over again. Am I missing something obvious?

 

[lemonade90]

I recall being puzzled, as I studied biostats for step I, why in the formula

Incidence x Duration = Prevalence

the prevalence does not inform / feed-back to incidence in such examples as HIV. I recall book tables and questions where prevalence is shrunk (cure) or increased (better treatment > longer lifespan), yet as numbers of HIV-positive individuals shrinks or grows, incidence remains the same. This would certainly apply to epidemiology of non-infectious diseases, but not diseases that can only originate from prevalence.

Now that I'm reviewing biostats for step II, this irked me all over again. Am I missing something obvious?

Are you asking why prevalence decreases if incidence is constant and people still have the the HIV virus? I am having difficulty understanding with the wording of your question.

 

[HelpPleaseMD]

BRS gave an example of white women and black women having the same incidence but different prevalence. The black women had a lower prevalence since they had a shorter duration of disease (died earlier). Not sure if I answered your question.

 

[tjquinn]

Treating a patient does not usually change incidence because you have not treated the original cause of the disease. For instance, our treatment for heart disease leads to an increase in the number of people that have heart disease (since they live longer). However, the original cause of the heart disease remains unchanged (diet, exercise, genetics, etc.).

Also, if we can cure AML then the prevalence of AML may decrease since less people will have the disease at any given time; but, the fact that we wait for AML to exist before we treat it means that the incidence of AML will remain the same.

The only way to alter incidence would be via prophylaxis or the use of screening tests to eliminate the disease before it rears its ugly head. For example, PAP smear reduces the incidence of cervical cancer because we can detect pre-cancerous lesions and treat the patient before they even develop the disease.

Remember, incidence is simply the number of new cases in a given period of time and therefore can only be reduced if you prevent new cases from existing.

Prevalence is simply the total number of cases in a given period of time and thus the act of curing or treating the disease can modify this.

 

[AndyRSC]

I think I didn't explain what I meant well enough, so here's an example:

I recall Kaplan videos and questions with up/down arrows confirming this premise.

tjquinn, it was a good write-up, but what I'm asking falls between the last two of your sentences. Why prevalence does not feed-back to incidence when new cases may only come from existing ones (infectious disease).

 

[lemonade90]

I think I didn't explain what I meant well enough, so here's an example:

I recall Kaplan videos and questions with up/down arrows confirming this premise.

tjquinn, it was a good write-up, but what I'm asking falls between the last two of your sentences. Why prevalence does not feed-back to incidence when new cases may only come from existing ones (infectious disease).

Your point is valid and I agree with you. If you have decreased prevalence, that means that you should have less incidence in diseases/organisms which have reservoirs only in humans (e.g. Tric Vag); fewer people with the disease should mean less opportunities for exposure. However if I recall correctly, I think HIV has parallels in other organisms and was thought to be transmitted from them (e.g. monkeys), so they do have a reservoir outside humans.

Perhaps a way of thinking about it is in terms of t=0 and t=1. At t=0, you will have decreased prevalence. Once you have decreased prevalence, at t=1, you will have decreased incidence as well. However there will be a time lag and it will not be immediate. I guess Kaplan was trying to ask the question at t=0.

 

[AndyRSC]

It's a good point, and I considered it as well; it would probably apply to decreased prevalence (cure), but not increased prevalence (longer lifespan).

 

[lemonade90]

It's a good point, and I considered it as well; it would probably apply to decreased prevalence (cure), but not increased prevalence (longer lifespan).

I guess you are right as in the case of infectious diseases either incidence or prev can be made the dependent and indep vars.