community sheet #1 Dr.faro8 - Tala Mazahreh

[Shadi Jarrar]

بسم الله الرحمن الرحيم

_______________________________
http://www.4shared.com/file/U75biPRj/Epidemiology_sheetJude.html
_______________________________

Epidemiology
Sheet #1 Epidemiology
Dr.Farouq Shakhatreh
By:Tala Mazahreh

Epidemiology : is defined as "the study of the distribution and determinants of a disease in a human population"
From the previous definition, there are 3 basic terms that we are going to discuss:
 Distribution of the disease:
which is studied according to 3 main categories:
Person : Personal characteristics such as age, gender, social class, size, family history of certain diseases and so on.
Place: any disease can be defined according to place which may be either the area of residence or the place of work.
Time: which can be divided into 3 variants
1)Long term variation (Secular trend)
2)Periodic or Cyclic Variation
3)Short term variation (Epidemic variation)


Now the question is, why do we (as dental students) concern about epidemiology (or even biostatistics)?!!
This is a very important subject for us, since we will be asked at some point in our career to do a research, and having the basic knowledge in epidemiology and biostatistics is a must for establishing a good research.
For example a group of dentists in Jordan were recently interested in doing a research about "Fluorosis" which is defined as high fluoride intake. Fluorosis has a high prevalence rate in south Jordan due to the high percentage of fluoride in water found in wells.
In such a research we do what is called "A case control study" to find out the relationship between two variants (Fluorosis and type of drinking water ), or to find out Odds ratio (OR) or the Relative Risk (RR)
{We will discuss these terms in more details later on}, and finally to able to come to a conclusion such as that
**high fluoride in water sources causes Fluorosis**

Another example for an epidemiological study that was done in Japan in 1954 in a number of villages, where 3 o them was found to have high amount of fluoride in drinking water, while on the other side of the country another two villages had very low amount of fluoride in drinking water, and it was found that the prevalence rate of dental caries was much higher in areas with low fluoride in drinking water . This is called Community Trial, and since that time a lot of epidemiological studies, covert studies, case control studies and many clinical trials were done in this issue which resulted in adding fluoride to the components of toothpastes…

Another epidemiological study( a cross-sectional study) was done in Jordan on children 13-14 years of age , and it was found that prevalence rate of Myopia "nearsightedness" was 25% , so any child in this age suffering from headache and tends to sit close to a screen or T.V. is most likely to be suffering from Myopia.

We should also be familiar with the terms Prevalence and incidence. Prevalence rate of dental caries in Jordan in children less than 10 years of age is around 75%

Another example relating epidemiology and biostatistics to our career, we have a piece of data:
Prevalence rate of dental caries among children in Amman is for example 60%, and the same percentage equals 72% in Irbid for example…..Now is the difference between these two percentages significant at the 5 percent level ( P-value=0.05 )?!
Here, since we already have percentages(the prevalence rate) then the most appropriate test to be used is the Chi-Square Test [because we have difference in percentages but never ever use the T-test] and this will give us the P-value.
Now,
If P-value <= 0.05………….then this difference is significant.
If P-value >=0.05…………..then it is not significant.

So in the above mentioned example , if the P-value was calculated and it was equal to 0.03 then we can say that the difference is Significant.

Example: The average height of Males is a certain population = 170 cm
The average height of females= 165 cm.
Is the difference in height significant?!
Let's assume that the P-value here is calculated (here by using the T-Test) and it was equal to 0.08, then we can say that the difference is Not Significant
When we say that the P-value equals 0.08, this means that 8% of the difference between the heights is by chance and 92% is different.
Always when we have difference in percentages use Chi-Square test, or you may (to a lesser extent) use Fisher's Exact Probability Test(which is a variant of Chi-Square test used in special conditions).
When you have difference between two means always use T-Test.
But we have to pay attention here, if we have 3 means we can't use the T-Test, in this case there's what's called ANOVA Test (Analysis of variants) which is based on a computer programme.
**We have three computer packages that are used in data analysis:
1)SPSS (Statistical Package for Social Science)
2)SAS
3)Epi Info (which is used by the WHO)


That's all for today 

P.S: This lecture was supposed to be an introduction to epidemiology, but obviously it was Not !! Everything discussed here concerning biostatistics will be discussed in details later on, don't worry about it =)