Homework #3: Rate Adjustment

Homework #3: Rate Adjustment - Session 6: Direct and...

Info iconThis preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
Epi 420: Session 6 (Adjustment of Rates) Page 1 of 7 Revised 04-12-2008 Session 6: Direct and Indirect Adjustment of Rates ADJUSTMENT OF RATES Comparing crude rates can sometimes be misleading. This is because other factors might be the root cause of differences in crude rates. Some definitions of terms are useful to fix in your mind the language of rate adjustment. We start by studying an event rate. The event is typically called the outcome or disease (D). We are investigating how this event rate varies across some other factor or variable which is called the exposure (E). The exposure is a very general concept and can refer to almost any variable or measure; when performing rate adjustment in descriptive epidemiology the exposure is typically a geographic unit such as cities, counties, states or countries. We also have a third factor or variable that we refer to as the confounder (C). The confounder is a factor that might distort the relationship between the exposure and disease. Confounding occurs when there is another factor (C) that is related to the disease (D) and is also differentially distributed between the groups (E) you are comparing. An example is provided when examining differences between crude mortality rates in two different populations (say, Sequim and Fremont). Let's say that the crude mortality rate is much higher in Sequim than Fremont. Before we implement a multi-billion dollar intervention to lower the mortality rate in Sequim we might want to know if some other factor could explain the difference in rates. Age is a factor that naturally comes to mind since it is strongly related to mortality. If Sequim and Fremont have very different age distributions this might provide a simple explanation for why the crude rates are different. The population with a larger proportion of older people will typically have a higher crude mortality rate. Suppose you want to know whether, after you account for age differences between the two populations, the difference in mortality rates persist. To answer this question you need a method of statistically removing the effects of age so that you can make a valid inference. The technique for accomplishing this is called rate adjustment or rate standardization . Standardization is an old-fashioned method going back to the time of John Snow and was first introduced by F.G.P. Neison at a presentation to the Statistical Society of London in 1844. While old, it is still widely used in epidemiology and definitely worth mastering. There are two distinct methods of standardization: direct and indirect . The direct and indirect methods each have a place in the epidemiologist's tool kit. Both methods can be used in a wide variety of situations. You should also note that while typically illustrated with mortality, the method is not limited to mortality and can be used for almost any dichotomous event (yes/no) you wish to study - births, marriages, divorces, pregnancies, etc. Similarly, the confounding factor or factors can encompass a broad range of characteristics
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 05/12/2008 for the course EPI 420 taught by Professor Goldberg during the Spring '08 term at University of Washington.

Page1 / 7

Homework #3: Rate Adjustment - Session 6: Direct and...

This preview shows document pages 1 - 2. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online