The next step in the problem solving methodology, is creating a conceptual framework. A conceptual framework is a visualization of how we think the world is working. And as Paul Hughes said, a conceptual framework is a frame that works to put those concepts into practice.

As you can see in our figure, once you have a problem definition that you think is close to your final version, remembering that you can revise your problem definition at any point in this methodology, you have meaningful measures of the magnitude of the problem. So now you're ready to move on to drawing a picture of the problem. That's really what the conceptual framework is, a picture of the problem that includes key determinants of the problem.

So before you start drawing your picture, you turn to the literature and existing information. You review peer-reviewed literature, Grey Literature which includes expert opinions. And consensus statements, you search the web, you look throughout your individual agency to see if there's additional information that may be unpublished. You use all of these resources to identify key determinants. Although there are maybe not a lot of literature about key determinants of infant mortality specifically in Baltimore City. There's certainly a lot of literature about key determinants of infant mortality in other settings that could be applicable to Baltimore City. So you search for information and literature and you start making lists of key indicators or determinants of infant mortality. When you're looking at the literature and other sources of information don't only observe the key determinants. Be sure you are paying attention to the outcomes in these studies and other pieces of information. Do they use infant mortality rate? What other indicators of infant mortality are being used? Create a list of the outcome indicators next to your list of key determinants.

Next, focus in on your key determinants. Sort that list of key determinants into proximal and distal determinants. Proximal key determinants are the factors closest to your outcome. We have two categories of proximal determinants. Direct causes of the outcome and the individual level characteristics of determinants that may or may not be causal. The importance of identifying proximal determinants is that these are the determinants we target for intervention. These proximal determinants help answer our question about how to best intervene. Intervening on proximal determinants typically results in a greater decrease in the problem, as compared to when we target distal determinants.

Next we identify the distal determinants in our list. And these are the determinants that are little bit further removed from the outcome. These are often times associated but not causally linked to the outcome. We'll be using two categories for distal determinants. The first are environmental determinants, which are characteristics of the environment. The other are non-modifiable distal determinants, and these are typically individual level characteristics that are not modifiable with public health intervention like changing someone's education level. Alternatively, they could be truly non-modifiable, like race, something you are born with. The importance of the distal determinants and more specifically the non-modifiable determinants is that these determinants help us to identify the group of people we should target for intervention. Now that we've sorted our key determinants, we can begin to draw them into a picture. The first box on the far right side is the outcome. For our problem definition we can simply write infant mortality. Notice we did not write infant mortality rate because we know that there are numerous statistics we could use to measure infant mortality. That being said, the infant mortality rate is standardized to the size of the population because the denominator is the number of live births. So there are benefits to using this specific indicator of infant mortality rate. So let's start drawing our picture. First, we draw in the proximal indicators that are the direct causes of infant mortality. Here I've identified congenital malformations, low birth weight and short gastration, respiratory distress, sudden infant death syndrome or SIDS, and accidental suffocation or strangulation bed. These determinates can directly result in infant mortality. I found them in the literature and other web based resources.

We can draw end arrows to show their direct link with infant mortality and we can also draw end arrows to show the relationships between these proximal causal key determinants.

Now we can move on to proximal key determinants that are individual level characteristics of mother and baby.

Let's pause right here to work it out.

You can write down the proximal individual level characteristics of mother and baby, and draw lines from these determinants to other determinants currently shown.

If you're unfamiliar with this content area of infant mortality rate, please do attempt this exercise with key determinants you think may fit the proximal, individual level characteristic category. It's okay to guess because drawing this framework is also an iterative process like the drafting of the problem definition. Click Continue when you're ready to move on.

Let's compare proximal individual level determinants of mother and baby. I have listed low maternal intake of magnesium and folic acid. Lack of prenatal care, smoking, alcohol and drug use, and malnutrition of either the mother or the baby, as proximal individual level determinants. You will notice only one is directly applicable to the baby, malnutrition. The others are characteristics of the mother. I have also drawn the lines between these determinants and our proximal causal determinants. Notice that none of the proximal mother-baby determinants have arrows directly drawn from them to infant mortality. If one of these proximal mother-baby determinants was directly causal with infant mortality, we would want to move it into the proximal causal determinants category.

Let's keep working on this exercise. The next category of determinants is distal environmental determinants. You give it a try.

Add distal environmental determinants on the appropriate arrows into our conceptual framework. Click Continue when you are ready to move on.

Let's compare again. I have added air pollution, parental chemical exposures and unsafe sleeping environments including sharing a bed with the infant. As distal environmental determinants,again, these determinants have been shown in the literature to be associated with infant mortality. You can see I've drawn in arrows that link these distal environmental determinants, to proximal causes of infant mortality. If applicable, the arrows could have also been drawn to the mother baby proximal characteristics.

Last category of determinants. Distal, non-modifiable determinants. Add your distal non-modifiable determinants in appropriate arrows. Click Continue when you're ready to move on. The distal non-modifiable determinants I identified were African American race of the mother and or baby, adolescent age of the mother, low educational attainment of the mother, and short inter-pregnancy interval. You could argue that these could be listed under the proximal mother baby category. I chose to put them under non-modifiable distal determinants because I like to use this category as the keeper of the determinants I will use to target sub-groups for intervention. This is a stylistic decision.

The two main uses of the conceptual framework are to identify the who and how of your intervention. For organizational purposes the determinants like guide who you'll intervene upon are best kept in the individual non-modifiable characteristics but could also be in the proximal individual level characteristics, if you find that makes more organizational sense to you. The determinants that will guide how you intervene, that is, what types of interventions would most likely result in a decrease in the infant mortality rate, are going to be found in the proximal causes. For example, in order to address the proximal cause determinant like accidental supplication or strangulation in bed, one intervention maybe to provide new mothers with this safe sleep space for their babies outside of their mothers bed. Such as a bassinet with sleep sacks or other materials to keep babies sufficiently warm without the risk of accidental strangulation.

Although we provide guidance for creating conceptual frameworks. There is a lot of art in creating these frameworks. With practice and experience, you will develop your own style for the conceptual framework that best suits you, and the public health problem that you are working on.

Conceptual Framework

Epidemiology in Public Health Practice

Meet the Instructors