Essay on Statistics 3.10

Submitted By cookiechick
Words: 918
Pages: 4

I am interested in the way that the gestation period during pregnancy affects the survival of an infant. Today, preterm babies have had a much greater chance of survival that a few decades ago, due to improvements in technology and advancements of treatment available. However, being born too early can still have health risks, and according to [http://www.birth.com.au/Premature-baby/Survival-of-preterm-babies-gestation.aspx#gestation], “the older the gestation of the baby, the better off they tend to be”. On the same site, it is stated that 80% if preterm births occur around 32 -37 weeks of pregnancy. About 20% are earlier than 32 weeks. Measures of general health, hospital admissions and long lasting illnesses showed a gradient of increasing risk of poorer outcome with decreasing gestation (http://www.bmj.com/content/344/bmj.e896). This suggests that the health outcomes of preterm/early term babies are worse off than full term babies. Below is a table showing the likelihood of survival with different ages of gestation.

COMPLETED WEEKS OF GESTATION AT BIRTH(using last menstrual period) | CHANCE OF SURVIVAL | 21 weeks and less | 0% | 22 weeks | 0-10%* | 23 weeks | 10-35% | 24 weeks | 40-70% | 25 weeks | 50-80% | 26 weeks | 80-90% | 27 weeks | >90% | 30 weeks | >95% | 34 weeks | >98% |
-------------------------------------------------
* Most babies at 22 weeks are not resuscitated because survival without major disability is so rare.
-------------------------------------------------
(http://www.spensershope.org/chances_for_survival.htm)

From this table we can observe that the fewer weeks of gestation, the less chance there is of survival for the baby. A baby born any less tha 25 weeks has a very low chance of survival. Therefore it seems reasonable to predict that in my data set, babies with a younger gestation “age” have a poorer chance of living than babies with an older gestation age. My investigative question is, “I wonder what the mean difference is between weeks of gestation before birth and the survival of the baby in the data set.” I chose “weeks of gestation as my numeric variable because very often babies are born prematurely and can be closely related to the survival of the baby (my categorical variable).

Summary of Gestation by Survival
Min. 1stQu. Median Mean 3rdQu. Max. Std.dev Sample.Size No 23 27.0 29 28.83 30 36 2.313 103
-------------------------------------------------
Yes 22 27.5 29 29.07 31 38 2.385 477

In my sample, there is a shift of ___ weeks of gestation from the mean of babies not surviving, from the mean of babies surviving. This suggests that the babies with shorter gestation period have only a slightly smaller chance of survival than babies with a longer gestation period. The distribution of both surviving and non-surviving babies is bell shaped, showing that most gestation ages are normally distributed. It also means that most values fall in the middle with few values falling to either of the extremes. The middle 50% of gestation of surviving babies is spread over ____, while the interquartile range for non-surviving babies is ____, so there is hardly more variation in the surviving babies in my sample. Furthermore, the standard deviation between the two are practically the same: Surviving babies have a standard deviation of 2.385, and non-surviving babies with a standard deviation of 3.313.

Bootstrap

The average difference in the sample between the mean gestation for surviving babies and non-surviving babies is