Effectiveness of Treatments for Patients Undergoing Surgery Report

This report presents a statistical investigation into the effectiveness of two treatments, Paracetamol and a placebo, on the time patients spend in surgery. The study involved 50 patients, divided into two groups, each receiving one of the treatments. Variables considered included the group of study, treatment type, age, gender, weight, height, and surgery time. The analysis employed the R studio software to analyze the data. The study aimed to determine if there was a significant difference in surgery time between the two treatment groups. The null hypothesis posited that the average surgery time for each treatment sample equals the average time of the entire population, while the alternative hypothesis suggested that they do not. The analysis revealed that the alternative hypothesis held true for both treatment groups. The linear regression model showed the presence of covariate variables affecting the surgery time, such as age, gender and weight. The findings suggest that Paracetamol serves better as a treatment substance during surgery and doctors should check other human factors that would affect the health of patients. The report includes descriptive analysis, t-tests, and predictive analysis using linear regression models to evaluate the impact of different variables on surgery time. The study concludes with a discussion of the findings and their implications for surgical practices.

Treatments 1
TREATMENTS EFFECTIVENESS TO PATIENTS UNDERGOING SURGERY
Name of Author
Name of Class
Name of Professor
Name of School
State and City of School
Date

Treatments 2
Abstract
The actual question that this report is actually going to answer is based on how two
treatments affect the time that patients undergoing surgery take in a surgery room. The first
treatment that was administered to the first group of patients was Paracetamol. The second group
was had a placebo treatment, a control group for the process. Paracetamol is actually used to
reduce acute pain in the body and soften the skin for easier medical operations during surgery.
The control treatment, a saline type of drug served the purpose of a Paracetamol.
After administering the two treatments that would determine how long a patient would
take in surgery, the time respective time was then recorded. Time was treated as the dependent
variable. Other variables are; a group of study, treatment, age, gender, weight, height and surgery
time.
The data was split into two blocks that were to be analyzed differently; the set of
hypotheses created was;
Null hypothesis: Average time of the treatment samples each equals the average time of the
entire population.
Alternative hypothesis: Average time of each treatment sample does not equal to the average
time of the entire population.
The methodological approach was a collection of the observational dataset. Which was
later analyzed in R studio as directed. With all the findings and the functions properly explained
in the data, methods, results and discussion sections.
Findings of the actual analysis are that the alternative hypothesis holds true for both the
sample of treatment datasets. The linear regression model to shows us how there exist covariate
variables which are actually affecting how long an individual takes in a surgery room.
The findings of the analysis are that Paracetamol serves better as a treatment substance
should be regularly during surgery. In addition, doctors should regularly check at other human
factors, that would affect the health or surgery process of a patient who undergoes surgery.
Introduction
As you can see from the study group variable of the data, it is evident that there was a
very large patient’s population. From the large population, an exercise of partitioning patients
into samples was conducted to help ease an analysis process. Using the entire population would
have proved to be costly. Therefore of this, a total sample of 50 was selected which was later
divided into blocks of twenty-five (25) each. The very first block was administered with
treatment one, whereas the second block was administered with treatment two. The treatments
would, later on, affect how long each patient lasted in a surgery room since the treatments would
soften the skin for easy cutting and reduce pain during and after surgery. The respective time
amounts that a patient who participated in the sample would then be recorded and used for
analysis. In the analysis process, age, weight, gender and height would serve as independent
variables. The variable Surgery time would serve as our response variable. Covariate variables
would then be.

Treatments 3
Since the Surgery time serves as our response variable, it is the variable that will enable us to
develop our set of hypotheses. The set of developed hypotheses would be something going like;
Null hypothesis: Average time of the treatment samples each equals the average time of the
entire population.
Alternative hypothesis: Average time of each treatment sample does not equal to the average
time of the entire population.
Data
The dataset was collected from an observational study. The observational study must
have been from those patients that visited a hospital wanting surgery to be performed on their
respective body parts. The actual sample is 50 in total and the sample, since there was a placebo
treatment and an actual treatment, was divided into two blocks. Each block had a farther sample
of 25 patients each.
The numbers of variables that are to be under scrutiny are seven in total and these are; a group of
study, treatment, age, gender, weight, height and surgery time. Surgery time is our dependent
variable, with variables like a group of study less important. The group of study variable is less
important because it only issues information of the study group that a patient originates from.
This is just a random pick and has no effect on the total time that a patient can take in a surgery
room when administered with any of the treatments. The actual variables that would act as
covariates are gender, age and weight. The reason for the existence of the covariate variables is
that gender in itself has different effects when it comes to hospital treatments. Both males and
females have got different physiological buildups. The difference in physiological buildups
affects the way in which both the genders respond to treatment. And therefore as you can see
each sample has a mix up of the genders just to help standardize the actual results output,
therefore there won't be a one-sided bias. As of age and weight, the older and heavier a person is,
the more different they will be responding to different types of treatments compared to the less
heavy and younger people. Hence these three variables are just as important covariate variables.
Moving on to the actual variables' types, we will be forced to introduce the software that we will
use for our data analysis at this point. For us to see the variable types as per the two data samples
blocks, we will use the command str(data name) (Marlow et al. 2015). From the command, we
will run one for the first treatment block of the dataset and another one for the second treatment
block dataset. This by far will give us different variables types for each block variable dataset.
The first dataset sets study group, treatment, age, gender, height and surgery time as integer
variables and leaves weight as a numeric variable. Moving on to the variable types, we have that
weight and heights are both numeric variables whereas study group, treatment, age, gender and
surgery time are integer variables. As from the variables' types, one is able to see that there are
no categorical or nominal or qualitative variables. This by chance shows us that we will not have
to change from nominal to numeric for the analysis that only needs numeric datasets or variables.
The grouping of the data is done in blocks of 25 each, this is because there were two treatments
that were to be administered and therefore we the sample data had to be sampled even farther.
This is done in groups of 25 each to help create balance.

Treatments 4
One of the procedures that were employed to reduce sampling error and invoke bias is having
both sets of treatment data sets having mixed variables of males and females in each. There is no
dataset that only has male and the other females, both have a mixture of both. This can be seen
by the numbers 1 and 2. Where 1 represents females and 2 represents females. A farther action
that reduces biases is the mixture of ages together and there is no dataset having very old people
whereas another has very young people.
Methods, Results and Discussion
For the analysis of the dataset, a request was placed for the use of R studio software. This
request with no doubt was followed to the latter. R studio serves best when it comes to data
analysis at it offers the best that R can offer. This is because it has up to four tabs that open up
when it comes to using it in an analysis. Of the four tabs, one offers a display area for the results
that have been analyzed. Therefore one is able to rectify a mistake when one is made because it
is not the part that you write your codes that will display your results. There is too another
window that is more of a help window that helps with analysis.
Moving to our data analysis, we will have to analyze the two treatment blocks of datasets
differently as we will be trying to answer to two hypotheses. Also, it is important to note that
analyzing the two datasets is very incorrect ass will bring about biases. The biases originate from
the fact that there are two treatments. Since there are two treatments, the separate analysis would
suffice as it helps us find out which treatment is better than the other.
Before any analysis the data has to be uploaded into R studio using the commands;
WorkBook1A=read.csv(file.choose(), header = T), WorkBook2=read.csv(file.choose(), header =
T) and WorkBook3=read.csv(file.choose(), header = T).
The further analysis we move into doing a descriptive analysis of the datasets using the code
summary(data set) of the different treatments which gives the mean, median, maximum values
and the minimum values of the respective variables.
Moving progressively into the analysis of the hypothesis to find out which of the actual
hypothesis bit holds true. Is it the null hypothesis or the alternative hypothesis? Having the
surgery time as our response variable, we will have to use the average of this variable in our
entire hypothesis testing. Therefore to get the mean of the surgery time to our whole sample
population, we will use the code; mean(DF) which give the value of up to 254.98. DF is the
dataset that we get after subsetting the WorkBook1A dataset by deleting the first six columns.
Since we want to test hypothesis using t-test, we will also get the mean for the surgery time
relative to the two types of treatment datasets. This eventually leads us into subsetting the first
treatment data set using the code; DF1=WorkBook2[, -c(1:6)] and applying the code; mean(DF1)
which gives a mean value of 252.88. The same is done to the subset the second treatment data set
using the code; DF2=WorkBook3[, -c(1:6)] and applying the code; mean(DF2) which gives the
mean of 257.08. (Wood, 2017)
The t-test thereafter comes in to play where we confidently use the code; t.test(DF1, mu =
254.98). This means that the first treatment dataset has the mean of the surgery time variable
equal to the mean of the entire population. Running the code, we get a p-value which is greater
than 0.5 and that stands at 0.8319. This gives us a strong loop to reject the null hypothesis and

Treatments 5
pick the alternative hypothesis. The same applies to the second treatment dataset where the p-
value stands at 0.8479.
As we wrap up our analysis we perform predictive analysis. This is done by conducting the
linear regression model on each dataset. Using the codes; LR1=WorkBook2[, -c(1:2)],
View(LR1), LR2=lm(Surgery.time~ ., LR1) and summary(LR2). This gives us how various
variables would affect the actual time that a patient takes in a surgery ward. Looking at the
estimate column after calling the summary function, we find that we have negative or positive
values. This means for every additional amount of a respective variable, time that is spent in the
surgery room is either taken away or added.
Father plot functions give diagnostics plots as well. When doing a boxplot, the only dataset that
belongs to the second treatment has a single outlier.
Conclusion
It is evident that there are several variables that affect how long a patient stays in a
surgery room and not only treatment as seen by results from the actual analysis conducted.
Doctors should, therefore, ensure that they note all the factors that are important to note and that
actually affect the medical life of each and every patient. This, when done, enhances surgery
processes and other medical attention provided to a patient.

Treatments 6
References
Marlow, C.A., Viskontas, I.V., Matlin, A., Boydston, C., Boxer, A. and Taylor, R.P., 2015. The
temporal structure of human gaze dynamics is invariant during free viewing. PloS one, 10(9),
p.e0139379.
Wood, S.N., 2017. Generalized additive models: an introduction with R. Chapman and
Hall/CRC.