Analysis Of Hepatitis Data: Association Between Spleen Palpable And Class

The Adverse Effects of Hepatitis on the Human Body

Hepatitis is commonly a virus affected disease that brings an inflammatory condition of the liver (Lok & McMahon, 2009). The adversities that Hepatitis generally causes in human body are:

• The lack of bile production
• Excretion of bilirubin in excess amount
• Fatigue
• Pain in abdomen

Spleen is located in the left upper quadrant of the abdomen in human body. Spleen controls many crucial functions of the body that are elimination of abnormal cells, storing of iron and blood and providing protection against infections. Human beings can live without proper function spleen. However, liver plays a crucial role to compensate that and it makes liver more vulnerable.

Except slender people, a normal-sized spleen could not be palpated during the physical test of the abdomen. It is often seen that while affected by hepatitis, size of spleen grows significantly. The incidence rate of death gets higher for hepatitis patients, when people are exposed to palpation of spleen. The situation gets worst and patient may die. Hence, spleen palpability could be a significant critical aspect that endanger and enhance the risk of life at the time of hepatitis.  

The diagnosed results of the patients of a hospital is the proposed data in this analysis. Dr. G.Gong of Carnegie-Mellon University cleaned data and presented in new form. The data set has total 20 variables. Among which “Age”, “Bilirubin”, “Alkaline Phosphate”, “SGOT”, “Albumin” and “Protime” are numerical variables. Rest of all the variables are nominal variables. Only two variables are taken under consideration for this report. These variables are “Class” and “Spleen Palpable”. The variables are nominal (categorical) in nature. Therefore, for finding the association between these two variables, the research carried out chi-square test.

Hepatitis is assumed to grow palpability of the spleen. The spleen gets more palpable if hepatitis infects the human body. The palpation causes the bigger size of the spleen. The palpable and enlarged spleen is a notable indication of a variety of illness. Supine and flexed knees are the prominent indications of splenic enlargement and palpability at the time of hepatitis. Extensive palpation of spleen generates weight loss, lymphadenopathy and night sweats in Hepatitis may lead patient to the death.

Figure 1: Palpation of Spleen

At the time of hepatitis, size of spleen increases up to three or four times (Smith and Custer, 1946). In this research, the researcher suspects that palpation of spleen is a prominent reason of death in hepatitis. The research is particularly aiming on this analysis.  

Data Set Used in the Analysis

The objective of the research on hepatitis is to find the association or independence between the chosen two factors “Class” and “Spleen Palpable”. “Class” refers two categories that are longevity and morbidity (Gerspacher-Lara et al., 1998). The probable solution and measures could be interpreted from the analysis of the research.

The data set undertaken for analysis is “Hepatitis” data set. As per the requirement of the research, the data is publicly available in several websites that stores data. The “UCI learning” website is one of them. The data is collected from online resources. Hence, it is a secondary data.

As per online reviews, the actual source of the data is unknown. At the initial stages, Diaconis and Efron (1983) worked with rough data. Then, Konenenko et al. in the year of 1987, filtered the data and attempted for analysis. That time “Hepatitis data” is treated as primary data. As far as reported lastly, eminent data analysts such as G.Gonj and Bojan Cestnik of Carnegie-Mellon University structured the Hepatitis data in November, 1988.

With the help of “Observation” method of survey procedure, the readings of sample parameters are granted. It is notable that “Observation” method is a field of enumeration betrothed for a defined topic accomplished by the technique of “Complete enumeration method” or “Simple random sampling”. The hepatitis data is collected here by complete enumeration method from a specific healthcare home. The data is collected through a specific span of time interval.  

The data is not manipulated deliberately. However, as the data is structured many times by handling of different data scientists. Therefore, little bit of bias may present here. However, the current data is reliable and valid.

To avoid bias, the researcher did not attempt to manipulate the data set in any circumstances.

Table 1: Table shows frequency distribution of “Class”

Class	Count of Class
Die	31
Live	119

Figure 2: Bar chart displays the frequency distribution of “Class”

Out of 150 patients, 31 patients have died of Hepatitis and 119 patients have survived from Hepatitis.

Table 2: Table shows frequency distribution of “Spleen Palpable”

Spleen Palpable	Count of Spleen Palpable
no	30
yes	120

Figure 3: Bar chart displays the frequency distribution of “Spleen Palpable”

Out of 150 patients, 30 patients did not have spleen palpation and 120 patients had spleen palpation.

Table 3: Table shows frequency distribution of “Spleen Palpable” with respect to “Class”

Class	Spleen Palpable	Count of Spleen Palpable
Die	no	12
Die	yes	19
Live	no	18
Live	yes	101

Figure 4: Grouped bar chart displays the frequency distribution of “Spleen Palpable” with respect to “Class”

1. 12 people whose spleen was not palpable, died of Hepatitis.
2. 19 people whose spleen was palpable, died of Hepatitis.
3. 18 people whose spleen was not palpable, survived from Hepatitis.
4. 101 people whose spleen was palpable, survived from Hepatitis.

From analysed data, it looks that the hepatitis patients whose spleen is not palpable, has a significant chance of curability. Simultaneously, spleen palpability generally increases the chance of death at hepatitis.

Frequency Distribution of Class and Spleen Palpable

Null hypothesis (H₀): There is no statistical significant association between “Class” and “Spleen Palpable”. That is these two factors are independent to each other.

Alternative hypothesis (H_A): There is statistical significant association between “Class” and “Spleen Palpable”.

     The level of significance of Chi-square statistic is assumed to be 0.05.

     Chi-square statistic is shown as: 

O_i = observed value, E_i = expected value, c = degrees of freedom (Sharpe, 2015).

Table 4: Frequency table for analysis

Rows: Class   Columns: Spleen Palpable

	no	yes	Missing (*)	All
die	12	19	1	31
live	18	101	4	119
All	30	120	*	150

Table 5: The basic pivot table of frequencies

Count of Class		Column Labels			Actual Relative
Row Labels		no	yes	Grand Total	frequency
Die		12	19	31	0.21
Live		18	101	119	0.79
Grand Total		30	120	150
Actual Relative frequency		0.2	0.8		1

From the basic pivot table, calculated expected frequency table is:

Table 6: The expected frequency table

Class/Spleen Palpable	no	yes	Grand Total
Die	6.2	24.8	31
Live	23.8	95.2	119
Grand Total	30	120	150

The calculated Chi-square statistic = [( = 8.549.

The degrees of freedom = (number of rows-1) *(number of column-1) = (2-1) * (2-1) = (1*1) =1.

Table 7: Table displays the outcomes of Chi-square test

	Chi-Square	DF	P-Value
Pearson	8.549	1	0.003
Likelihood Ratio	7.616	1	0.006

The Chi-square statistic () is 8.549. The right tail p-value of Pearson Chi-square statistic = 0.003. The calculated p-value is less than 0.05 (critical p-value).

In this data, out of 155 observations, 5 observations have missing values. The researcher neglected these missing values and chi-square test is executed with the cleaned data. The “live” (a type of “Class”) is an indicator of “Survivability” and “Curability” from Hepatitis. It looks that the hepatitis patients whose spleen is not palpable, has a significant chance of curability. Simultaneously, spleen palpability generally increases the chance of death at hepatitis.

Figure 5: Chi-square distribution plot with 1 degrees of freedom (d.f. = 1)

Therefore, the null hypothesis of absence association of two factors or presence of independence is rejected at 5% level of significance. The alternative hypothesis of association of Class and Spleen Palpable is accepted with 95% probability.

Hence, the presence of spleen palpability significantly influences the curability in terms of “die” and “live”. Two variables are dependent to each other. These two factors are not at all independent to each other. In other words, spleen palpable varies according to the change of living status (“Class”).

Conclusion:

It is very challenging for diagnosing “Hepatosplenomegaly” which is the condition when spleen swell beyond their normal size due to viral hepatitis (Beasley, 1988). From the analysis the conclusion that could be drawn is- the presence palpation of Spleen generally increases the risk of life. The longevity of life gets more vulnerable at the time of hepatitis when Spleen palpability increases. Such type of disorder could be manifested by severe hepatitis. However, if it is not fixed, the patient suffering from hepatitis may be significantly affected. Enlarged and palpable spleen is a serious issue that needs appropriate medical treatment. Otherwise, hepatitis would directly damage health condition and as a result patient may receive a death. Patients should focus on palpation of their spleen to prevail the risk of Hepatitis.

References:

Beasley, R. P. (1988). Hepatitis B virus. The major etiology of hepatocellular carcinoma. Cancer, 61(10), 1942-1956.

Gerspacher-Lara, R., Pinto-Silva, R. A., Serufo, J. C., Rayes, A. A., Drummond, S. C., & Lambertucci, J. R. (1998). Splenic palpation for the evaluation of morbidity due to schistosomiasis mansoni. Memórias do Instituto Oswaldo Cruz, 93, 245-248.

Lok, A. S., & McMahon, B. J. (2009). Chronic hepatitis B: update 2009. Hepatology, 50(3), 661-662.

Sharpe, D. (2015). Your chi-square test is statistically significant: Now what?. Practical Assessment, Research & Evaluation, 20.

Smith, E. B., & Custer, R. P. (1946). Rupture of the spleen in infectious mononucleosis: a clinicopathologic report of seven cases. Blood, 1(4), 317-333.