Deakin University MIS7771: Analyzing UnitedHealth Claims Data

This report presents a statistical analysis of UnitedHealth Group's medical malpractice claims data. The analysis addresses specific questions regarding claim payment amounts, claimant demographics (age, insurance status), and claim severity. Key findings include the average claim payment amount, confidence intervals for claimant age and insurance coverage, and validation of industry trends related to claim severity. The report also examines the impact of private attorney representation on claim amounts and investigates potential differences in claim characteristics for orthopedic surgeons compared to other specialists. The analysis leverages descriptive statistics, confidence interval estimation, and hypothesis testing to provide insights into the claims data, concluding that certain assertions regarding claim severity and specialist types are not supported by the data. Desklib provides similar solved assignments and past papers for students.

DEAKIN UNIVERSITY
DATA ANALYTICS AND VISUALISATION
MIS7771
STUDENT ID
[Pick the date]

Introduction
The cost of medical malpractice in the US is very high and exceeds $ 50 billion. Due to this,
there is a upward pressure on the health care costs coupled with a higher premium for
insurance related to medical malpractice. One of the known health insurance provider in US
is the UnitedHealth Group which aims for a better understanding of the claims raised on
account of medical malpractice. In order to achieve this objective, claims data for 200
random claimants has been provided which comprises of information related to the claim
amount, severity, gender, hiring of private attorney or not, age, marital status
presence/absence of insurance coupled with speciality. The given report aims to present the
findings of the statistical analysis of the data provided in wake of the specific questions that
my manager Edmond Kendrick has raised in the email. Through the requisite statistical
analysis, the report would aim to answer the queries raised in relation to the claims data.
Data Analysis
The various issues as highlighted in the manager’s mail are addressed below.
1) An overall summary of the claim payment amount has been indicated in Appendix 1. The
average or mean claim payment amount has come out to be $ 73,457.49. Also, it is
estimated that 50% of provided sample claim amount is lesser than or equal to $72,571.38.
Further, the claim value of $ 5,400 seems to have the highest frequency in the sample data.
The dispersion of the claim amount values in the sample data seem to be quite high
considering that the lowest value of claim amount is $ 1,547 while the highest value of
claim amount is $ 228,724. 80. Also, the standard deviation of the claim amount from the
mean or average claim value stands at $ 32,178.50. Additionally, it is noticeable that the
distribution of the claim amount is not symmetric considering the fact that there are certain
claim amounts which are unusually large and hence considered outliers in the provided
sample data.
2) A) With a likelihood of 95%, it can be stated that the average age of the claimants for the
population would lie between 42 years and 47 years. The requisite computation for the
same is highlighted in Appendix 2A.
1

B) With a likelihood of 95%, it can be stated that the mean proportion of claimants with
“No Insurance” for the population would lie between 5.03% and 12.97%. The requisite
computation for the same is highlighted in Appendix 2B.
3) A) Based on the given claim sample data, it would be appropriate to conclude with 95%
confidence that the average claim amount paid by the industry has dropped below $
77,500. The requisite computation for the same is highlighted in Appendix 3A.
B) Using the claim severity data presented, it can be stated with 95% likelihood that the
study indicating that 75% of the patients fall either in “MILD” or “MEDIUM” severity
condition continues to be valid for the current year as well. The computation for the same
is illustrated in Appendix 3B.
C) Considering the gender trends in the provided claims data concerning the severity of
the claims, it can be inferred that there no significant difference in proportion of male and
female patients in the category of "MILD” or “MEDIUM” claims. The relevant
computation for the same is illustrated in Appendix 3C.
D) Also, the given sample data does suggest that the payment amount does tend to depend
on the fact whether the claimant is represented by a private attorney or not. This is
apparent from the fact that the average claim amount of claimants represented by private
attorneys tends to be higher than the corresponding claim amount of claimants not
represented by private attorneys. The relevant computation for the same is illustrated in
Appendix 3D.
E) The given data provided on claims does not lend support to the assertion that private
attorneys tend to have higher representation for “SEVERE” claims as compared to
“MEDIUM” claims. Hence, the statement is not valid. The relevant calculations to support
the above conclusion are illustrated in Appendix 3E.
4) A) The given claims data does not lend support to the assertion that “SEVERE” claims
tend to be higher for Orthopedic surgeon in comparison to other specialists. Infact, the
results derived in Appendix 4A tend to highlight that the difference between the
2

“SEVERE” claims proportion for the two does not show any significant difference and
hence can be assumed to be same.
B) The given claims data does not lend support to the assertion that average claim amount
for “SEVERE” claims tend to be higher for Orthopedic surgeon in comparison to other
specialists. Infact, the results derived in Appendix 4B tend to highlight that the difference
between the “SEVERE” claims average amount for the two does not show any significant
difference and hence can be assumed to be same.
Conclusion
Based on the above analysis, useful conclusions can be drawn about the claims data. It may
be concluded that the average age of the claimants tends to lie between 42 and 47 years. Also,
most of these claimants tend to have insurance since only a very small proportion (about 5-
12%) does not have insurance. The average claim amount has now dropped below $ 77,500.
Further, it can also be concluded that 75% of the claims belong to the “MILD” or
“MEDIUM” category and hence only 25% of the claims fall in the “SEVERE” category.
Also, there are no gender specific differences between the proportions of “MILD” or
“MEDIUM” category claims. Besides, it may be also concluded that the average claim
amount tends to be higher when a private attorney represents the claimant. However, no
significant difference is observed between the representation proportion of “MEDIUM” and
“SEVERE” claims with regards to private attorney. Also, the assertions regards higher
proportion and average claim amount for orthopaedic related “SEVERE” claims in
comparison with other specialists has been found incorrect as no evidence is present for the
same.
3

Appendices
APPENDIX 1
APPENDIX 2A
4

APPENDIX 2B
APPENDIX 3A
5

APPENDIX 3B
6

APPENDIX 3C
APPENDIX 3D
7

APPENDIX 3E
8

APPENDIX 4A
APPENDIX 4B
9