Description and Analysis some adj

1 | P a g e
Erin Dinkel
ECO4935
November 6, 2016
Description and Analysis Paper
I INTRODUCTION
Healthcare has been a debated topic for some time, especially when it comes to the health
of children. Programs have been put in place in an attempt to provide children with healthcare,
such as the Children’s Health Insurance Program (CHIP), Medicaid, Obamacare, or private health
insurance.1 The assumption is that families that cannot afford health insurance for their children
would not be able to afford any surgeries as well, thus showing a positive relationship. This study
will attempt to find a correlation between the number of pediatric surgeries performed and the
percentage of children who are insured by Medicaid in comparison to those insured with private
insurance. The hypothesis is that in pediatric surgical areas, with a higher percentage of children
insured by Medicaid, there will be more pediatric surgeries performed. The study will only be
conducted through data collected from northern New England, thus the information cannot be
generalizable for all of the country. However, it will give a glimpse into relationships between
children’s healthcare and other variables.
II SAMPLE
The sample was taken from a study of pediatric surgical areas in the northern New England
region, by The Dartmouth Atlas of Health Care.2 The sample consists of thirty-one pediatric
surgical areas in the northern New England Region, including Maine, New Hampshire, and
1 KidsHealth from Nemours- How to Find Affordable Healthcare
2 The Dartmouth Atlas of Children’s Health Care in Northern New England

2 | P a g e
Vermont. The sample contains information on geography, demographics, physician workforce,
and surgeries. Additionally, there is data on whether or not Medicaid or private insurance was used
for surgery in each pediatric surgery area, as well as, Census data on the number of people with a
bachelor’s degree or higher3 and the percent of the black population.4
III DEPENDENT VARIABLE
Total Amount of Tympanostomy Surgeries
The dependent variable is the amount of tympanostomy surgeries that are done, per 1,000
children, in each of the hospitals studied. This surgery is the placement of a tube in the tympanic
membrane to equalize the pressure. This is done to treat ear fluid that builds up in the inner ear,
which can lead to “otitis media” as well as loss of hearing.5 This surgery is being considered as a
variable in the study because otitis media is the most common diagnosis among children and is the
second most common diagnosis in medicine in general.6 For this reason a proportionately high or
low number of these surgeries compared to the population could potentially show a correlation
between the two variables.
Typanostomy surgeries can become a necessity for children who experience frequent ear
infections. Ear infections can be either viral or bacterial. The infection causes swelling in the
middle ear, preventing air from the throat to reach the ear. This blockage can create a vacuum
which brings in fluid from the nose into the middle ear area, and is unable to be drained from the
swollen tube.7 Treatment for ear infections can be antibiotics, if it is bacterial, or simply waiting
3 U. S. Census Bureau, Bachelor's degree or higher, percent of persons age 25 years+, 2010-2014
4 U.S. Census Bureau, 2010 Census of Population Black or African American alone, percent,
April 1, 2010
5 Medscape- Ear Tube Insertion
6 A Report of the Dartmouth Atlas Project- Typanostomy Tube Placement
7 Ear Infections Cause- WebMD

3 | P a g e
until it passes, if it is viral. If frequent ear infections continue, tympanostomy surgery is necessary.
The tube placed in the ear during the surgery can be temporary, in which the tube will fall out
naturally within 6 months, or long-term, in which the tubes should be removed after a longer
period.8
IV INDEPENDENT VARIABLES
Percent of Children in Poverty (POV)
One independent variable is of the percentage of children living in poverty in the specified
pediatric surgical area. This variable will be an indicator if the children in this neighborhood are
insured with healthcare or not. Studying this variable in comparison to the number in surgeries
will give some indication if there is a relationship between the two. The Affordable Care Act
implemented an expansion of Medicaid in order to extend coverage to low-income Americans. In
a 2012 study, it was found that 63% of the people at the federal poverty line (FPL) were eligible,
and in eighteen states the thresholds for Medicaid were at or above the FPL.9
Additionally, poverty can have an impact on the need for typanostomy surgery. Ear
infections can be caused by bacteria or a virus. Children living in poverty may be exposed to less
sanitary conditions, which could introduce them to more bacteria for potential ear infections.
Moreover, bacterial ear infections can only be treated with antibiotics. Meaning, children in
poverty may not have the means to afford antibiotics and will not be properly treated and therefore
be prone to more frequent ear infections.
8 Ear Tubes- American Academy of Otolaryngology
9 Medicaid Expansions from 1997 to 2009 Increased Coverage and Improved Access and Mental
Health Outcomes for Low-Income Parents.

4 | P a g e
Type of Insurance (INS)
The type of insurance used will be another independent variable. It is intuitive that this
variable be measured in conjunction with the number of surgeries done to determine if there is a
strong correlation between the two. For children who are hospitalized, Medicaid is the largest
payer, and it can account for 40% of hospital discharges on a national scale.10 This will be
measured using a dummy variable. It is important to note that there is some data missing as to the
exact number of surgeries that use Medicaid or private insurance, however this is a small amount
of surgeries. If Medicaid is used in the pediatric surgery area, then it was assigned a value of 1. If
private insurance was used in the pediatric surgery area, then a value of 0 was assigned.
Physician Workforce (PHY)
The physician’s work force in each area will be measured as an independent variable as
well. The data measures this variable as the number of Otolaryngologists (ear, nose, and throat
physicians) per 100,000 children and adults. This is important to consider because it could be a
factor in why there are a lot or little surgeries. For the purposes of this study we will assume that
the number of doctors in an area will have some affect as to the number of surgeries that are
performed as well.
Percent of Black Population (BLK)
This variable is being used to control for demographics differences. This variable will show
if the demographics have a strong correlation with the amount of surgeries. The hypothesis is that
it will not be significant as the amount of surgeries will be largely correlated with income and type
of insurance, rather than demographics.
10 Medicaid, Hospital financial stress, and the Incidence of Adverse Medical Events for Children

5 | P a g e
Percent of Bachelor’s Degree or Higher (BA)
Similarly, this variable of percent of Bachelor’s Degrees of higher will be used to control
for demographic differences. The hypothesis for this variable is that people who receive a
Bachelor’s Degree or higher tend to earn higher wages, and can therefore afford other forms of
health insurance other than Medicaid, typically private insurance.
V RESULTS
With such a small sample size with the number of variables I wish to study, I intend to run
multiple regressions in order to test how certain variables will affect the results. For the first
regression I will include all of the variables and use the following model.
Tympanostomies=0+1(POV)+2(INS)+3(PHY)+4(BLK)+5(EDU)+E
This regression model resulted in the following relationship between the number of
tympanostomy surgeries per 1,000 and the variables.
Tympanostomies=10.056-.253(POV)+3.389(INS)+.467(PHY)-.063(BLK)-
.060(EDU)+E
For this regression, the adjusted R-Squared value demonstrates that 31% of the
tympanostomy surgeries per 1,000 are congruent with this model. Consequently, the presence of
Medicaid is expected to increase the amount of tympanostomy surgeries by 3.389 per 1,000 as
compared to private insurance.
In this study the null hypothesis is that 2=0, meaning that the presence of Medicaid as
compared with private insurance has no effect on the number of pediatric tympanostomy surgeries
performed. In this regression, the p-value for Insurance was .0002. This value is less than .05,
meaning I can reject the null hypothesis at the 95% confidence level. This value is also less than
.01, meaning that I can also reject the null hypothesis at the 99% confidence level.

6 | P a g e
Variable Coefficients P-value
Intercept 10.05602108 0.000752641
Insurance 3.389285714 0.00015289
Otolaryngologists 0.467175077 0.045102065
Black -0.063452297 0.233765409
% withBA -0.060317864 0.305334699
% childreninpoverty -0.253043096 0.030178483
As for the other variables, the number or pediatric surgeons available present a p-value of
.045, which is less than .05, meaning that I can reject the null hypothesis at the 95% confidence
level. However, .045 >.01, meaning I cannot reject the null hypothesis at the 99% level as with the
Insurance variable. As expected, more surgeons are associated with more surgeries, since more
doctors available can perform more surgeries. For percent black, .234 >.05, so I cannot reject the
null hypothesis. This is not surprising, as the hypothesis presumes that the surgeries would have a
higher correlation with income and insurance rather than demographics. For percent with a
bachelor’s degree, .305 >.05, so I cannot reject the null hypothesis. This is expected because, the
hypothesis is that people with bachelor’s degrees earn a higher income and can therefore afford
private insurance and not Medicaid. Finally, for percent of children in poverty .03 < .05, so I can
reject the null hypothesis with 95% confidence. This is also expected having a higher percent of
children in poverty is associated with a lower percent of surgeries. This is expected if children who
are poor have less access to medicine and surgeries.
This model suggests that, with the significant variables (Insurance and Poverty), the
amount of tympanostomy surgeries would increase by 3.4 per 1,000 for every percentage increase
for those with Medicaid compared to those with private insurance, and decrease by .25 per 1,000
for every percentage increase in poverty. Both are at least significant at the 95% confidence level;
the Insurance variable is also significant at the 99% confidence level.

7 | P a g e
A note should be made about this regression that there is a possibility of multicollinearity
between the Insurance variable and the percent of children in poverty. It is known that many of the
families that receive Medicaid, do so because they live at or below the poverty line. This can also
affect the amount of pediatric surgeries that occur as well. For this reason, I tested the correlation
coefficient of all the variables.
Insurance
% children in
poverty Black % with BA Otolaryngologists
Insurance 1
% childrenin
poverty -2.40692E-17 1
Black -6.66912E-18 0.050726965 1
% withBA 0 -0.580147263 0.236587343 1
Otolaryngologists 7.35437E-18 -0.282722946 0.136771418 0.535800828 1
The above chart demonstrates that variables are not correlated because their correlation
coefficient’s absolute value is less than 0.8. For this reason, I have decided not to run separate
regression with each of these variables. At first glance it may seem odd that there is such an
extremely small correlation coefficient between the insurance and children in poverty. This is
because the insurance variable is a dummy. This means that the variable is the presence of
Medicaid compared to private insurance, which is why the correlation coefficient seems much
lower than expected.
The next regression will omit the percent of the black population variable and will use the
following model.
Tympanostomies=0+1(POV)+2(INS)+3(PHY)+5(EDU)+E
Following this regression model resulted in the following relationship between the number
of tympanostomy surgeries per 1,000 and the variables, omitting the percent of the black
population.

8 | P a g e
Tympanostomies=10.648-.277(POV)+3.397(INS)+.471(PHY)-.077(EDU)+E
For this regressions the adjusted R-Squared value demonstrates that 31% of the tympanostomy
surgeries per 1,000 are congruent with this model. Consequently, the presence of Insurance is
expected to increase the amount of tympanostomy surgeries by 3.397 per 1,000.
Similarly, as before, the null hypothesis is that 2=0. In this regression, the p-value for
Insurance was .00009. This value is less than .05, meaning I can reject the null hypothesis at the
95% confidence level. This value is also less than .01, meaning that I can also reject the null
hypothesis at the 99% confidence level. As expected this is due to the fact that the presence of
Medicaid compared to private insurance makes it more affordable to get surgery for families who
are of lower income.
Intercept 10.6475147 9.39713E-05
Insurance 3.396551724 9.05792E-05
% withBA -0.076738905 0.156567176
.040, which is less than .05, meaning that I can reject the null hypothesis at the 95% confidence
level. However, .04 >.01, meaning I cannot reject the null hypothesis at the 99% level as with the
Insurance variable. As expected, more surgeons are associated with more surgeries, since more
doctors available can perform more surgeries.
For percent with a bachelor’s degree, .157 >.05, so I cannot reject the null hypothesis. This
is expected because, the hypothesis is that people with bachelor’s degrees earn a higher income
and can therefore afford private insurance and not Medicaid. Finally, for percent of children in

9 | P a g e
poverty .011 < .05, so I can reject the null hypothesis with 95% confidence. This is also expected
having a higher percent of children in poverty is associated with a lower percent of surgeries. This
is expected if children who are poor have less access to medicine and surgeries.
This model suggests that, with the significant variables (Insurance, Poverty, and Surgeons),
the amount of tympanostomy surgeries would increase by 3.4 per 1,000 for every percentage
increase for those with Medicaid compared to those with private insurance, decrease by .27 per
1,000 for every percentage increase in poverty, and increase by .47 per 1,000 for every percentage
increase in pediatric surgeons. Each of these are at least significant at the 95% confidence level,
the Insurance variable is also significant at the 99% confidence level.
I will then run the regression including the black population variable and excluding the
percent of bachelor degrees or higher, and will use the following model.
Typanostomies=0+1(POV)+2(INS)+3(PHY)+4(BLK)+E
of tympanostomy surgeries per 1,000 and the variables, omitting the percent of people with
bachelor degrees or higher.
Tympanostomies=7.762-.187(POV)+3.389(INS)+.360(PHY)-.079(BLK)+E
For this regressions the adjusted R-Squared value demonstrates that 31% of the
Medicaid as compared to private insurance is expected to increase the amount of tympanostomy
surgeries by 3.389 per 1,000.
The null hypothesis is that 2=0. In this regression, the p-value for Insurance was .0002.
This value is less than .05, meaning I can reject the null hypothesis at the 95% confidence level.

10 | P a g e
This value is also less than .01, meaning that I can also reject the null hypothesis at the 99%
confidence level. This is expected for the same reasons stated earlier.
Intercept 7.761585728 3.64449E-05
Insurance 3.389285714 0.000151109
Black -0.079269423 0.122069654
.081, which is greater than .05, meaning that I cannot reject the null hypothesis. This is different
from the other regressions which showed it as significant, and is not expected. This may be because
the bachelor’s degree variable has been taken out and therefore doesn’t account for the surgeons
who do have a bachelor’s degree. For the percent of black population variable, the p-value is .122,
which is greater than .05, meaning I cannot reject the null hypothesis. This is not surprising, as the
hypothesis presumes that the surgeries would have a higher correlation with income and insurance
rather than demographics.
Finally, for percent of children in poverty .051 > .05, so I cannot reject the null hypothesis
with 95% confidence, however I can reject it with 90% confidence. This is expected for the same
intuition that was stated earlier. Only Insurance was significant in this regression at the 95%,
although the percent of children in poverty was also significant with 90% confidence.
This model suggests that, with the significant variable (Insurance), the amount of
tympanostomy surgeries would increase by 3.4 per 1,000 for every percentage increase for those
with Medicaid compared to those with private insurance. This is the only variable that was
significant with 95% confidence.

11 | P a g e
Finally, I will run the regression using neither the percent black population variable nor the
percent of bachelor’s degrees of higher variable, using the following model.
Tympanostomies=0+1(POV)+2(INS)+3(PHY)+E
of tympanostomy surgeries per 1,000 and the variables, omitting the percent of people with
bachelor degrees or higher and the percent of the black population.
Tympanostomies=7.897-.202(POV)+3.397(INS)+.308(PHY)+E
For this regressions the adjusted R-Squared value demonstrates that 30% of the
Medicaid is expected to increase the amount of tympanostomy surgeries by 3.397 per 1,000
compared to private insurance.
The null hypothesis is that 2=0. In this regression, the p-value for Insurance was .0001.
This value is less than .05, meaning I can reject the null hypothesis at the 95% confidence level.
This value is also less than .01, meaning that I can also reject the null hypothesis at the 99%
confidence level. Again, this result is expected as in the other regressions.
Variables Coefficients P-value
Intercept 7.896801126 1.41572E-05
Insurance 3.396551724 0.000101701
% childrenin
poverty -0.201692627 0.032828091
.120, which is greater than .05, meaning that I cannot reject the null hypothesis. This is not an
expected result and could be due to the limited number of variables in this regression. For the
percent of children in poverty .033 < .05, so I can reject the null hypothesis with 95% confidence.

12 | P a g e
This is expected as in the other regressions. So, Insurance and the percent of children in poverty
were significant in this regression.
This model suggests that, with the significant variable (Insurance and Poverty), the amount
of tympanostomy surgeries would increase by 3.4 per 1,000 for every percentage increase for those
with Medicaid compared to those with private insurance and increase by 0.2 per 1,000 for every
percentage decrease in poverty. Both variables were significant with 95% confidence, and
Insurance was also significant with 99% confidence.
With each of the variations of regressions, it was found to be consistent that the Insurance
variable was significant and we can reject the null hypothesis with 99% confidence. This supports
my hypothesis that the presence of Medicaid is highly correlated with the number of pediatric
surgeries as compared to the presence of private insurance. Presumably this is due to the fact that
the presence of Medicaid makes it more affordable to get surgery for families who are of lower
income. Although it is a possibility that there is some multicollinearity between the presence of
Medicaid and the percent of children in poverty, through the correlation coefficient it was shown
that it was not correlated enough to be of concern for this study. However, it is also important to
recognize that this study was conducted with a small sample, meaning the results are not
completely generalizable. Nonetheless the results still demonstrate a strong correlation, at the 99%
confidence level, between Medicaid as compared with private insurance and pediatric surgery, as
well as a correlation with the percent of children living in poverty, which rejected the null
hypothesis at the 95% confidence level.

13 | P a g e
Descriptive Statistics
Typanostomy Insurance Otolaryngologists
Mean 8.151785714 Mean 0.5 Mean 3.835714286
StandardError 0.498868792 StandardError 0.067419986 StandardError 0.290794127
Median 7.7 Median 0.5 Median 3.55
Mode 7.5 Mode 0 Mode 3.9
StandardDeviation 3.733192203 StandardDeviation 0.504524979 StandardDeviation 2.176103989
Sample Variance 13.93672403 Sample Variance 0.254545455 Sample Variance 4.735428571
Kurtosis 4.953731858 Kurtosis -2.075471698 Kurtosis 7.07892278
Skewness 1.575803011 Skewness 2.93069E-17 Skewness 2.125859571
Range 21.7 Range 1 Range 11.7
Minimum 2.5 Minimum 0 Minimum 0.7
Maximum 24.2 Maximum 1 Maximum 12.4
Sum 456.5 Sum 28 Sum 214.8
Count 56 Count 56 Count 56
Black % with BA % children in poverty
Mean 3.196428571 Mean 31.48928571 Mean 12.99636786
StandardError 1.122355374 StandardError 1.396662717 StandardError 0.621965694
Median 1.1 Median 29.55 Median 12.83515
Mode 0.6 Mode 45.6 Mode 13.1
StandardDeviation 8.398938554 StandardDeviation 10.45166674 StandardDeviation 4.654365066
Sample Variance 70.54216883 Sample Variance 109.2373377 Sample Variance 21.66311417
Kurtosis 23.19114773 Kurtosis -0.946948239 Kurtosis -0.630265979
Skewness 4.847194847 Skewness 0.05316005 Skewness 0.187193226
Range 45.2 Range 38.3 Range 17.0501
Minimum 0.4 Minimum 10.8 Minimum 4.9499
Maximum 45.6 Maximum 49.1 Maximum 22
Sum 179 Sum 1763.4 Sum 727.7966
Count 56 Count 56 Count 56

14 | P a g e
Regression for All Variables:
Typanostomies=0+1(POV)+2(INS)+3(PHY)+4(BLK)+5(EDU)+E
Regression Statistics
Multiple R 0.612549269
R Square 0.375216606
Adjusted RSquare 0.312738267
StandardError 3.094863019
Observations 56
ANOVA
df SS MS F Significance F
Regression 5 287.6109662 57.52219323 6.005547047 0.000199823
Residual 50 478.9088553 9.578177105
Total 55 766.5198214
Variables Coefficients Standard Error t Stat P-value Lower 95% Upper 95%
Intercept 10.05602108 2.801156852 3.589952869 0.000752641 4.429731957 15.6823102
Insurance 3.389285714 0.827136934 4.097611381 0.00015289 1.727932289 5.05063914
Otolaryngologists 0.467175077 0.227313689 2.05519993 0.045102065 0.010602096 0.923748059
Black -0.063452297 0.052644906 -1.205288458 0.233765409 -0.169192702 0.042288108
% withBA -0.060317864 0.058239873 -1.035679865 0.305334699 -0.177296091 0.056660364
% childreninpoverty -0.253043096 0.113408641 -2.231250576 0.030178483 -0.480831056 -0.025255136

15 | P a g e
Regression without black population
Typanostomies=0+1(POV)+2(INS)+3(PHY)+5(EDU)+E
R Square 0.361326302
AdjustedRSquare 0.313124514
Observations 58
ANOVA
Regression 4 279.3070384 69.82675961 7.496118169 7.31443E-05
Residual 53 493.6979616 9.315055878
Total 57 773.005
Variables Coefficients StandardError t Stat P-value Lower95% Upper 95%
Intercept 10.6475147 2.519157917 4.226616613 9.39713E-05 5.594723798 15.7003056
Insurance 3.396551724 0.801509605 4.237693102 9.05792E-05 1.788927044 5.004176404
Otolaryngologists 0.470723546 0.223728626 2.10399337 0.040140048 0.02198075 0.919466341
% withBA -0.076738905 0.053398436 -1.437100251 0.156567176 -0.183842604 0.030364793

16 | P a g e
Regression without bachelor’s degrees
Typanostomies=0+1(POV)+2(INS)+3(PHY)+4(BLK)+E
R Square 0.361813343
Observations 56
ANOVA
Regression 4 277.3370994 69.33427485 7.228481011 0.000108613
Residual 51 489.182722 9.591818079
Total 55 766.5198214
Intercept 7.761585728 1.715432758 4.524564249 3.64449E-05 4.317710764 11.20546069
Insurance 3.389285714 0.827725717 4.094696642 0.000151109 1.727556998 5.05101443
Otolaryngologists 0.360139158 0.202605495 1.777538946 0.081442565 -0.046608346 0.766886662
Black -0.079269423 0.050416867 -1.572279842 0.122069654 -0.180485506 0.02194666
% childreninpoverty -0.187164196 0.093957053 -1.992018575 0.051735945 -0.375790851 0.00146246

17 | P a g e
Regression without black population or bachelor’s degrees
Typanostomies=0+1(POV)+2(INS)+3(PHY)+E
R Square 0.336439031
Observations 58
ANOVA
Regression 3 260.0690529 86.68968429 9.126369438 5.53228E-05
Residual 54 512.9359471 9.498813836
Total 57 773.005
Intercept 7.896801126 1.653904991 4.774640121 1.41572E-05 4.580921265 11.21268099
Insurance 3.396551724 0.80937668 4.196503075 0.000101701 1.773849183 5.019254265
Otolaryngologists 0.308004382 0.194855237 1.580683112 0.119790791 -0.082656846 0.698665611

18 | P a g e
References
AmosWEB is Economics: Encyclonomic WEB*pedia. (n.d.). Retrieved October 11, 2016, from
http://www.amosweb.com/cgi-bin/awb_nav.pl?s=wpd
B. A. (n.d.). Adenoidectomy (Adenoid Removal). Retrieved October 10, 2016, from
http://my.clevelandclinic.org/health/treatments_and_procedures/hic-adenoidectomy-adenoid-
removal
Reilly, B. K. (2016, February 5). Ear Tube Insertion. Retrieved October 10, 2016, from
http://emedicine.medscape.com/article/1890757-overview
Ear Tubes. (2016). Retrieved December 07, 2016, from http://www.entnet.org/content/ear-tubes
Goodman, D. C., Morden, N. E., Ralston, S. L.(2013). Retrieved October 10, 2016, from
http://www.dartmouthatlas.org/downloads/atlases/NNE_Pediatric_Atlas_121113.pdf
How to Find Affordable Health Care. (n.d.). Retrieved October 11, 2016, from
http://kidshealth.org/en/parents/find-care.html
Kaiser Comission on Medicaid Facts. (May 2011). Retrieved November 6, 2016 from
https://kaiserfamilyfoundation.files.wordpress.com/2013/01/8188.pdf
McMorrow, S., Kenney, G., Long, S. (Aug 2016). Retrieved October 11, 2016, from
http://go.galegroup.com/ps/retrieve.do?tabID=T002&resultListType=RESULT_LIST&searchRe
sultsType=SingleTab&searchType=AdvancedSearchForm&currentPosition=1&docId=GALE%
7CA460185818&docType=Article&sort=RELEVANCE&contentSegment=&prodId=HRCA&c
ontentSet=GALE%7CA460185818&searchId=R1&userGroupName=gain40375&inPS=true
Phillips, M. (October, 2016) Personal Interview.
Smith, R. B., Gynan, L., Fairbrother, G. (2012). Retreived October 10, 2016, from
http://go.galegroup.com/ps/retrieve.do?tabID=T002&resultListType=RESULT_LIST&searchRe
sultsType=SingleTab&searchType=AdvancedSearchForm&currentPosition=27&docId=GALE
%7CA298292928&docType=Article&sort=RELEVANCE&contentSegment=&prodId=HRCA&
contentSet=GALE%7CA298292928&searchId=R3&userGroupName=gain40375&inPS=true
Staff, B. M. (2015, July 17). Tonsillectomy. Retrieved October 10, 2016, from
http://www.mayoclinic.org/tests-procedures/tonsillectomy/basics/definition/prc-20019889

Description and Analysis some adj

More Related Content

What's hot (18)

Viewers also liked (19)

Similar to Description and Analysis some adj (10)

Description and Analysis some adj