University of Southern California Dissertations and Theses

Validation of the Children’s International Mucositis Evaluation Scale (ChIMES) in pediatric cancer and SCT

(USC Thesis Other)

Validation of the Children’s International Mucositis Evaluation Scale (ChIMES) in pediatric cancer and SCT

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Validation of the Children’s International Mucositis Evaluation Scale (ChIMES) in
Pediatric Cancer and SCT

By Xichen Jin

A Thesis Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the Requirements for the Degree
MASTER OF SCIENCE
(APPLIED BIOSTATISTICS AND EPIDEMIOLOGY)

May 2019

2
TABLE OF CONTENTS
ACKNOWLEDGEMENT ...............................................................................................................3
ABSTRACT .....................................................................................................................................4
INTRODUCTION ...........................................................................................................................5
DATA AND METHODS ................................................................................................................6
DATA ..........................................................................................................................................6
METHODS .................................................................................................................................7
RESULTS ......................................................................................................................................10
DISCUSSION ................................................................................................................................16
REFERENCES ..............................................................................................................................18
APPENDIX I .................................................................................................................................20
APPENDIX II ................................................................................................................................25

3
Acknowledgement

I would like to thank my thesis mentor Dr. Ha Dang for her guidance and suggestions. Thank
you for giving me this opportunity to join this wonderful project, and always being patient and
supportive to me.

I also want to appreciate Dr. Todd Alonzo and Dr. Meredith Franklin for being my committee
members and giving me useful advise to improve my work.

4

Abstract

Our research data is from ACCL1031, a multicenter randomized double blinded placebo-
controlled clinical trial at Children’s Oncology Group (COG), that evaluated the effectiveness of
oral mucositis prevention of Caphosol. The study data contain information of patients from 4 to
21 years old who underwent myeloablative hematopoietic stem cell transplantation (HSCT). This
thesis project aims to validate the internal consistency and construct validity (via convergent
validity) of proxy-reported Children’s International Mucositis Evaluation Scale (ChIMES)
compared to the World Health Organization (WHO) Oral Toxicity Scale, 10-point mouth pain
categorical rating scale, and the Oral Mucositis Daily Questionnaire (OMDQ). The data of all
four rating scales were collected by trained healthcare provider starting one day before surgery
for up to 22 consecutive days. Cronbach’s alpha was calculated for ChIMES rating scale for all
patients by day.

The results based on study ACCL1031 showed a good-to-excellent internal consistency and
moderate-to-good convergent validity for healthcare provider proxy-reported ChIMES in
children with cancer and undergoing HSCT, and there are significant associations between
ChIMES and WHO scale, OMDQ, and 10-point mouth pain categorical scale.
5
Introduction

Oral mucositis is defined as inflammation and ulceration of the mouth mucosa with
pseudomembrane (Scardina et al, 2010). It is one of the most common adverse reaction of
chemotherapy and hematopoietic stem cell transplantation (HSCT). Depending on the regimen of
therapy and characteristics of patients, the range of prevalence could be 10% to 100% (Raber-
Durlacher et al, 2010). The symptom of oral mucositis is erythema followed by white
desquamating plaques, and it is painful even with a gentle touch and reduces the patient’s quality
of life. Dependable mucositis measures are needed for the identification of useful therapeutic
plans. As a result, a pediatric-specific measure for oral mucositis called the Children’s
International Mucositis Evaluation Scale (ChIMES) was developed for self-reporting by children
from 8 to 18 years old and proxy-reported by parents for children under 12 years old. ChIMES
contains 7 questions that ask about pain level, effects on mouth activities, and oral and throat
medication. These questions together evaluate the severity of oral mucositis. On the basis of a
previous study (Jacobs et al, 2013), the reliability and validity of self-reported and proxy-
reported ChIMES have been confirmed.

Aside for ChIMES, there are other measures for the severity of oral mucositis such as World
Health Organization (WHO) Mucositis Scale, the Oral Mucositis Daily Questionnaire (OMDQ),
and 10-point mouth pain categorical rating scale. ACCL1031 described the psychometric
properties of healthcare provider proxy-reported ChIMES, using WHO scale, OMDQ, and 10-
point mouth pain categorical rating scale as criterion.
6
Data and Methods

Data
The original data includes baseline characteristics of the study cohort and outcome variables.
Since Caphosol treatment and Palifemin may influence the condition of oral mucositis according
to previous study, patients who received Caphosol treatment and Palifemin 30 days before
enrolling into ACCL1031 (ClinicalTials.gov identifier: NCT01305200) were excluded (Treister
et al, 2016).

The baseline characteristics of the ACCL1031 study cohort are age, gender, method of payment,
diagnosis type (malignant or non-malignant), transplant type (autologous or allgeneic), graft
source (bone marrow, peripheral blood stem cells, or umbilical cord blood), conditioning (Total
Body Irradiation (TBI), busulphan, melphalan, or fludarabine), characteristics in allogeneic
group, Graft-versus-host Disease (GVHD) prophylaxis with methotrexate (yes or no), days of
engraftment, opioids dose and narcotic usage (yes or no).

Outcome variables include ChIMES total and percentage scores, World Health Organization
(WHO) Mucositis Scale, Oral Mucositis Daily Questionnaire (OMDQ), and 10-point mouth pain
categorical rating scale. Outcome variables were assessed during the transplant and recovery
phase (Day -1 to Day +20). If a patient was discharged before Day +20, assessments stoppped on
day of discharge.

7
Proxy-reported ChIMES score was the instrument we want to validate, other instruments were
used to construct validation. ChIMES includes 7 questions to assess mouth pain. The 1
st
to 4
th

questions have score ranging from 0 to 5, and the 5
th
to 7
th
questions have score ranges 0 to 1.
The ChIMES Total Score ranging from 0 to 23 if all questions are answered. Mucositis is worse
with higher ChIMES score. When calculating the ChIMES Total Score, missing value and
response of “I can’t tell” were scored as 0. ChIMES Percentage Score was computed as ChIMES
Total Score divided by the total maximum score, where the elements with “I can’t tell” response
were omitted from the total maximum score.

The WHO scale ranges from 0 to 4 and higher score means worse mucositis (WHO Handbook
for reporting results of cancer treatment, 1979).

The OMDQ includes 7 questions, but the 7
th
question was excluded from ACCL1031 study since
this question measures amount of diarrhea and performed poorly with negligible correlation with
other questions based on previous study (Manji et al, 2012). The OMDQ ranges from 0-24 and
higher score indicates worse mucositis. Six questions of OMDQ were considered individually for
all analyses.

The mouth pain categorical rating scale ranges from 0 to 10, and higher score indicates worse
pain.

Methods

8
All patients who had outcome measures for ChIMES total and percentage score, WHO scale,
OMDQ, and 10-point mouth pain categorical rating scale were considered for the analyses in this
thesis.

We used Cronbach’s alpha which was developed by Lee Cronbach in 1951 (Cronbach, 1951)
and corresponding 95% confidence interval to evaluate the internal consistency of the ChIMES
rating scale. Cronbach’s alpha, which ranges between 0 and 1, provides a measure of the internal
consistency (reliability) of a questionnaire or scale. Internal consistency describes the extent to
which all the items in a rating scale measure the same concept or construct (Tavakol & Dennick,
2011). Reliability estimates show the amount of measurement error in a rating scale, which
means this estimate is the correlation of rating scale with itself. The 95% confidence interval for
Cronbach’s alpha was provided using the exact method by Koning & Frances (Kromrey et al,
2010). Cronbach’s alpha larger than 0.7 suggests acceptable internal consistency (Streiner &
Norman, 2008). Higher Cronbach’s alpha denotes stronger internal consistency. Additional
details about Cronbach’s alpha and its interpretation are provided in Appendix II section 1.
Based on previous studies, the Cronbach’s alpha was 0.95 for parent proxy-report, 0.93 for
children aged 12 to 18 years self-report, and 0.95 for children aged 8 to <12 years self-report
ChIMES (Jacobs et al, 2013). Since our study concerns healthcare provider proxy-report
ChIMES, the result might be reported less precise than parent proxy-report and children self-
report ChIMES. As such, for this study we slightly decrease the expectation of Cronbach’s alpha
to alpha > 0.8 (Streiner & Norman, 2008).

9
We used Spearman’s correlation coefficient, which is a non-parametric analog of Pearson’s
correlation coefficient, i.e. it makes no assumption regarding the distribution of variables, to
evaluate the construct validity via convergent validity of ChIMES with WHO, mouth pain scale,
and 6 questions of OMDQ. Spearman’s correlation coefficient captures monotonic
(increasing/decreasing) relationship between two continuous or ordinal variables (not necessarily
linear). The range of Spearman’s correlation coefficient is from -1 to 1. When there is no
repeated value in the data, -1 and 1 denote a perfect Spearman’s correlation when each of the
variables is a perfect monotone function of the respective one. Zero denotes a total lack of
association. Additional details about Pearson’s correlation and its interpretation are provided in
Appendix II section 2. Based on previous studies, all Spearman’s correlation coefficients are
larger than 0.5 across respondent types (parent proxy-report, children aged 12 to 18 years self-
report, and children aged 8 to <12 years self-report ChIMES) (Jacobs et al, 2013). Since our
study was about healthcare provider proxy-report ChIMES, the result might be reported less
precise than parent proxy-report and children self-report ChIMES. So, we slightly decrease the
expectation of Spearman’s correlation coefficient. For this thesis project, we anticipated at least
fair positive correlation, r>0.25 (Colton, 1974) between ChIMES and each of the 3 rating scales.

To take into account the fact that the same patient provided multiple measures, generalized linear
mixed models were used to obtain p-values for Spearman’s correlation coefficient estimates
(Roy, 2006). For the purpose of this analysis, WHO, mouth pain scale, and OMDQ were
considered as the dependent variables and the ChIMES total and percentage scores were the
independent variables. We specified the distribution as multinomial to acquire p-values.
Significance level was set at 0.05. All statistical analyses were conducted using SAS 9.4.
10
Results

There were 220 eligible patients enrolled in this study. Description of basic characteristics of all
patients are shown in Table 1. Most of patients were between age 11 to 14 (63 patients, 28.6%)
and 15 to 18 (58 patients, 26.4%), the other 3 age groups have similar percentage of total eligible
patients, around 15%. About 56% of the study participants were male. Nearly half of patients
used private insurance (45.5%). Patients with malignant conditions (90.0%) is far more prevalent
than patients with non-malignant conditions (10.0%). Seventy four percent of patients
transplanted by allogeneic HSCT with and without TBI or melphalan, while 25.9% of patients
transplanted by autologous HSCT with and without TBI or melphalan. Nearly 90.0% of patients
received their graft from bone marrow and PBSC. There are 141 patients with TBI and/or
melphalan (64.1%), and 79 patients without TBI and melphalan (35.9%). Seventy percent of the
patients received HSCT from a fully matched unrelated marrow or peripheral blood stem cell
donor and HLA matched family members. For graft-versus-host disease (GVHD) prophylaxis,
96 patients (66.7%) received methotrexate, while 48 patients (33.3%) did not. There were 157
patients with reported days of engraftment, the median (range: min, max) day of engraftment was
13 (8, 34). The usage of narcotic and opioids was known for 210 participants. Among 176
patients who reported opioids dose, the median (range: min, max) dose of opioids used in M
equivalent per kg per day was 5.27 (0.04, 731.04). The interquartile range (IQR) and
corresponding frequencies of cumulative dose of opioids were shown in Appendix I, A1.

Table 1: Baseline Characteristics of the Study Cohort (N=220)

Characteristics n (%)
Age in Years
11
4-7 38 (17.3%)
8-10 29 (13.2%)
11-14 63 (28.6%)
15-18 58 (26.4%)
19-21 32 (14.5%)
Gender
Male 118 (53.6%)
Female 102 (46.4%)
Method of Payment
Private insurance 101 (45.5%)
Medicaid 80 (36.0%)
Medicaid and medicare 1 (0.5%)
Medicare 7 (3.1%)
Medicare and private insurance 5 (2.3%)
Self-pay (no insurance) 1 (0.5%)
Military sponsored 1 (0.5%)
Other 15 (6.7%)
No means of payment 11 (4.9%)
Diagnosis
Malignant conditions 198 (90.0%)
Non-malignant conditions 22 (10.0%)
Narcotic Use
Yes 189 (85.9%)
No 21 (9.5%)
Missing 10 (4.6%)
Median Opioids dose (range) (N=176)
5.27 (0.04,
731.04)
Transplant Type
Allogeneic HSCT with TBI or melphalan 99 (45.0%)
Autologous HSCT with TBI or melphalan 42 (19.1%)
Allogeneic HSCT without TBI or melphalan 64 (29.1%)
Autologous HSCT without TBI or melphalan 15 (6.8%)
Graft Source
Bone marrow 105 (47.7%)
Bone marrow; cord blood 1 (0.5%)
Cord blood 23 (10.5%)
Peripheral blood stem cells (PBSC)
91 (41.3%)
Conditioning
With TBI and/or Melphalan 141 (64.1%)
Without TBI/Melphalan 79 (35.9%)
Characteristics in Allogeneic Group
Fully matched unrelated marrow or peripheral blood stem cell donor 61 (27.7%)
HLA matched family member 53 (24.1%)
12
HLA matched or partially mismatched cord blood (class I or II) 23 (10.5%)
HLA matched sibling or parents 4 (1.8%)
Partially matched family donor (class I) 1 (0.5%)
Partially mismatched unrelated marrow or peripheral blood stem cell
donor
21 (9.5%)
Missing 57 (25.9%)
GVHD prophylaxis
Methotrexate 96 (43.6%)
No methotrexate 48 (21.8%)
Missing 76 (34.6%)
Median Day of Engraftment (range) (N=147)
13 (8, 34)

*Abbreviations: TBI = total body irradiation; GVHD = graft-versus-host disease.

Median and interquartile range (IQR) of baseline and maximum scores were reported to describe
the characteristics of each mucositis rating scale. The baseline score was taken from the first day
of evaluation of each patient, and the maximum score was determined as the largest score of
each patient across all returned rating scale measures. For ChIMES, medians of the ChIMES
total score and the ChIMES percentage score were both calculated. Median was calculated
separately for each element of OMDQ (from OMDQ1 to OMDQ6). Medians and corresponding
IQRs were also computed for the total scores of WHO and 10-point Mouth Pain Categorical
Rating Scale. Each patient may have up to 22 days of mucositis measures. The results are shown
in Table 2.

Table 2. Characteristics of Outcomes
Outcomes N (baseline) Baseline
N
(maximum)
Maximum
Median ChIMES Total
Score (IQR)
193 0 (0, 1) 198 19 (12, 21)
Median ChIMES
Percentage Score (IQR)
193
0.0% (0.0%,
4.3%)
198
82.6% (52.2%,
91.3%)
13

Internal Consistency
For the following analysis, we refer to day of mucositis measure as iteration. There are 22
iterations in ChIMES, WHO, OMDQ, and 10-point mouth pain categorical scale. Cronbach’s
alpha and corresponding 95% confidence interval of raw alpha were obtained to assess the
internal consistency of ChIMES rating scale for each iteration. The alphas were calculated using
7 elements of ChIMES rating scale. All raw Cronbach’s alphas were above 0.88, and there are
minimal differences between the raw and their corresponding standardized alphas (all above
0.90) because the ranges of 7 elements in ChIMES do not vary significantly from each other. The
overall raw alpha was 0.91 with corresponding standardized alpha equal to 0.94, which exceeded
our expectation of Cronbach’s alpha at least 0.8.
Median World Health
Organization Mucositis
Scale (IQR)
198 0 (0, 0) 198 3 (2, 3)
Median Oral Mucositis
Daily Questionnaire 1
(IQR)
198 0 (0, 0) 198 3 (2, 4)
Median Oral Mucositis
Daily Questionnaire 2
(IQR)
198 0 (0, 0) 198 2 (0, 3)
Median Oral Mucositis
Daily Questionnaire 3
(IQR)
198 0 (0, 0) 198 3 (2, 3)
Median Oral Mucositis
Daily Questionnaire 4
(IQR)
198 0 (0, 0) 198 3 (2, 4)
Median Oral Mucositis
Daily Questionnaire 5
(IQR)
198 0 (0, 0) 198 4 (2, 4)
Median Oral Mucositis
Daily Questionnaire 6
(IQR)
198 0 (0, 0) 198 2 (1, 3)
Median 10-Point Mouth
Pain Categorical Rating
Scale (IQR)
198 0 (0, 0) 198 6 (4, 8)
14

Table 3. Cronbach’s Alpha for internal consistency of ChIMES

Property N ChIMES Total Score
Reliability Cronbach’s alpha
Internal
consistency
Raw (95%CI*) Standardized
Iteration 1 193 0.886 (0.860, 0.909) 0.900
Iteration 2 194 0.904 (0.882, 0.923) 0.920
Iteration 3 194 0.905 (0.883, 0.924) 0.921
Iteration 4 191 0.905 (0.884, 0.925) 0.922
Iteration 5 192 0.901 (0.879, 0.921) 0.927
Iteration 6 193 0.905 (0.883, 0.924) 0.928
Iteration 7 192 0.909 (0.888, 0.927) 0.942
Iteration 8 187 0.900 (0.877, 0.920) 0.933
Iteration 9 189 0.898 (0.874, 0.918) 0.928
Iteration 10 189 0.895 (0.871, 0.916) 0.928
Iteration 11 190 0.889 (0.863, 0.911) 0.914
Iteration 12 189 0.892 (0.867, 0.914) 0.926
Iteration 13 187 0.904 (0.881, 0.924) 0.941
Iteration 14 185 0.899 (0.875, 0.920) 0.928
Iteration 15 182 0.896 (0.871, 0.918) 0.923
Iteration 16 173 0.890 (0.863, 0.914) 0.915
Iteration 17 169 0.909 (0.886, 0.929) 0.934
Iteration 18 158 0.893 (0.865, 0.916) 0.916
Iteration 19 150 0.895 (0.867, 0.919) 0.925
Iteration 20 142 0.893 (0.864, 0.918) 0.921
Iteration 21 131 0.902 (0.874, 0.925) 0.929
Iteration 22 123 0.884 (0.850, 0.913) 0.920
Overall 3893 0.906 (0.901, 0.910) 0.935

*95% confidence intervals derived from Koning & Frances Exact method.

Construct Validity
OMDQ, WHO, and 10-point mouth pain categorical rating scale were used as criterion to
evaluate the construct validity via convergent validity of ChIMES by Spearman’s rank-order
correlation. ChIMES total and percentage score were both used to acquire the correlations with
each of the three criterion scales, and the number of patients available for these analyses is 193.
15
There are 22 iterations in each pair of correlation. All Spearman’s correlation coefficients
showed positive correlation between total and percentage score ChIMES and 6 elements of
OMDQ, WHO, and 10-point mouth pain categorical rating scale (p<0.0001). The range of
overall Spearman’s correlation estimates are between 0.57 and 0.85 (Table 4), which exceeded
our expectation of at least 0.25. Spearman’s correlations were computed by iteration were
summarized in Table 2 in Appendix I. All p-values are <0.0001, which means there are
statistically significant association between ChIMES total and percentage score and OMDQ,
WHO, and 10-point mouth pain categorical rating scale (Table 4).

Table 4. Overall Spearman’s correlation and p-value from linear mixed model

Property
Spearman’s Correlation (P-value*)
Pain
Scale
OMDQ
1
OMDQ
2
OMDQ
3
OMDQ
4
OMDQ
5
OMDQ
6
WHO
ChIMES
Total
Score
0.795
(<0.000
1)
0.842
(<0.000
1)
0.577
(<0.000
1)
0.817
(<0.000
1)
0.807
(<0.000
1)
0.817
(<0.000
1)
0.686
(<0.000
1)
0.727
(<0.000
1)
ChIMES
Percentag
e Score
0.794
(<0.000
1)
0.843
(<0.000
1)
0.578
(<0.000
1)
0.818
(<0.000
1)
0.809
(<0.000
1)
0.819
(<0.000
1)
0.688
(<0.000
1)
0.725
(<0.000
1)

*P-values derived from generalized linear mixed model
16
Discussion

In previous study, the reliability, validity, and possibility of response of a pediatric-specific
measure of oral mucositis called Children’s International Mucositis Evaluation Scale (ChIMES)
has been verified in children and adolescents with cancer and experiencing HSCT (Jacobs et al,
2013) in both child self-report and parent proxy-report. In this thesis project, we investigated the
internal consistency and construct validity of ChIMES in measuring mucositis in Pediatric
Cancer and SCT patients aged 4 to 21 years from ACCL1031.

Cronbach’s alphas for internal consistency of ChIMES suggested a good to excellent internal
consistency in ChIMES and were higher than anticipated (alpha>0.8). Convergent validity as
accessed by the results from Spearman’s correlation coefficients showed moderate to good
positive correlations between ChIMES and OMDQ, WHO, and 10-point mouth pain categorical
rating scale. Furthermore, ChIMES is significantly associating with OMDQ, WHO, and 10-point
mouth pain categorical rating scale at α=0.05 level.

Spearman’s correlation method gives rough estimate of correlation between 2 measures;
however, this analysis ignores the correlation structure of the repeated measures that is present in
the data. Having said that, estimating correlation through a mixed model required additional
assumptions, particularly about correlation structures which is beyond the scope of this thesis
project. It would be of interest to address the limitations identified in a future project or studies to
confirm the results obtained in this thesis.

17
In summary, based on ACCL1031 data, the results from this thesis project suggest that there is a
good internal consistency and construct validity of proxy-reported Children’s International
Mucositis Evaluation Scale (ChIMES) assessed by healthcare provider in children with cancer
and undergoing myeloablative hematopoietic stem cell transplantation (HSCT). We conclude
that the ChIMES scale can be used with confidence to report the severity of oral mucositis.

18
References

1. Cronbach, L. J. (1951). Coefficient alpha and the internal structure of
tests. Psychometrika,16(3), 297-334. doi:10.1007/bf02310555
2. Colton, T. (1974). Statistics in medicine (1st ed.). Boston: Little, Brown.
3. D Kromrey, Jeffrey., Romano, Jeanine., & Hibbard, Susan. (2010). ALPHA_CI: A SAS
® Macro for Computing Confidence Intervals for Coefficient Alpha
4. Introduction to linear mixed models. (n.d.). Retrieved from
https://stats.idre.ucla.edu/other/mult-pkg/introduction-to-linear-mixed-models/
5. Manji, A., Tomlinson, D., Ethier, M., Gassas, A., Maloney, A., & Sung, L. (2012).
Psychometric properties of the Oral Mucositis Daily Questionnaire for child self-report
and importance of mucositis in children treated with chemotherapy. Supportive Care in
Cancer, 20(6), 1251–1258. https://doi.org/10.1007/s00520-011-1211-z
6. Nathaniel Treister, Michael Nieder, Christina Baggott, Ellen Olson, Lu Chen, Ha
Dang, … Lillian Sung. (2016). Caphosol for prevention of oral mucositis in pediatric
myeloablative haematopoietic cell transplantation. British Journal of Cancer, 116(1), 21–
27. https://doi.org/10.1038/bjc.2016.380
7. Roy, A. (2006). Estimating Correlation Coefficient between Two Variables with
Repeated Observations using Mixed Effects Model. Biometrical Journal, 48(2), 286–301.
https://doi.org/10.1002/bimj.200510192
8. Raber-Durlacher, J., Elad, S., & Barasch, A. (2010). Oral mucositis. Oral Oncology,
46(6), 452–456. https://doi.org/10.1016/j.oraloncology.2010.03.012
19
9. Scardina, G., Pisano, T., Messina, P. (2010). Oral mucositis. Review of literature. The
New York State Dental Journal, 76(1), 34–38. Retrieved from
http://search.proquest.com/docview/733847480/
10. Streiner, D., & Norman, G. (2008). Health Measurement Scales A practical guide to their
development and use (4th ed.). Oxford: Oxford University Press.
11. Swinscow, T., & Campbell, M. (1997). Statistics at square one (9th ed. / rev. by M.J.
Campbell.). London: BMJ Pub. Group.
12. S Jacobs, C Baggott, R Agarwal, T Hesser, T Schechter, P Judd, … L Sung. (2013).
Validation of the Children’s International Mucositis Evaluation Scale (ChIMES) in
paediatric cancer and SCT. British Journal of Cancer, 109(10), 2515–25122.
https://doi.org/10.1038/bjc.2013.618
13. Tavakol, M., & Dennick, R. (2011). Making sense of Cronbach’s alpha. International
Journal of Medical Education, 2, 53–55. Retrieved from
http://search.proquest.com/docview/898889039/
14. WHO (World Health Organization) Handbook for reporting results of cancer treatment.
(1979). Geneva: WHO.

20
Appendix I

A1. Opioids Dose and Narcotic Use

Property Cumulative dose of opioids
Quartile 0-0.37 0.37-2.37 2.37-13.41 >13.41
N (%) 52 (24.8%) 53 (25.2%) 52 (24.8%) 53 (25.2%)

A2. Spearman’s Correlation by Iteration for Convergent Validity of ChIMES

Property N ChIMES Total Score ChIMES Percentage Score
Correlation
with Mouth
Pain Scale
Spearman’s r Spearman’s r
Iteration 1 193 0.693 0.692
Iteration 2 194 0.724 0.723
Iteration 3 194 0.778 0.778
Iteration 4 191 0.768 0.768
Iteration 5 192 0.841 0.840
Iteration 6 193 0.784 0.784
Iteration 7 192 0.810 0.811
Iteration 8 187 0.825 0.824
Iteration 9 189 0.783 0.781
Iteration 10 189 0.691 0.679
Iteration 11 190 0.750 0.746
Iteration 12 189 0.707 0.709
Iteration 13 187 0.744 0.741
Iteration 14 185 0.813 0.805
Iteration 15 182 0.801 0.798
Iteration 16 173 0.807 0.805
Iteration 17 167 0.714 0.711
Iteration 18 158 0.695 0.698
Iteration 19 149 0.686 0.687
Iteration 20 141 0.734 0.733
Iteration 21 131 0.614 0.614
Iteration 22 122 0.531 0.527

Correlation
with OMDQ1
Spearman’s r Spearman’s r
Iteration 1 193 0.804 0.804
21
Iteration 2 194 0.771 0.770
Iteration 3 194 0.791 0.791
Iteration 4 191 0.807 0.807
Iteration 5 192 0.851 0.850
Iteration 6 193 0.840 0.840
Iteration 7 192 0.842 0.844
Iteration 8 187 0.870 0.868
Iteration 9 189 0.832 0.836
Iteration 10 189 0.790 0.796
Iteration 11 190 0.804 0.816
Iteration 12 189 0.800 0.806
Iteration 13 187 0.816 0.819
Iteration 14 185 0.864 0.867
Iteration 15 182 0.838 0.836
Iteration 16 173 0.810 0.811
Iteration 17 167 0.775 0.779
Iteration 18 158 0.745 0.750
Iteration 19 149 0.738 0.739
Iteration 20 141 0.756 0.755
Iteration 21 131 0.638 0.652
Iteration 22 122 0.588 0.584

Correlation
with OMDQ2
Spearman’s r Spearman’s r
Iteration 1 193 0.489 0.488
Iteration 2 194 0.492 0.492
Iteration 3 194 0.520 0.521
Iteration 4 191 0.468 0.467
Iteration 5 192 0.573 0.572
Iteration 6 193 0.573 0.576
Iteration 7 192 0.548 0.545
Iteration 8 187 0.616 0.614
Iteration 9 189 0.648 0.651
Iteration 10 189 0.609 0.613
Iteration 11 190 0.634 0.648
Iteration 12 189 0.648 0.658
Iteration 13 187 0.490 0.491
Iteration 14 185 0.580 0.579
Iteration 15 182 0.562 0.566
Iteration 16 173 0.548 0.541
Iteration 17 167 0.455 0.454
Iteration 18 158 0.514 0.512
Iteration 19 149 0.518 0.517
Iteration 20 141 0.361 0.362
22
Iteration 21 131 0.368 0.367
Iteration 22 122 0.306 0.303

Correlation
with OMDQ3
Spearman’s r Spearman’s r
Iteration 1 193 0.658 0.659
Iteration 2 194 0.723 0.722
Iteration 3 194 0.788 0.788
Iteration 4 191 0.763 0.763
Iteration 5 192 0.847 0.848
Iteration 6 193 0.806 0.804
Iteration 7 192 0.813 0.816
Iteration 8 187 0.834 0.834
Iteration 9 189 0.833 0.838
Iteration 10 189 0.773 0.776
Iteration 11 190 0.779 0.792
Iteration 12 189 0.806 0.810
Iteration 13 187 0.794 0.797
Iteration 14 185 0.819 0.825
Iteration 15 182 0.829 0.828
Iteration 16 173 0.799 0.795
Iteration 17 167 0.772 0.774
Iteration 18 158 0.737 0.736
Iteration 19 149 0.682 0.685
Iteration 20 141 0.660 0.660
Iteration 21 131 0.611 0.608
Iteration 22 122 0.529 0.525

Correlation
with OMDQ4
Spearman’s r Spearman’s r
Iteration 1 193 0.625 0.626
Iteration 2 194 0.687 0.686
Iteration 3 194 0.756 0.756
Iteration 4 191 0.744 0.744
Iteration 5 192 0.816 0.817
Iteration 6 193 0.814 0.813
Iteration 7 192 0.796 0.799
Iteration 8 187 0.823 0.822
Iteration 9 189 0.827 0.839
Iteration 10 189 0.786 0.792
Iteration 11 190 0.799 0.809
Iteration 12 189 0.784 0.790
Iteration 13 187 0.792 0.795
Iteration 14 185 0.822 0.828
23
Iteration 15 182 0.822 0.822
Iteration 16 173 0.794 0.790
Iteration 17 167 0.720 0.728
Iteration 18 158 0.720 0.727
Iteration 19 149 0.733 0.737
Iteration 20 141 0.680 0.682
Iteration 21 131 0.637 0.638
Iteration 22 122 0.486 0.482

Correlation
with OMDQ5
Spearman’s r Spearman’s r
Iteration 1 193 0.690 0.690
Iteration 2 194 0.685 0.684
Iteration 3 194 0.751 0.751
Iteration 4 191 0.740 0.738
Iteration 5 192 0.856 0.856
Iteration 6 193 0.824 0.821
Iteration 7 192 0.797 0.801
Iteration 8 187 0.850 0.859
Iteration 9 189 0.802 0.810
Iteration 10 189 0.786 0.795
Iteration 11 190 0.738 0.753
Iteration 12 189 0.773 0.778
Iteration 13 187 0.786 0.789
Iteration 14 185 0.828 0.831
Iteration 15 182 0.819 0.825
Iteration 16 173 0.807 0.805
Iteration 17 167 0.756 0.763
Iteration 18 158 0.750 0.757
Iteration 19 149 0.754 0.755
Iteration 20 141 0.727 0.730
Iteration 21 131 0.617 0.626
Iteration 22 122 0.559 0.555

Correlation
with OMDQ6
Spearman’s r Spearman’s r
Iteration 1 193 0.484 0.483
Iteration 2 194 0.512 0.512
Iteration 3 194 0.598 0.599
Iteration 4 191 0.549 0.548
Iteration 5 192 0.648 0.648
Iteration 6 193 0.694 0.697
Iteration 7 192 0.729 0.726
Iteration 8 187 0.745 0.745
24
Iteration 9 189 0.754 0.763
Iteration 10 189 0.717 0.727
Iteration 11 190 0.648 0.661
Iteration 12 189 0.707 0.697
Iteration 13 187 0.706 0.707
Iteration 14 185 0.738 0.744
Iteration 15 182 0.691 0.695
Iteration 16 173 0.645 0.646
Iteration 17 167 0.609 0.613
Iteration 18 158 0.583 0.585
Iteration 19 149 0.559 0.560
Iteration 20 141 0.500 0.505
Iteration 21 131 0.478 0.477
Iteration 22 122 0.290 0.287

Correlation
with WHO
Spearman’s r Spearman’s r
Iteration 1 193 0.569 0.568
Iteration 2 194 0.616 0.616
Iteration 3 194 0.654 0.652
Iteration 4 191 0.664 0.662
Iteration 5 192 0.747 0.746
Iteration 6 193 0.738 0.732
Iteration 7 192 0.640 0.643
Iteration 8 187 0.712 0.710
Iteration 9 189 0.597 0.602
Iteration 10 189 0.657 0.643
Iteration 11 190 0.685 0.681
Iteration 12 189 0.638 0.639
Iteration 13 187 0.678 0.676
Iteration 14 185 0.658 0.651
Iteration 15 182 0.695 0.696
Iteration 16 173 0.700 0.699
Iteration 17 167 0.696 0.695
Iteration 18 158 0.611 0.611
Iteration 19 149 0.759 0.757
Iteration 20 141 0.744 0.749
Iteration 21 131 0.656 0.652
Iteration 22 122 0.511 0.506

25

Appendix II

1. Cronbach’s alpha:

The formula of Cronbach’s alpha is:
α=
𝑁∙𝑐̅ 𝑣̅ +(𝑁−1)∙𝑐̅
N=the number of items in a rating scale
𝑐̅=average covariance between item-pairs
𝑣̅=average variance

The rule of thumb for interpreting Cronbach’s alpha is (Streiner & Norman, 2008):
Cronbach’s alpha Internal consistency
a³0.9 Excellent
0.9>a³0.8 Good
0.8>a³0.7 Acceptable
0.7>a³0.6 Questionable
0.6>a³0.5 Poor
0.5>a Unacceptable

26
The alpha increases when the items in a rating scale are correlated to each other. The value of
alpha is also related to the length of the rating scale. To increase the alpha, we should add more
related items to test the same concept on the rating scale.

The SAS syntax PROC CORR statement with ALPHA option was used to compute the
Cronbach’s alpha.

2. Spearman’s Correlation Coefficient:

The general procedure for Spearman’s correlation is:

• Rank X from smallest to largest (Rx), assign average rank for ties
• Rank Y from smallest to largest (Ry), assign average rank for ties
• Calculate d=𝑅
/
−𝑅
0
for each subject
• Then the Spearman’s correlation is given by 𝛾
2
=𝜌
4
5
, 4
8
=
9:;(4
5
, 4
8
)
<
=
5
<
=
8
, where r denotes
the usual Pearson correlation, but applied to the rank variables; c𝑜𝑣(𝑅
/
, 𝑅
0
) is the
covariance of the rank variables; 𝜎
4
5
and 𝜎
4
8
are the standard deviations of the rank
variables.
• Only when all n ranks are distinct integers, the Spearman’s correlation can be given by
𝛾
2
=1−
A∑C
D
E
F(F
E
GH)
, where 𝑑
J
=R(𝑋
J
)−R(𝑌
J
) is the difference between the two ranks of
each observation; n is the number of observations.
27
• Test statistic: t=
O
P
√FGR
SHGO
P
E
, df=n−2.

Spearman’s correlation coefficient is practicable in both continuous and ordinal data, and it is
robust to outliers. The correlation will be high when two variables have similar rank, and it will
be low when the rank are dissimilar for two variables.

The rule of thumb for interpreting Spearman’s correlation coefficient is (Colton, 1974):

Spearman’s’s
Correlation coefficient
Correlation
r³0.75 Very good-to-excellent correlation
0.75>r³0.5 Moderate-to-good correlation
0.5>r³0.25 Fair correlation
0.25>r Negligible or not correlated

The PROC CORR statement with SPEARMAN’S option was used to obtain this correlation
coefficient in SAS.

Abstract (if available)

Abstract Our research data is from ACCL1031, a multicenter randomized double-blinded placebo-controlled clinical trial at Children’s Oncology Group (COG), that evaluated the effectiveness of oral mucositis prevention of Caphosol. The study data contain information of patients from 4 to 21 years old who underwent myeloablative hematopoietic stem cell transplantation (HSCT). This thesis project aims to validate the internal consistency and construct validity (via convergent validity) of proxy-reported Children’s International Mucositis Evaluation Scale (ChIMES) compared to the World Health Organization (WHO) Oral Toxicity Scale, 10-point mouth pain categorical rating scale, and the Oral Mucositis Daily Questionnaire (OMDQ). The data of all four rating scales were collected by trained healthcare provider starting one day before surgery for up to 22 consecutive days. Cronbach’s alpha was calculated for ChIMES rating scale for all patients by day. ❧ The results based on study ACCL1031 showed a good-to-excellent internal consistency and moderate-to-good convergent validity for healthcare provider proxy-reported ChIMES in children with cancer and undergoing HSCT, and there are significant associations between ChIMES and WHO scale, OMDQ, and 10-point mouth pain categorical scale.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Predicting ototoxicity evaluated by SIOP in children receiving cisplatin

PDF

A novel risk-based treatment strategy evaluated in pediatric head and neck non-rhabdomyosarcoma soft tissue sarcomas (NRSTS) patients: a survival analysis from the Children's Oncology Group study...

PDF

Instability of heart rate and rating of perceived exertion during high-intensity interval training in breast cancer patients undergoing anthracycline chemotherapy

PDF

Predicting hospital length of stay (LOS) using the National Inpatient Sample

PDF

Spatial modeling of non-tailpipe emissions and its association with children's lung function

PDF

An analysis of disease-free survival and overall survival in inflammatory breast cancer

PDF

Development and validation of survey instrument designed for cervical cancer screening in Malawi, and other low resource settings

PDF

Effect of biomass fuel exposure on infant respiratory health outcomes in Bangladesh

PDF

Risk factors and survival outcome in childhood alveolar soft part sarcoma among patients in the Children’s Oncology Group (COG) Phase 3 study ARST0332

PDF

Cross-sectional association of blood pressure, antihypertensive medications, MRI volumetric measures and cognitive function scores in an aging population

PDF

Evaluating the use of friend or family controls in epidemiologic case-control studies

PDF

Diet quality and pancreatic cancer incidence in the multiethnic cohort

PDF

A comparison of three different sources of data in assessing the adolescent and young adults cancer survivors

PDF

Long-term effects on vertebral body height growth of dose sculpting intensity modulated radiation therapy for children with neuroblastoma

PDF

Examining exposure to extreme heat and air pollution and its effects on all-cause, cardiovascular, and respiratory mortality in California: effect modification by the social deprivation index

PDF

Effects of post-menopausal hormone therapy on arterial stiffness in the ELITE trial

PDF

Evaluation of factors influencing Los Angeles Tiered-Dispatch System’s improvement on bystander CPR rate and inter reliability between electronic patient care report (ePCR) and 911 call review on...

PDF

Comparison of participant and study partner predictions of cognitive impairment in the Alzheimer's disease neuroimaging initiative 3 study

PDF

The evaluation of the long-term effectiveness of zero/low fluoroscopy workflow in ablation procedures for the treatment of paroxysmal and persistent atrial fibrillation

Asset Metadata

Creator Jin, Xichen (author)

Core Title Validation of the Children’s International Mucositis Evaluation Scale (ChIMES) in pediatric cancer and SCT

School Keck School of Medicine

Degree Master of Science

Degree Program Applied Biostatistics and Epidemiology

Publication Date 04/24/2019

Defense Date 04/24/2019

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag Children's International Mucositis Evaluation Scale,internal consistency,OAI-PMH Harvest,oral mucositis,pediatric

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Dang, Ha (committee chair), Alonzo, Todd (committee member), Franklin, Meredith (committee member)

Creator Email jxc0122@outlook.com,xichenji@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-142098

Unique identifier UC11675577

Identifier etd-JinXichen-7232.pdf (filename),usctheses-c89-142098 (legacy record id)

Legacy Identifier etd-JinXichen-7232.pdf

Dmrecord 142098

Document Type Thesis

Format application/pdf (imt)

Rights Jin, Xichen

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA