11 September 2008
Item reduction and psychometric validation of the Oily Skin Self
Assessment Scale (OSSAS) and the Oily Skin Impact Scale (OSIS).
Supplementary file
APPENDIX 1. ADOLESCENT FOCUS GROUPS
The initial development of the oily skin instruments was based on qualitative data obtained
from focus groups conducted with adults with oily skin. However, following these focus
groups it was decided to conduct further focus groups with adolescents with oily skin, in order
to ensure that all concepts of importance to adolescents are included in the Oily Skin Self
Assessment Scale (OSSAS), and to assess the face and content validity of the OSSAS in the
adolescent population.
OBJECTIVES
The objective of the adolescent focus groups was to evaluate the face and content validity of
the OSSAS in the adolescent population (ages 12-17). This involved asking questions to
evaluate whether all concepts of importance to adolescents are included in the OSSAS, and
to assess whether the instructions, items and response options were understood and
interpreted correctly by the adolescents with oily skin. Also it was important to identify any
OSSAS items that are inappropriate or unsuitable for adolescent patients.
METHODOLOGY
Four focus groups were conducted each including 7-10 adolescents with oily skin, who had a
desire to improve their oily skin using treatment. As oily skin is personal and subjective,
diagnosis of oily skin was determined by the subjects themselves. The focus groups occurred
in four US cities (Dallas, Philadelphia, Miami and Los Angeles) to ensure there was no
geographical bias in the data obtained from the focus groups.
Subjects were included in the study if they suffered from oily skin (in their own opinion), if they
wanted their skin to be less oily and were interested in using a treatment that may improve
oily skin. All subjects were between 12-17 years old, spoke English as their primary language
and were willing and able to participate in a two hour group discussion about how oily skin
affects their life. All subjects understood and were willing to sign an informed consent form
for the study (a parent also provided informed consent) prior to their involvement in focus
groups, in a manner that followed the US Health Insurance Portability and Accountability Act
(HIPAA) guidelines. Since this research was qualitative and observational (not involving any
study interventions or medications), and given that medical professionals were not involved in
PI5542B OS Validation Manuscript Supplementary file v2.0
Page 1 of 8
11 September 2008
recruiting subjects, approval of the study from an institutional review board was not sought.
All subjects were compensated for their participation.
Subjects were excluded from the study if they had a life-threatening health condition, had
another more severe skin condition such as psoriasis in addition to oily skin, had an
uncontrolled psychiatric condition or was unable to participate in a two hour discussion on the
impact of oily skin on their life.
In two of the focus groups patients had acne in addition to their oily skin, thus allow an
evaluation of whether there were any differences in the concepts of importance to the two
groups of patients. The final composition of the focus groups was as follows:

One focus group in males who were experiencing oily skin with no additional acne

One focus group in males who were experiencing acne in addition to their oily skin

One focus group in females who were experiencing oily skin with no additional acne

One focus group in females who were experiencing acne, in addition to their oily skin.
Focus group procedure
Each of the four focus groups were conducted by the same moderate who used a structured
discussion guide to enable the discussion whilst ensuring conformity and consistency across
the groups. The discussion guide for those groups that included patients with acne had
additional questions pertaining to pimples. Each focus group lasted approximately 2 hours.
During each focus group, participants were asked to discuss their experience of having oily
facial skin, how they determined the severity of their oily skin, the impact oily skin has had on
their lives and the emotional impact of having oily skin. They also completed the pre-item
reduction OSSAS developed based on adult qualitative research and provided feedback on
the questions, instructions format and response options.
RESULTS
A total of 34 subjects participated in the focus groups with seven to ten subjects in each
group. Of these subjects 53% (n=18) had acne in conjunction with their oily skin and 47%
(n=16) subjects had oily skin only. Fifty six percent of the subjects were female (n=19) and
44% (n=15) were male. The majority of the sample was Caucasian (58%), 21% were African
American, 18% were Hispanic and 3% were mixed African/Caucasian. Thirty-five percent of
the sample perceived themselves to have mild oily skin, 44% moderate oily skin and 21% of
the subjects perceived themselves to have severe oily skin. Of the subjects who had acne in
conjunction with their oily skin 50% (n=9) perceived themselves to have mild acne, 33% (n=6)
moderate acne and 17% (n=3) of subjects perceived themselves to have severe acne.
PI5542B OS Validation Manuscript Supplementary file v2.0
Page 2 of 8
11 September 2008
When asked to describe their oily skin, the most commonly used words were ‘Shiny’ (used by
12 participants), ‘Greasy’ (n=7), ‘Annoying’ (n=6), ‘Ugly’ (n=6) and ‘Irritating’ (n=5). Oily skin
was described as feeling like a ‘film’ or ‘extra layer’ and the look of oily skin was characterized
as ‘bothersome as it stands out to other people’.
skin made them feel ‘Very self-conscious’.
The adolescents described that their oily
Findings were very consistent with focus groups
carried out previously with adult subjects; however the emotional impact of oily skin due to
‘teasing’ from peers is likely to be more profound in adolescent subjects than in adults.
Findings were also consistent between the groups conducted with subjects with acne, and
those without acne.
When commenting on their review of the OSSAS, the adolescents generally felt the
questionnaire was ‘through’ and ‘pretty easy’ to complete and covered most of the aspects
that were important to assessing oily skin.
The adolescents did however criticize the
questionnaire as being repetitive and commented that some of the instructions or questions
were difficult to answer or understand. One suggestion made by subjects in two of the focus
groups was a recommendation to change the Likert response scale to a 0-10 numeric rating
scale to make the questionnaire easier to complete. It was commented that the questionnaire
did not include questions asking about the impact of emotions, social self-consciousness,
problems with shaving (males), and factors that contribute to the oiliness of their skin (e.g.
diet, stress, exercise, hot climate/weather). It should be noted however that the adolescent
subjects were not shown the OSIS, an instrument that would assess the emotional impacts of
oily skin.
Based on the four focus groups, the OSSAS was shown to have content validity as a
measure of severity of oily skin in that it captured important aspects of how adolescents
assess the oiliness of their skin. Input and suggestions from the adolescents focus groups
were considered when reducing and modifying the items contained in the OSSAS and OSIS,
and some modifications were made on the basis of their comments.
PI5542B OS Validation Manuscript Supplementary file v2.0
Page 3 of 8
11 September 2008
APPENDIX 2. PATIENT ACNE CHARACTERISTICS
For those patients with acne in addition to oily skin a number of acne characteristics were
captured and analyzed. First, as part of the patient demographic form patients rated the
severity of their acne by answering the question ‘Currently is your acne:’ with the responses
of ‘mild’, ‘moderate’, ‘severe’ and ‘extreme’. Table 1 displays the results of the acne severity
rating for the 135 patients who rated their facial acne, demonstrating a range of self rated
acne severity.
Table 1. Patient Rating of Acne Severity
Acne severity n (%)
Mild
Moderate
Severe
Extreme
Missing from 196 total subjects
135 (100)
65 (48)
55 (41)
15 (11)
0 (0)
61
The recruiting clinician at each site also performed a facial acne lesion count on all patients.
The mean number of total acne lesions and the inflammatory and non-inflammatory lesions is
provided in Table 2. The results show that the mean number of total acne lesions was 9.25,
with approximately half of each of this number being inflammatory lesions (4.29) and NonInflammatory lesions (4.96) respectively.
Table 2. Patient Acne Lesion Characteristics
Total Acne Lesions
n
Mean (SD)
Median
Min, Max
Inflammatory Acne Lesions
n
Mean (SD)
Median
Min, Max
Non-Inflammatory Acne Lesions
n
Mean (SD)
Median
Min, Max
PI5542B OS Validation Manuscript Supplementary file v2.0
196
9.25 (13.53)
4.00
0.00, 87.00
196
4.29 (6.06)
2.00
0.00, 33.00
196
4.96 (9.38)
0.00
0.00, 74.00
Page 4 of 8
11 September 2008
A cross tabs of the number of each type of lesion experienced by the patient (comedone,
papule, pustule and nodule) crossed with the severity rating as described by the patient was
conducted Results are provided in Table 3 to Table 6.
Table 3. Facial Comedone counts compared with acne severity level
Current acne
Mild
Moderate
Severe
N
49
45
11
Comedone counts
Mean (se)
Median
3.3 (0.9)
1
10.1 (2.2)
5
5.4 (2.5)
0
Min, Max
0, 34
0, 74
0, 25
Table 4. Facial Papule counts compared with acne severity level
Current acne
Mild
Moderate
Severe
N
48
45
11
Papule counts
Mean (se)
Median
2.6 (0.6)
1
6.4 (0.7)
5
8.1 (2.2)
8
Min, Max
0, 17
0, 19
0, 19
Table 5. Facial Pustule counts compared with acne severity level
Current acne
Mild
Moderate
Severe
N
48
45
11
Pustule counts
Mean (se)
Median
0.9 (0.4)
0
1.8 (0.5)
0
0.7 (0.5)
0
Min, Max
0, 18
0, 18
0, 5
Table 6. Facial Nodule counts compared with acne severity level
Current acne
Mild
Moderate
Severe
N
46
45
11
Nodule counts
Mean (se)
Median
0.1 (0.1)
0
0.1 (0.1)
0
0.1 (0.1)
0
PI5542B OS Validation Manuscript Supplementary file v2.0
Min, Max
0, 4
0, 4
0, 1
Page 5 of 8
11 September 2008
APPENDIX 3. ITEM REDUCTION OF THE OSSAS
Following the initial item reduction analyses, and qualitative data from the adolescent focus
groups a number of items were deleted from the OSSAS.
Table 7 provides details of the
items that were deleted and a rationale for the deletion.
Table 7. OSSAS item deletion table
Item
A. Please check ALL the areas of the face that are
usually oily?
B. Which is the MOST bothersome oily area of your
face?
1. Currently, how bothersome is your MOST
bothersome area (referring to your selection above)?
3. Have you had sunburn?
4. Has it been very humid?
5. Has it been very hot?
6. For female patients only – what is the length of time
since your last menstrual period?
7. Prior to answering the questions in this section,
please indicate the degree to which you feel you can
rate the oiliness of your skin without touching or looking
at it by placing a checkmark below your answer
9. How itchy does your skin feel?
10. How hot does your skin’s surface feel?
12. How much do your pores feel blocked or clogged?
18. How enlarged to your skin pores look?
15. Overall, without touching or looking at your skin, how
dry does your skin feel?
20. How uneven does your skin surface look due to
flaking or dryness?
25. Overall how dry does your skin look?
34. Overall how dry does your skin seem when
touched?
17. How flushed (red) is your skin?
21. How visible is your acne?
22. How visible are your blackheads?
23. How visible are your pimples?
27. How bumpy does the surface of your skin feel when
stroked?
29. How oily do your fingers look after stroking your
skin?
30. How oily do your nails look after stroking your skin?
32. How moist is your skin when touched?
Reason for deletion
These introductory items asked the subjects to
identify areas of the face that were most oily.
However, following discussion it was decided to
concentrate on the face as a whole, therefore these
items were deleted.
These introductory questions asked about
covariates of oily skin and were included for
validation, but are not necessary as part of the oily
skin measure itself.
This item was concluded to be confusing and was
originally included to aid validation, however does
not belong as part of the oily skin measure.
Deleted due to poor psychometric performance and
low importance ratings given by subjects in the
online focus groups.
Adolescent subjects found this item problematic
and difficult to assess
Assessing the dryness of skin is not part of the
concept of oily skin and would not add any
additional information
Assessing redness was not part of the concept of
oily skin, and the item failed to load on any factor in
the factor analysis
It was concluded that acne would be assessed by
an additional measure specific to acne.
This item was confusing, and it was unclear what
the item was measuring. The item was given a low
importance rating by subjects in the online focus
groups.
These items were proxy measures of oily skin and it
was felt that they would be more accurately
assessed by direct measures of oily skin.
Item loaded on two scales in the factor analysis
In addition to the deletion of some of the items, a number of revisions were recommended
based on the qualitative adolescent focus group data and the item reduction process. The
revisions made to the OSSAS items and the rationale for each revision are detailed in Table
8.
PI5542B OS Validation Manuscript Supplementary file v2.0
Page 6 of 8
11 September 2008
Table 8. Revisions to the OSSAS
Revision to instrument
An item assessing ‘how ‘greasy’ the face
feels’ was added to the questionnaire
Rationale for revision
Adolescent focus groups indicated the
importance of this wording to assess skin
oiliness. The item was previously omitted as
it was felt ‘greasy’ had too much of a
negative connotation.
This item was added to provide a single item
that may assess the overall oiliness of the
skin
It was felt that made more sense for subjects
to assess the oiliness of the face as a whole.
An overall item assessing the oiliness of the
‘skin on your face’ was added to the
questionnaire
The items asking the subject to select the
area of the face that was most ‘bothersome’
was removed and replaced by an
assessment of the face as a whole for visual
and sensation items.
The response options were revised to rate
the oily skin from ‘Not at all’ to ‘Extremely’ on
a scale of 0-10
Evidence from the adolescent focus groups
revealed that they found the original
response options difficult to understand and
that they would prefer this format.
APPENDIX 4. CONCURRENT VALIDITY
The concurrent validity of the OSSAS scales was assessed by examining correlations of the
OSSAS domain scores with the three Skindex-29 scale scores (Table 9) and the four AcneQoL scales scores (Table 10). Pearson’s correlations are presented below, all correlations
were statistically significant (p<0.0001).
Table 9. Concurrent Validity: correlations of the OSSAS with the Skindex-29 scores at
baseline (N=195)*
Domain
Skindex/
Skindex/
Skindex/
Emotions
Symptoms
Functioning
OSSAS: Sensation
0.38
0.38
OSSAS: Visual
0.30
0.23
OSSAS: Tactile
0.30
0.20
OSSAS: Blotting
0.25
0.20
*All correlations were statistically significant at the p<0.0001 level.
0.32
0.25
0.24
0.24
Table 10. Concurrent Validity: Correlations of the OSSAS with the Acne-QoL scores at
baseline (N=120)*
Acne-QoL
Domain
Acne-QoL
Acne-QoL
Self-
Acne-QoL
Role-
Acne
perception
Role-social
emotional
symptom
OSSAS: Sensation
-0.35
-0.23
-0.35
OSSAS: Visual
-0.18
-0.08
-0.23
OSSAS: Tactile
-0.24
-0.10
-0.22
OSSAS: Blotting
-0.17
-0.10
-0.17
*All correlations were statistically significant at the p<0.0001 level.
PI5542B OS Validation Manuscript Supplementary file v2.0
-0.34
-0.23
-0.21
-0.19
Page 7 of 8
11 September 2008
The concurrent validity of the OSIS domains was assessed by examining correlations of the
OSIS domain scores with the three Skindex-29 scale scores (Table 11) and the four AcneQoL domain scores (Table 12). Pearson’s correlations are presented below, all correlations
were statistically significant (p<0.0001).
Table 11. Concurrent validity: correlations of OSIS scores with Skindex-29 scores at
visit 1 (n=195)*
Domain
Skindex/
Skindex/
Skindex/
Emotions
Symptoms
Functioning
OSIS: Self Image
0.69
0.37
OSIS: Annoyance
0.73
0.44
*All correlations were statistically significant at the p<0.0001 level.
0.56
0.59
Table 12. Concurrent validity: correlations of OSIS scores with Acne-QoL scores at
visit 1 (N=196)*
Domain
Acne-QoL
Acne-QoL
Acne-QoL
Acne-QoL
Self-perception
Role-social
Role-emotional
Acne symptom
OSIS: Self Image
-0.67
-0.53
-0.62
OSIS: Annoyance
-0.65
-0.57
-0.73
*All correlations were statistically significant at the p<0.0001 level.
PI5542B OS Validation Manuscript Supplementary file v2.0
-0.48
-0.62
Page 8 of 8