online Appendix 2

Full text

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068 suppl2. https://doi.org/10.18549/PharmPract.2018.01.1068
Online Appendix 2: Factor analysis
Self-efficacy
Initially, we tested the factorability of six items of question PB8 to measure Self Efficacy.
Items were correlated at least 0.4 with at least one other item which suggests
factorability of the items. Kaiser-Meyer-Olkin value of 0.849 confirms sample adequacy
for factor analysis. Communalities are all above 0.5 suggests that each item shares
common variance with other items. All the above mentioned indicators allowed us to run
factor analysis with all the items of question PB8.
Total Variance Explained
Factor
Initial Eigenvalues
Extraction Sums of Squared Loadings
Total
% of Variance
Cumulative %
Total
% of Variance
Cumulative %
1
3.910
65.161
65.161
3.508
58.461
58.461
2
.654
10.892
76.053
3
.592
9.858
85.912
4
.400
6.667
92.578
5
.263
4.385
96.964
6
.182
3.036
100.000
Extraction Method: Principal Axis Factoring.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
2
Factor Matrixa
Factor
Item
1
PB8b
How sure are you that you could prescribe in a
clinical area that you are familiar with?
.829
PB8a
How sure are you that you could perform a
patient assessment to prescribe?
.814
PB8e
How sure are you that you could initiate new
therapy for a patient?
.809
PB8f
How sure are you that you could accept
responsibility for medication management?
.749
PB8d
How sure are you that you could adapt a
prescription for patients starting a new therapy?
.728
PB8c
How sure are you that you could prescribe in a
clinical area that you are not familiar with?
.641
Extraction Method: Principal Axis Factoring.
a. 1 factors extracted. 5 iterations required.
First, we ran principle component analysis without any rotation to identify composite
scores for the underlying factors. First factor only had Eigenvalue over one and that
factor explained variance of 65.16%. The scree plot also suggested one factor solution.
Then we ran exploratory factor analysis. We did not use any rotation as we have only
one factor. We found only one factor with loadings of all items above 0.6. All the items of
the factor explain self-efficacy belief of prescribing activities such as level of confidence
of assessing patient, prescribing in familiar clinical area or non-familiar clinical area,
starting a new therapy, adapting prescription, managing medication. We also tested the
internal consistency of the scale using Cronbach’s alpha. The alpha value was 0.892
suggesting higher reliability of self-efficacy scale.
Impact of prescribing on practice
We tested the factorability of nine items of questionnaire to measure impact of
prescribing on practice in the survey (i.e. Question PB7). All the items, except PB7 c, d
and i, were correlated by at least 0.3 with at least one other item. We removed PB7 c, d
and i items from factor analysis due to low correlation (i.e. below 0.3). Kaiser-Meyer-
Olkin value of 0.737 confirms sample adequacy for factor analysis. Communalities are all
above 0.5 suggests that each item shares common variance with other items. All the
above mentioned indicators allowed us to run factor analysis with the six items of
question PB7.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
3
Total Variance Explained
Factor
Initial Eigenvalues
Extraction Sums of Squared Loadings
Rotation Sums
of Squared
Loadingsa
Total
% of Variance
Cumulative %
Total
% of Variance
Cumulative %
Total
1
2.942
49.039
49.039
2.554
42.571
42.571
2.219
2
1.315
21.924
70.963
.946
15.769
58.340
1.901
3
.672
11.197
82.160
4
.450
7.492
89.652
5
.345
5.751
95.403
6
.276
4.597
100.000
Extraction Method: Principal Axis Factoring.
a. When factors are correlated, sums of squared loadings cannot be added to obtain a total variance.
Pattern Matrixa
Factor
Item
1
2
PB7a
To what extent has prescribing impacted the
following for you: Job satisfaction?
.918
PB7b
To what extent has prescribing impacted the
following for you: Professional image?
.741
PB7h
To what extent has prescribing impacted the
following for you: Quality of physician
relationship?
.542
PB7g
To what extent has prescribing impacted the
following for you: Quality of patient care?
.454
.406
PB7e
To what extent has prescribing impacted the
following for you: Time spent with patient?
.824
PB7f
To what extent has prescribing impacted the
following for you: Time spent assessing
patients?
.818
Extraction Method: Principal Axis Factoring.
Rotation Method: Oblimin with Kaiser Normalization.
a. Rotation converged in 5 iterations.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
4
Factor Correlation Matrix
Factor
1
2
1
1.000
.394
2
.394
1.000
Extraction Method: Principal Axis
Factoring.
Rotation Method: Oblimin with
Kaiser Normalization.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
5
First, we used principle component analysis without any rotation to identify composite
scores for the underlying factors. First two factors only had Eigenvalue over one and
these two factors explained variance of 49% and 22% respectively. The leveling of
eigenvalues after two factors in the scree plot also suggested solution of two factors.
Then we ran exploratory factor analysis with varimax and oblimin rotation. Correlation of
0.4 between two factors suggested choosing oblimin solution. Three items PB7 a, b, h
were loaded in factor 1 with loading factor above 0.5 and two items- PB7 e and f were
loaded in factor 2 with loading factor above 0.8. Items loaded in factor 1 explained the
impact of prescribing on positive professionalism such as job satisfaction, professional
image, quality of physician relationship. Items, loaded in factor 2, explained the impact
on patient care for example time spent with patient or for patient assessment. Item PB7
g, explaining impact on quality of patient care, were loaded in both factor with loading
factor above 0.4. Due to similarities of this item with impact on patient care we included
this item in factor 2 (i.e. impact on patient care). We also tested the internal consistency
of both factors using Cronbach’s alpha. The alpha value of impact on positive
professionalism and impact on patient care scale were 0.759 and 0.775 respectively
suggesting higher reliability of the scales.
Support for prescribing adoption
Initially, we tested the factorability of nine items of questionnaire to measure support for
prescribing adoption in the survey (i.e. Question PB6). All the items were correlated by at

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
6
least 0.3 with at least one other item which suggests factorability of the items. Kaiser-
Meyer-Olkin value of 0.850 confirms sample adequacy for factor analysis. Communalities
of 0.3 or above suggest that each item shares common variance with other items. All the
above mentioned indicators allowed us to run factor analysis with the nine items of
question PB6.
Total Variance Explained
Factor
Initial Eigenvalues
Extraction Sums of Squared Loadings
Rotation Sums
of Squared
Loadingsa
Total
% of Variance
Cumulative %
Total
% of Variance
Cumulative %
Total
1
4.085
45.394
45.394
3.596
39.954
39.954
3.237
2
1.111
12.344
57.738
.727
8.074
48.028
2.755
3
.782
8.687
66.425
4
.714
7.931
74.356
5
.630
6.995
81.351
6
.578
6.425
87.776
7
.466
5.173
92.949
8
.381
4.233
97.182
9
.254
2.818
100.000
Extraction Method: Principal Axis Factoring.
a. When factors are correlated, sums of squared loadings cannot be added to obtain a total variance.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
7
Pattern Matrixa
Factor
Item
1
2
PB6c
To what extent do the following factors affect
your prescribing activities: My practice
environment?
.797
PB6a
To what extent do the following factors affect
your prescribing activities: Pharmacy staffing
at my practice location?
.690
PB6b
To what extent do the following factors affect
your prescribing activities: Access to patient
information?
.688
PB6i
To what extent do the following factors affect
your prescribing activities: Employer's
expectations?
.539
PB6d
To what extent do the following factors affect
your prescribing activities: Patient
expectations?
.530
PB6g
To what extent do the following factors affect
your prescribing activities: My education and
training?
.366
-.312
PB6h
To what extent do the following factors affect
your prescribing activities: Requirement to
document patient care?
.361
-.301
PB6e
To what extent do the following factors affect
your prescribing activities: Relationships with
physicians?
-.955
PB6f
To what extent do the following factors affect
your prescribing activities: Relationships with
other health care professionals?
-.751
Extraction Method: Principal Axis Factoring.
Rotation Method: Oblimin with Kaiser Normalization.a
a. Rotation converged in 5 iterations.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
8
Factor Correlation Matrix
Factor
1
2
1
1.000
-.595
2
-.595
1.000
Extraction Method: Principal Axis
Factoring.
Rotation Method: Oblimin with
Kaiser Normalization.
First, we used principle component analysis without any rotation to identify composite
scores for the underlying factors. First two factors only had Eigenvalue over one and
these two factors explained variance of 41% and 10% respectively. The leveling of
eigenvalues after two factors in the scree plot also suggested solution of two factors.
Then we ran exploratory factor analysis with varimax and oblimin rotation. Correlation of
0.6 between two factors suggested choosing oblimin solution. Five items – PB6 a, b, c, d,
and i were loaded in factor 1 with loading factor above 0.53 and two items- PB6 e and f
were loaded in factor 2 with loading factor above 0.7. Items PB6 a,b,c and i of factor 1
explained the support from practice environment such as access to patient information,
staffing at practice location, employer’s expectation, practice environment. Other item
PB6 d of factor 1 explains patient expectation which we considered as part of support
from practice environment as because patient expectation has a profound impact on the

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
9
practice environment. Items, loaded in factor 2, explained the support from relationship
with healthcare professionals (HCPs). We eliminated PB6 g and h from the scales as
these two items’ loading factors were below 0.4 in both factors. We also tested the
internal consistency of both factors using Cronbach’s alpha. The alpha value support
from practice environment scale and support from relationship with healthcare
professionals scale were 0.78 and 0.851 respectively suggesting higher reliability of the
scales.
Use of Electronic Health Record (HER)-Netcare
We tested the factorability of five items of question GP8 to measure level of EHR use. All
the items were correlated by at least 0.3 with at least one other item which suggests
factorability of the items. Kaiser-Meyer-Olkin value of 0.668 confirms sample adequacy
for factor analysis. Communalities above 0.6 suggest that each item shares common
variance with other items. All the above mentioned indicators allowed us to run factor
analysis with the five items of question GP8.
Total Variance Explained
Factor
Initial Eigenvalues
Extraction Sums of Squared
Loadings
Rotation Sums of Squared
Loadings
Total
% of Variance
Cumulative %
Total
% of Variance
Cumulati
ve %
Total
% of
Varian
ce
Cumulative
%
1
2.159
43.178
43.178
1.746
34.930
34.930
1.729
34.581
34.581
2
1.334
26.687
69.865
.704
14.089
49.018
.722
14.438
49.018
3
.659
13.181
83.046
4
.486
9.715
92.761
5
.362
7.239
100.000
Extraction Method: Principal Axis Factoring.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
10
Rotated Factor Matrixa
Factor
Item
1
2
GP8e
In Netcare, I look up: - Medical history such as
diagnostic tests and discharge or admission
history
.848
GP8d
In Netcare, I look up: - Lab values
.723
GP8c
In Netcare, I look up: - Medication
history/allergies/refills including Pharmaceutical
Information Network
.694
GP8b
In Netcare, I look up: - Double doctoring or
multiple pharmacies
.655
GP8a
In Netcare, I look up: - Demographic information
including personal health care numbers (number
from Alberta Health card)
.522
Extraction Method: Principal Axis Factoring.
Rotation Method: Varimax with Kaiser Normalization.a
a. Rotation converged in 3 iterations.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
11
Factor Transformation Matrix
Factor
1
2
1
.992
.129
2
-.129
.992
Extraction Method: Principal Axis
Factoring.
Rotation Method: Varimax with
Kaiser Normalization.
First, we used principle component analysis without any rotation to identify composite
scores for the underlying factors. First two factors only had Eigenvalue over one and
these two factors explained variance of 43% and 27% respectively. The leveling of
eigenvalues after two factors in the scree plot also suggested solution of two factors.
Then we ran exploratory factor analysis with varimax and oblimin rotation. Correlation of
0.1 between two factors suggested choosing varimax solution. Three items – GP8 c, d, e
were loaded in factor 1 with loading factor above 0.6 and two items- GP8 a and b were
loaded in factor 2 with loading factor above 0.5. Items, loaded in factor 1, explained the
EHR use for intense patient care such as checking patient history, lab values, and
pharmaceutical information. Items of factor 2 explained the technical use of EHR for
example demographic or doctor visit information. We also tested the internal
consistency of both factors using Cronbach’s alpha. The alpha value of the EHR use for
intense patient care scale was 0.8 suggesting higher reliability of the scales. The alpha
value of technical use of EHR scale was 0.512.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
12
Prescribing belief
Initially we tested the factorability of five items of question PP1 to prescribing belief. All
the items, except PP1 a and c, were correlated by at least 0.3 with at least one other
item which suggests factorability of the items. Kaiser-Meyer-Olkin value of 0.606
confirms sample adequacy for factor analysis.
Total Variance Explained
Factor
Initial Eigenvalues
Extraction Sums of Squared Loadings
Rotation Sums of Squared
Loadings
Total
% of
Variance
Cumulative %
Total
% of
Variance
Cumulative %
Total
% of
Variance
Cumulative
%
1
1.667
33.344
33.344
1.039
20.784
20.784
1.034
20.679
20.679
2
1.059
21.182
54.527
.194
3.883
24.667
.199
3.988
24.667
3
.954
19.079
73.606
4
.743
14.868
88.473
5
.576
11.527
100.000
Extraction Method: Principal Axis Factoring.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
13
Rotated Factor Matrixa
Factor
Item
1
2
PP1b
Patients are responsible for ensuring they
have a sufficient supply of medications?
.666
PP1d
Pharmacist prescribing increases
pharmacists' professional liability?
.640
PP1e
Pharmacists should only extend refills once?
.405
PP1a
Pharmacist prescribing is an extension of the
role that pharmacists already fulfill?
.375
PP1c
Pharmacist prescribing helps patients avoid
physician follow-up?
Extraction Method: Principal Axis Factoring.
Rotation Method: Varimax with Kaiser Normalization.
a. Rotation converged in 3 iterations.

Guirguis LM, Hughes CA, Makowsky MJ, Sadowski CA, Schindel TJ, Yuksel N, Faruquee CF. Development and validation of a survey
instrument to measure factors that influence pharmacist adoption of prescribing in Alberta, Canada. Pharmacy Practice 2018 Jan-
Mar;16(1):1068. https://doi.org/10.18549/PharmPract.2017.01.1068
www.pharmacypractice.org (eISSN: 1886-3655 ISSN: 1885-642X)
14
Factor Transformation Matrix
Factor
1
2
1
.997
-.079
2
.079
.997
Extraction Method: Principal Axis
Factoring.
Rotation Method: Varimax with
Kaiser Normalization.
First, we used principle component analysis without any rotation to identify composite
scores for the underlying factors. First two factors only had Eigenvalue over one and
these two factors explained variance of 33% and 21% respectively. The leveling of
eigenvalues after two factors in the scree plot also suggested solution of two factors.
Then we ran exploratory factor analysis with varimax and oblimin rotation. Correlation of
0.1 between two factors suggested choosing varimax solution. Three items – PP1 b, d, e
were loaded in factor 1 with loading factor above 0.4. Two items PP1a and c were
eliminated as they did not load above 0.4 in any factor. Items loaded in factor 1 explained
the prescribing belief of pharmacist. We also tested the internal consistency of the factor
using Cronbach’s alpha. The alpha value of the prescribing belief was 0.583.