A total of 200 Umrah pilgrims responded to this study. On data screening, no missing data was found. The age of the participants from this study ranged from 18 to 80 years old with a mean age of 39.13 (SD 16.03). The females (65.5%) dominated the number of pilgrims. The socio-demographic characteristics of the participants are shown in Table 1.
Table 1: Socio-demographic characteristics of participants (n=200)
Variables
|
|
Mean (SD)
|
Frequency (%)
|
Age (years)
|
|
39.13 (16.029)
|
|
Gender
|
Male
|
|
65 (35.5)
|
Female
|
|
131 (65.5)
|
Ethnicity
|
Malay
|
|
197 (98.5)
|
Indian
|
|
1 (0.5)
|
Others
|
|
2 (1.0)
|
Marital status
|
Single
|
|
89 (44.5)
|
Married
|
|
109 (54.5)
|
Divorced/widowed
|
|
2 (1.0)
|
Occupation
|
Student
|
|
19 (9.5)
|
Civil servant
|
|
37 (18.5)
|
Private sector
|
|
95 (47.5)
|
Pensioner
|
|
22 (11)
|
Housewife
|
|
15 (7.5)
|
Self-employed
|
|
12 (6)
|
Highest level of education
|
PhD
|
|
4 (2.0)
|
Master’s degree
|
|
13 (6.5)
|
Bachelor’s degree
|
|
42 (21.0)
|
Diploma
|
|
73 (36.5)
|
Secondary school
|
|
54 (27.0)
|
Primary school
|
|
14 (7.0)
|
History of vaccination
|
Meningococcal vaccine
|
|
60 (30)
|
Influenza (flu) vaccine
|
|
29 (14.5)
|
Pneumococcal vaccine
|
|
24 (12.0)
|
Presence of Co-morbidities
|
Chronic lung disease
|
|
1 (0.5)
|
Neuromuscular disease
|
|
9 (4.5)
|
Allergic rhinitis
|
|
2 (1.0)
|
Diabetes
|
|
6 (3.0)
|
Hypertension
|
|
29 (14.5)
|
Heart disease
|
|
2 (1.0)
|
Chronic kidney disease
|
|
2 (2)
|
Immune deficiency disorders
|
|
1(0.5)
|
In the knowledge section, IRT analysis results showed an acceptable range for both difficulty (− 3 to + 3) and the discrimination parameter on each of the items in all the sub domains. The sub-domains are SD1 (K1i, K1ii, K1iii), SD2 (K2i, K2ii, K2iii, K2iv, K2v, K3), SD3 (K4i, K4ii, K4iii, K4iv, K4v and K4vi), SD4 (K5i, K5ii, K5iii, K5iv and K5v), prevention practices (K6i, K6ii, K6iii, K6iv and K6v) and SD5 (K7i, K7ii, K7iii, K8 and K9) covering the aetiology, transmission, risk factors, complications, preventive practices and the use of personal protective equipment (PPE). However, all the items were retained because they had acceptable difficulty and discrimination values. The amount of information tapped by the items between − 3 and + 3 difficulty range was 93.1%. The unidimensionality assumption was not supported by the modified parallel test at α = 0.05 (p = 0.010). In terms of internal consistency reliability, Cronbach’s alpha was 0.9. IRT analysis for the psychometric characteristics of the domain as shown in Table 2.
Table 2: Result of the IRT analysis in the knowledge section (n = 200)
|
|
Items
|
b
|
a
|
λ
|
χ2 (df = 8)
|
P values
|
1
|
|
Flu-like illnesses are caused by:
|
|
|
|
|
|
|
i
|
Allergies
|
-0.41
|
3.50
|
0.9
|
100.55
|
<0.001
|
|
ii
|
Bacteria
|
-0.54
|
2.16
|
0.78
|
62.98
|
<0.001
|
|
iii
|
Virus
|
-1.04
|
3.12
|
0.87
|
120.2
|
<0.001
|
2
|
|
Flu-like illnesses are spread by:
|
|
|
|
|
|
|
i
|
Air
|
-1.10
|
3.39
|
0.88
|
24.34
|
0.002
|
|
ii
|
Dust
|
-0.86
|
2.24
|
0.79
|
26.19
|
0.001
|
|
iii
|
Sharing towels with an infected person
|
-0.32
|
3.95
|
0.94
|
52.78
|
<0.001
|
|
iv
|
Water
|
0.24
|
2.32
|
0.82
|
88.76
|
<0.001
|
|
v
|
Shaking the hands of an infected person with a cough and/or cold
|
-0.16
|
1.90
|
0.75
|
52.42
|
<0.001
|
3
|
|
Flu-like illnesses are spread quickly
|
-1.17
|
1.41
|
0.64
|
101.32
|
<0.001
|
4
|
|
The following persons are at an increased risk of flu-like illnesses:
|
|
|
|
|
|
|
i
|
Asthmatics
|
-0.87
|
2.83
|
0.86
|
43.63
|
<0.001
|
|
ii
|
Diabetics
|
0.40
|
4.32
|
0.93
|
21.16
|
0.007
|
|
iii
|
People with arthritis
|
0.43
|
2.34
|
0.80
|
50.84
|
<0.001
|
|
iv
|
Senior citizens aged 65 and older
|
-0.57
|
1.94
|
0.75
|
29.26
|
<0.001
|
|
v
|
Smokers
|
-0.14
|
3.08
|
0.87
|
75.46
|
<0.001
|
|
vi
|
Those in crowded places/among a lot of people
|
-1.13
|
1.83
|
0.73
|
49.71
|
<0.001
|
5
|
|
What are the complications of flu-like illnesses?
|
|
|
|
|
|
|
i
|
Bronchitis
|
-0.14
|
2.93
|
0.91
|
126.49
|
<0.001
|
|
ii
|
Difficulty in breathing
|
0.64
|
6.26
|
0.88
|
22.80
|
0.004
|
|
iii
|
Multi-organ failure
|
0.55
|
2.85
|
0.89
|
170.57
|
<0.001
|
|
iv
|
Pneumonia
|
-0.27
|
2.76
|
0.90
|
91.12
|
<0.001
|
6
|
|
The following practices can help protect you from flu-like illnesses:
|
|
|
|
|
|
|
i
|
Covering your nose with your hands
|
-0.67
|
1.75
|
0.92
|
71.99
|
<0.001
|
|
ii
|
Ensuring a healthy diet
|
-1.05
|
2.29
|
0.50
|
38.14
|
<0.001
|
|
ii
|
Receiving vaccinations
|
-0.80
|
2.26
|
0.85
|
66.51
|
<0.001
|
|
iv
|
Washing your hands with hand sanitizers
|
-0.86
|
6.29
|
0.78
|
15.08
|
<0.001
|
|
v
|
Wearing a face mask
|
-1.22
|
5.07
|
0.71
|
10.75
|
<0.001
|
7
|
|
The following are reasons for wearing a mask:
|
|
|
|
|
|
|
i
|
Being in crowded places
|
-1.03
|
6.11
|
0.97
|
11.82
|
<0.001
|
|
ii
|
Being near people who are coughing
|
-1.26
|
4.83
|
0.96
|
20.75
|
0.008
|
|
iii
|
When I am sick
|
-0.91
|
4.33
|
0.94
|
49.03
|
<0.001
|
8
|
|
A cloth facial mask is as effective as a 2-ply surgical facial mask
|
1.02
|
1.29
|
0.60
|
60.85
|
<0.001
|
9
|
|
If I am not sick, the used face mask can be stored in a bag for later use
|
0.72
|
1.56
|
0.67
|
182.10
|
<0.001
|
For the attitude domain, the two-factor model was then tested by CFA using an MLR estimation method. MLR was used because the data did not follow a multivariate normal distribution required by the MLR. Satisfactory model fitness was not demonstrated by the initial 12-item factor. To achieve the model fitness, the maximum likelihood (ML) values were examined and re-analysed to achieve a better model fit. To be included in the model, items with high correlated errors within the same factor will be considered. The two-factor model showed a good fit (χ2 [df = 6] = 43, p < 0.001; CFIrobust = 0.928; TLIrobust = 0.890; RMSEArobust = 0.063; SRMR = 0.079) as shown in Table after correlated errors (A12A↔A13, r = 0.341; A3↔A9, r = -0.267; A5A↔A5B, r = 0.265; A8↔A7, r = 0.268; A8↔A9, r = 0.240; A10↔A4, r = -0.237; A10↔A7, r = -0.191; A3↔A5B, r = 0.267; A9↔A5B, r = -0.168; A10↔A5B, r = 0.205) were added. However, the two sub-domains under attitude (barriers to compliance and self-motivation) have correlation between them of r = 0.444. The composite reliability of the barriers to compliance and self-motivation factor all have satisfactory cut-off value of >0.7 as summarize in Table 3.
Table 3: Results of CFA of the attitude section
Factors
|
Items
|
Factor loading
|
Reliability (Raykov’s rho)
|
Barriers to compliance
|
A3: Since the bird flu, SARS, MERS-COV and H1N1 crises are over, I no longer need to worry about contracting flu-like illnesses
|
0.696
|
0.76
|
A8: I am generally opposed to wearing a face mask
|
0.555
|
A9: Flu vaccinations have unpleasant side effects
|
0.376
|
A10: I am influence by negative news about flu vaccines
|
0.751
|
A11: It is too much trouble to get a flu vaccine
|
0.751
|
Self-motivation
|
A4 If I have a flu-like illness, I may spread it to others
|
0.516
|
0.72
|
A5: I feel that someone that have influenza-like illness should:
|
|
A5A: cover his mouth and nose with his bare hand when coughing or sneezing
|
0.603
|
A5B: cover his mouth and nose with a handkerchief when coughing or sneezing
|
0.402
|
A6: Influenza vaccines protects hajj pilgrims from influenza
|
0.75
|
A7: Using a hand wash can prevent you from getting flu like illness
|
0.652
|
A12A: I think coughs and the flu can be prevented by wearing a mask outside my house
|
0.424
|
A13: Wearing a well-fitting face mask is effective in preventing flu-like illnesses
|
0.431
|
For practice domain which comprises of 13 items, the two-factor model was analyzed by CFA. The model showed an acceptable fitness, as shown in Table 4 (χ2 [df = 64] = 31.49, p < 0.001; CFIrobust = 0.903; TLIrobust = 0.882; RMSEArobust = 0.073; SRMR = 0.067). The correlations between the factors were: Healthy-lifestyle↔Prevention-practices (r = 0.471). The composite reliability of the healthy lifestyle and prevention practices factors were above the cutoff value of 0.7 (Raykov’s rho = 0.863 and 0.827), despite the low standardized loading for item P7.
Table 4: Results of CFA of the practice domain
Factors
|
Items
|
Factor loading
|
Reliability (Raykov’s rho)
|
Health lifestyle
|
P1: I eat vegetables
|
0.918
|
0.863
|
P2: I eat fruits
|
0.888
|
P5: I use soap to wash my hands
|
0.664
|
Prevention practices
|
P4: When wearing a mask, I test it to ensure it fits properly
|
0.535
|
0.827
|
P6: I use disinfectant or disposable wipes or hand gel to wash my hands
|
0.483
|
P7: I use a washable cloth handkerchief to clean my hands
|
0.284
|
P8: I wash my hands after:
|
|
P8A: touching the personal items of someone who has a cough and/or cold
|
0.744
|
P8B: shaking hands with people who have a cough and/or cold
|
0.787
|
P8C: touching doorknobs
|
0.692
|
P9: I refrain from:
|
|
P9A: being close to those who cough or sneeze
|
0.562
|
P9B: shaking the hands of those who have a cough and/or cold
|
0.577
|
P9C: often touching my nose
|
0.365
|
P10: I received the flu vaccine
|
0.511
|
The model showed an acceptable fitness for both attitude and practice. In the attitude domain, two-factor model showed a good fit (χ2 [df = 6] = 43, p < 0.001; CFIrobust = 0.928; TLIrobust = 0.890; RMSEArobust = 0.063; SRMR = 0.079) after correlated errors (A12A↔A13, r = 0.341; A3↔A9, r = -0.267; A5A↔A5B, r = 0.265; A8↔A7, r = 0.268; A8↔A9, r = 0.240; A10↔A4, r = -0.237; A10↔A7, r = -0.191; A3↔A5B, r = 0.267; A9↔A5B, r = -0.168; A10↔A5B, r = 0.205) were added. For the practice domain, the fitness indices (χ2 [df = 64] = 31.49, p < 0.001; CFIrobust = 0.903; TLIrobust = 0.882; RMSEArobust = 0.073; SRMR = 0.067) are well represented. The fitness indices are summarized in Table 5.
Table 5: Fit Indices for Confirmatory Factor Models
Factors
|
No of items
|
|
|
Goodness of fit indices
|
X2(df)
|
P value
|
df
|
CFI
|
TLI
|
RMSEA
|
SRMR
|
Attitude
|
12
|
76.8 (43)
|
< 0.001
|
66
|
0.928
|
0.890
|
0.063
|
0.079
|
Practice
|
13
|
121 (76)
|
< 0.001
|
64
|
0.903
|
0.882
|
0.073
|
0.067
|