Respondent characteristics
A total of 194 patients were approached, of which 28 did not give consent to take part. Thus, the remaining 166 participants agreed and completed the survey giving a response rate of 85.6%. The number of male participant showed preponderance (54.8%). Most of the respondents (29.5%) were between the ages of 18-27 years with a mean age of 38.5±13.8 years. More than half of the participants (56.6%) were married; and about 53 percent had attended their primary (1-8 grades) education and above. Nearly half (48.8%) of the participants claimed as they have lived with DM for about 1-5 years with a mean (±SD) of 2.5 ± 0.9 years. Marked number of (58.4%) of respondents used insulin therapy for about 1-5 years with a mean (±SD) duration of 2.3 ± 0.8 years. More than half (55.4%) of the participants obtain insulin merely through purchase (Table 1).
Table 1: Distribution of selected characteristics of patients with diabetes mellitus by sex (N=166); 2019
Variables
|
Total
|
Male
|
Female
|
P-value
|
N (166) %(100.0)
|
N (91) % (54.8)
|
N (75) %(45.2)
|
Age in years: 18-27
28-40
41-50
>50
Mean ± SD
|
49 (29.5)
43 (25.9)
43 (25.9)
31 (18.7)
38.5±13.8
|
28
22
22
19
|
21
21
21
12
|
0.778
|
Residence: Rural
Urban
|
86 (51.8)
80 (48.2)
|
43
48
|
43
32
|
//
|
Marital status: Single
Married
Divorced
Widowed
|
56 (33.7)
94 (56.6)
10 (6)
6 (3.6)
|
31
52
8
0
|
25
42
2
6
|
0.02
|
Occupation: Farmer
Employer
Merchant
Housewife
Student
|
34 (20.5)
31 (18.7)
38 (22.8)
36 (21.7)
27 (16.3)
|
31
18
21
8
13
|
3
13
17
28
14
|
<0.001
|
Educational status: Illiterate
Read & write only
Primary and secondary education
College & above
|
48 (28.9)
30 (18.1)
52 (31.3)
36 (21.7)
|
27
17
31
16
|
21
13
21
20
|
0.546
|
Duration of DM (in years): 0.25-1
>1-5
>5-10
>10
Mean ± SD
|
15 (9)
81 (48.8)
40 (24.1)
30 (18.1)
2.5 ± 0.9
|
5
38
23
25
|
10
43
17
5
|
0.002
|
Duration of Insulin therapy
(in years): 0.25-1
>1-5
>5-10
>10
Mean ± SD
|
16 (9.6)
97 (58.4)
39 (23.5)
14 (8.4)
2.3 ± 0.8
|
6
44
27
14
|
10
53
12
0
|
<0.001
|
Getting of insulin: Freely
Payment
|
74 (44.6)
92 (55.4)
|
39
52
|
35
40
|
//
|
Knowledge score
The overall median (IQR) knowledge level of the study subjects about insulin storage and handling techniques was 9 (7.8-11), out of 14 (64.3%). Most of the participants (44%) had moderately adequate knowledge, while nearly one-third of them (31.3%) had adequate knowledge, and the rest of the patients (24.7%) had inadequate knowledge levels. Large number of patients (82%) knew that outdated (expired) insulin should not be used. Less than half (44%) of the respondents were aware of the acceptable distance in regards to injection site rotation on the same site, which is known to be one thumb (Figure 1).
A non-parametric Kruskal Wallis H test was performed to compare the effect of educational status, duration of insulin therapy, and duration of diabetes on patients’ level of knowledge for insulin storage and handling techniques. The test showed that the levels of education X2(3) =18.9, P<0.001; duration of insulin therapy X2(3) =11.7, P=0.008; and duration of diabetes X2(3) =10.7, P=0.014 had significantly effects on patients’ knowledge level. The Pairwise multiple comparisons at an adjusted alpha level of 0.0125 were used to compare all pairs of groups, and patients who achieved colleges and above (Median (Mdn) =12) had a higher median knowledge level than illiterates (Mdn=8), P<0.001. Patients who had been on insulin for 5-10 years (Mdn=10) had a better knowledge level as compared with who had been on for three months to one year (Mdn=8), P=0.001 (Table 2).
Table 2: Kruskal Wallis H test for predictor variables on the level of knowledge and practices on insulin storage and injection differences among respondents’ in northwest Ethiopia primary hospitals, Gondar, 2019 (N=166).
Variables
|
Knowledge score
|
Practice score
|
Median (IQR)
|
Test Statistics (X2), (df)
|
P-value
|
Median (IQR)
|
Test Statistics (X2), (df)
|
P-value
|
Educational status:
Illiterate
Read & write only 1ry and 2ndry education
College & above
|
8 (6-10)
9 (8-11)
9.5 (8-10)
12 (8-13)
|
18.89, (3)
|
<0.001**
|
28 (27-30)
28 (24-33)
31 (29.3-35)
32 (29.3-34)
|
25.86, (3)
|
<0.001**
|
Occupation: Farmer
Employer
Merchant
Housewife
Student
|
|
|
0.076
|
28 (27-31.3)
32 (28-35)
29 (25.8-31)
31 (28-34)
32 (31-34)
|
23.24, (4)
|
<0.001**
|
Years of insulin therapy:
0.25-1
>1-5
>5-10
>10
|
8 (6-9)
9 (7-10.5)
10 (8-12)
11 (8.5-12)
|
11.71, (3)
|
0.008**
|
28.5 (21.3-30.8)
31 (28.5-34)
29 (27-31)
28.5 (27-32.5)
|
15.85, (3)
|
0.001*
|
Years of disease: 0.25-1
>1-5
>5-10
>10
|
8 (6-10)
9 (7-10)
10 (8-12)
10.5 (7.8-12)
|
10.67, (3)
|
0.014*
|
30 (22-32)
31 (28-34)
30.5 (27.3-33.8)
28 (27-32.5)
|
8.55, (3)
|
0.036*
|
Knowledge level:
Adequate (>10.5)
Moderate (8-10.5)
Inadequate (<7)
|
//
|
//
|
//
|
29 (23-31)
30 (28-32.5)
32.5 (29-35)
|
19.26, (2)
|
<0.001**
|
* Statistically significant effect on patients’ knowledge and practice scores at p=0.05.
** With Pairwise multiple comparisons of Kruskal Wallis 1-way ANOVA (k-samples) there is a significant difference among the groups.
To determine the difference in knowledge level the differences of the patients’ overall knowledge levels between the two group predictor variables such as residency, ways to get insulin, and mocked injection technique during first training, we employed a Mann-Whitney U test. Significant difference in the patients’ overall knowledge level across different in residency (P=0.001), ways to get insulin (P<0.001) and mocked injection technique during first training (P=0.016) was noted. The median knowledge score level of urban dwellers (Mdn=10) was higher than those of rural dwellers (Mdn=8). Similarly, patients charged for insulin (Mdn=10) had a better knowledge level than those who obtain insulin for free (Mdn=8). Patients who mocked injection technique during first training (Mdn=9.5) had slightly better scores compared with those who did not (Mdn=8.5) (Table 3).
Table 3: The Mann-Whitney U test for the median scores of level of knowledge differences of between categories of predictor variables (N=166).
Variables
|
Overall knowledge score
|
Median (IQR)
|
Mann-Whitney U test
|
Z-score
|
P-value
|
Sex: Male
Female
|
|
|
|
0.291
|
Residence: Urban
Rural
|
10 (8-12)
8 (7-10)
|
2410
|
-3.36
|
0.001*
|
Getting of insulin: Free
Payment
|
8 (7-10)
10 (9-12)
|
1874
|
-5.01
|
<0.001*
|
Trained on insulin injection: Yes
No
|
|
|
|
0.075
|
Mocking the injection technique during first training: Yes
No
|
9.5 (8-12)
8.5 (6-10)
|
1959.5
|
-2.41
|
0.016*
|
* Statistically significant effect on patients’ knowledge levels at p<0.05.
Practice score
The participants’ insulin storage and injection practices were assessed by using the stated 14-item questionnaire. The median (IQR) practice level of study subjects was 31 (28-33.3) out of 56 (55.4%). Most of the participants (64.5%) had fair practice. Only 1.2% of them had good practice and the rest (34.3 %) portrayed poor practice. As illustrated in figure 2, the majority of the patients (73.3%) mixed the cloudy insulin Neutral Protamine Hagedorn (NPH) prior to use. One-third (33.8 %) of patients had ever injected their insulin through their clothes.
A Kruskal-Wallis H test identified the effects of different potential predictor variables on patients’ level of practice for insulin storage and handling techniques. Education X2(3) =25.9, P<0.001; occupation X2(4) =23.2, P<0.001; duration of insulin therapy X2(3) =15.9, P=0.001; duration of diabetes X2(3) =8.6, P=0.036; and patients knowledge levels X2(2) =19.3, P<0.001 were significantly alter the patients’ practice levels. As to the Pairwise multiple comparisons at an adjusted alpha level of 0.0125 for education levels, patients who completed primary and secondary educations (Mdn=31) had better practice levels than those who were not educated (Mdn=28), P=0.001. Those who achieved colleges and above (Mdn=32) exhibited higher practice levels than those who did not educated (P<0.001), and those who read and write only (Mdn=28), P=0.01. The Pairwise multiple comparisons at an adjusted alpha level of 0.01 for occupations indicated that students (Mdn=32), compared with farmers (Mdn=28), P=0.003 and merchants (Mdn=29), P=0.002, correctly demonstrate insulin handling techniques. The pairwise multiple comparisons at an adjusted alpha level of 0.017; patients who scored adequate knowledge level (Mdn=32.5) had better practice levels compared with those having moderate adequate (Mdn=30), P=0.008 and inadequate knowledge levels (Mdn=29), P<0.001 (Table 2). A Spearman’s correlation was run to determine the relationship between the patients’ knowledge level in their insulin handling techniques and their practice of insulin injections. The test showed as there was a moderate, positive correlation between the knowledge and practice level (rs =0.425, P<0.001).
Insulin self-administration skill assessment
Based on the observational checklists used to assess the patients’ skills related to insulin self-administration, a large numbers of participants (94.6%) correctly demonstrated the injection sites and about 70% of the participants properly indicated the pattern of injection site rotations. In contrast, about half of the respondents either performed incorrectly or skipped very critical and important steps such as shaking of cloudy NPH insulin, skin pinching and 45o injection skill, and drawing of insulin from the vials (Table 4).
Table 4: Observational checklist of patients’ skill related to self-insulin administration
Items
|
Correct
|
Incorrect
|
Skipped
|
N (%)
|
N (%)
|
N (%)
|
Showed injection sites
|
160 (94.6)
|
4 (2.4)
|
2 (1.2)
|
Showed injection site rotations
|
116 (69.9)
|
46 (27.7)
|
4 (2.4)
|
Showed how to shake NPH
|
92 (55.4)
|
44 (26.5)
|
30 (18.1)
|
Showed how to pinch (fold) skin and inject with (45o)
|
108 (65.1)
|
54 (32.5)
|
4 (2.4)
|
Showed how to draw insulin from the vial
|
86 (51.8)
|
49 (29.5)
|
31 (18.7)
|