Factors associated with non-adherence during tuberculosis treatment among patients treated with DOTS strategy in Jayapura, Papua Province, Indonesia

Figures & data

Figure 1. Sampling process.

Table 1. Treatment history, demographic and household characteristics associated with adherence during tuberculosis treatment among tuberculosis patients from Jayapura City.

	Cases (poor adherence) (n,%) (N = 81)	Controls (good adherence) (n,%) (N = 183)	p-value^§
Basic & treatment characteristics
Age in years, median (25th–75th percentile)	26 (22–35)	32 (24–40)	0.009^#
Female sex	38 (46.9)	71 (38.8)	0.22
Sputum acid-fast bacilli positive at treatment initiation	54 (66.7)	127 (69.4)	0.70
Treatment outcome			<0.001
Cured	12 (14.8)	132 (72.1)
Treatment completed	1 (1.2)	51 (27.9)
Lost to follow up	68 (84.0)	0 (0.0)
Findings from interview
Place of interview			<0.001
In health center	32 (39.5)	121 (66.1)
In respondent’s home	49 (60.5)	62 (33.9)
General health
Karnofsky score <90%* at interview	13 (16.0)	8 (3.3)	<0.001
Having TB-symptoms at interview^£	54 (66.7)	80 (43.7)	0.001
At-risk for alcoholism^†	12 (14.8)	10 (5.5)	0.022
Demographics
Papuan ethnicity	60 (74.1)	105 (57.4)	0.01
No or low-level education only‡	12 (14.8)	34 (18.6)	0.57
Job type before tuberculosis			0.54
Unemployed	7 (8.6)	15 (8.2)
Housewife	22 (27.2)	45 (24.6)
Civil servant	4 (4.9)	18 (9.8)
Self-employed, trading	16 (19.8)	42 (23.0)
Student	17 (21.0)	25 (13.7)
Other (incl. farmers, fishermen)	15 (18.5)	38 (2.8)
Household factors
Household income, median (25^th–75^th percentile) in (×10^6IDR)	1.2 (0.9–2.0)	1.5 (1.0–2.0)	0.42^#
>1 Persons per 10m^2 in house	17 (21.0)	49 (26.8)	0.14
Household member ever treated for TB	29 (35.8)	37 (20.2)	0.01

^§p-values were calculated using chi-square test for proportions except the continuous variable using ^#Mann-Whitney equality-of-populations rank test to test differences between medians.

*Karnofsky score to describe performance status (Oxford Textbook of Palliative Medicine, Oxford University Press, 1993). A score of 90 means has minor signs of disease, less than 90% shows more

signs of disease and unable to carry normal activity.

^£The following TB symptoms were reported most frequently: coughing that last three or more weeks, coughing of blood, chest pain, pain with breathing or coughing, unintentional weight loss, fatigue, fever, night sweats, chills, loss of appetite

^†At-risk for alcoholism was defined as a TWEAK-score of >2 [17].

^‡Low-level education defined as no education or did not finish primary school.

IDR, Indonesian Rupiah

Table 2. Experience, perception, and knowledge about tuberculosis associated with adherence status among tuberculosis patients.

	Cases (poor adherence) (n,%) (N = 81)	Controls (good adherence) (n,%) (N = 183)	p-value^§
Treatment experience
Has not received education about TB from the TB nurse	12 (14.8)	8 (4.4)	0.007
Receiving loose drug rather than fix drug combination	12 (14.8)	7 (3.8)	0.001
Reports to have a treatment observer	57 (70.4)	138 (75.4)	0.48
Experiencing side effects during TB treatment*	33 (40.7)	45 (24.6)	0.012
Mobility and access
Moved residence in the last year	9 (11.1)	3 (1.6)	0.001
Travel time >15 minutes to the public health center	15 (18.5)	32 (17.5)	0.97
Perceives that transport to health center was difficult due to distance or costs	10 (12.3)	2 (1.1)	<0.001
Retained job during TB treatment	60 (74.1)	148 (80.9)	0.28
Perceives that costs of TB treatment are too high	23 (28.4)	29 (15.8)	0.02
Patient’s knowledge about TB
Does not know what causes TB	63 (77.8)	95 (51.9)	<0.001
Does not know how TB is transmitted	41 (50.6)	34 (18.6)	<0.001
Thinks that TB cannot be cured	9 (11.1)	0 (0.0)	<0.001
Unaware about the duration of a full TB treatment course	10 (12.3)	4 (2.2)	0.001
Unaware about consequences of incomplete TB treatment‡	15 (19.0)	6 (3.3)	<0.001

^§p-values were calculated using chi-square tests for proportions.

*This concerns the following symptoms: feeling sick or dizzy, skin rashes, pin and needles, flu like symptoms.

‡ Two cases did not answer this question, and therefore, the percentages were calculated on a total of 79 cases and 183 controls (total population: 262).

Table 3. Univariable and multivariable odds ratios (OR) and 95% confidence intervals (95% CI) for factors associated with non-adherence to tuberculosis (TB) treatment^‡.

	Univariable				Multivariable^‡
Risk factors	OR	95% CI			aOR*	95% CI
Individual characteristic
Age < 35yrs	2.2	1.2	-	3.9	3.1	1.5	-	6.6
Papuan ethnicity	2.1	1.2	-	3.8	NS
At risk for alcoholism	3.0	1.2	-	7.3	NS
Family characteristics
>1 person per 10m^2 in household	2.0	1.1	-	3.4	NS
Having a household member treated for TB	2.2	1.2	-	3.9	2.5	1.2	-	5.2
Access and mobility
Moved residence in the last year	7.5	2.0	-	28	9.6	2.2	-	42
Distance/travel costs perceived as problem	12.7	2.7	-	60	10.2	1.7	-	60
Finds costs of TB treatment too high	2.1	1.1	-	3.9	NS
Treatment experience
Received loose drugs (instead of fixed dose combination)	4.4	1.7	-	12	3.8	1.2	-	12
Having side effect from TB treatment	2.1	1.2	-	3.7	NS
Has not received education about TB	3.8	1.5	-	9.7	3.5	1.0	-	12
Patient knowledge
Does not know the cause of TB	3.2	1.8	-	5.9	2.4	1.1	-	5.4
Does not know how TB is transmitted	4.5	2.5	-	8.0	3.8	1.8	-	7.8
Thinks that TB cannot be cured	3.5	2.9	-	4.3
Unaware about the duration of a full TB treatment course	6.3	1.9	-	21	NS
Unaware about the consequences of incomplete TB treatment	6.9	2.6	-	19	10.3	3.1	-	34

^‡NS: not statistically significant in multivariable model (applying backward stepwise logistic regression; retaining all variables with a p-value of <0.05 in the model)

*Adjusted for all other variables with an adjusted OR in this table.

^‡The total subject analyzed for the final model were 262 (79/81 cases, and 183/183 controls).

Chan A, Pristach E, Welte J, et al. Use of TWEAK test in screening for alcoholism/heavy drinking in three populations. Alcohol Clin Exp Res. 1993;17:1188–1192.

 PubMed Web of Science ®Google Scholar