Factors associated with tuberculosis contact tracing from the perspective of patients: a cross-sectional survey across three districts in Ghana

Abstract

Background

In Ghana, tuberculosis (TB) case detection remains suboptimal, with only 29–34% of estimated TB cases detected, well below the African regional average of 47% and the World Health Organization (WHO) target of 70%. Contact tracing is an important strategy for identifying exposed contacts and potentially undiagnosed TB cases, thereby supporting early diagnosis and interruption of transmission. This study examined patient-reported factors associated with successful TB contact tracing in three selected districts across three regions in Ghana.

Methods

This study used an analytical cross-sectional study design. A systematic sampling technique was used to recruit TB patients aged 18 years and above from TB treatment registers. Data was collected using a structured interviewer-administered questionnaire by trained research assistants. Descriptive statistics were computed, and logistic regression analysis was used to determine factors associated with contact tracing from the patient’s perspective. Variables with log-likelihood ratio p-value ≤ 0.2 in bivariable analysis were included in multivariable models. Statistical significance was set at p < 0.05.

Results

A total of 316 tuberculosis patients were included, with a median age of 45 years (IQR: 35–56), and 29.8% aged ≥ 55 years. Overall, 70.3% (222/316) of patients had their contacts successfully traced, while 29.7% (94/316) did not have their contacts traced. Factors independently associated with successful contact tracing included having at least secondary education (aOR = 4.04; 95% CI: 1.78–9.20; p = 0.001), perceived influence of social stigma (aOR = 3.14; 95% CI: 1.84–5.34; p < 0.001), willingness to participate in contact tracing (aOR = 8.14; 95% CI: 1.55–42.74; p = 0.013), perceived access to a health facility (aOR = 4.01; 95% CI: 1.32–12.17; p = 0.014) and perceived religious beliefs (aOR = 5.47; 95% CI: 3.06–9.76; p < 0.001). Conversely, patients residing in Ho Municipality were significantly less likely to have their contacts traced compared to those in the Achiase District (aOR = 0.04; 95% CI: 0.00–0.67; p = 0.025). Given the perception-based measurement of several explanatory variables, these findings should be interpreted cautiously.

Conclusion

TB contact tracing in the study setting was moderately high but remains suboptimal. The findings suggest that successful contact tracing is associated with educational level, district of residence, willingness to participate, and patient-reported perceptions of social stigma, health facility access and religious beliefs. Improving TB contact tracing requires a comprehensive, patient-centered approach to address implementation gaps.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12879-026-13597-9.

Background

Tuberculosis (TB) remains one of the leading causes of infectious disease-related mortality globally, with significant impacts on public health and socioeconomic development [1]. TB, a communicable bacterial disease caused by Mycobacterium tuberculosis that primarily affects the lungs, is transmitted via airborne droplets and is treatable with appropriate antimicrobial therapy [2]. Although TB is preventable and curable, transmission persists when infectious cases are not detected early and when exposed contacts are not identified, evaluated, and managed in a timely manner [1, 2]. Contact tracing is therefore a key component of TB control because it enables health systems to identify individuals who have been exposed to infectious TB patients and link them to screening, diagnosis, preventive therapy, or treatment where appropriate [3, 4].

In 2024, an estimated 10.7 million people fell ill with TB worldwide, and a total of 1.23 million people died from it [5]. The African region accounts for over 33% of global TB deaths, with co-infections of TB and HIV being particularly prevalent [2]. In 2024, the estimated TB incidence in Ghana was 126 cases per 100,000 population and TB deaths in HIV-negative people was 9,300 [6]. While mortality rates have slightly reduced by 29%, it remains a major public health challenge [6].

In Ghana, TB case detection remains below global targets, suggesting that a substantial proportion of cases may remain undiagnosed within communities [7, 8]. Contact tracing can help address this gap by identifying exposed household and close contacts who may have active TB disease or latent TB infection [3, 4]. However, the effectiveness of contact tracing depends not only on the availability of health services but also on the cooperation of TB patients, their willingness to disclose contacts, and the ability of health workers to follow up disclosed contacts [3, 9, 10]. To reduce transmission and contain the disease, the World Health Organization (WHO) has recommended contact tracing as an effective approach to identify, diagnose and treat individuals exposed to active TB patients [11]. However, despite the effectiveness of implementing this initiative, reports suggest that TB transmission particularly remains in some selected districts in Ghana [12].

Evidence from previous studies suggests that several patient- and system-level factors may influence TB contact tracing. Patient-related factors include knowledge of TB transmission, willingness to disclose contacts, fear of stigma, beliefs about TB, and trust in health workers [3, 13–16]. Sociocultural influences, including family support, community attitudes, cultural beliefs, and religious beliefs, may also shape how patients engage with contact tracing activities and TB-related health services [3, 17]. At the health-system level, provider communication, accessibility of health facilities, availability of diagnostic services, and delays in follow-up may affect whether disclosed contacts are successfully traced [16, 18, 19].

In Ghana, only 29–34% of TB cases have been detected, well below the African regional average of 47% and the WHO target of 70% [9, 20]. Existing studies have documented barriers to TB diagnosis and treatment, including stigma, misconceptions, economic constraints, access barriers, and health-system limitations [10, 19, 21]. However, less is known about patient-reported factors specifically associated with the successful tracing of contacts after TB patients have disclosed them. Understanding these factors is important for strengthening patient-centred and context-specific contact tracing strategies. This study therefore examined patient-reported socio-demographic, individual-level, health-system, and sociocultural factors associated with successful TB contact tracing among TB patients in three selected districts across three regions in Ghana.

Methods

Study design and setting

An analytical cross-sectional study was conducted to examine factors associated with contact tracing from the perspective of TB patients. The study was carried out between January-April 2025 in TB diagnostic and treatment facilities in three districts in Ghana: Bole District in the Savannah Region, Ho Municipality in the Volta Region, and Achiase District in the Eastern Region. These districts were selected to provide representation across different geographic and health service contexts.

Bole District has an estimated population of 115,800 and is served by 53 health facilities, including Community-based Health Planning and Services (CHPS) compounds. The study was conducted at the district hospital which is the main TB diagnostic and treatment centre in the district. Ho Municipal, the capital of the Volta Region, has an estimated population of 180,420 and is served by 58 health facilities. The study was conducted at the municipal hospital which is one of the TB diagnostic and treatment centres in the municipality. Achiase District has an estimated population of 56,348 and is served by 25 health facilities. The study was conducted at the district hospital which is the main TB diagnostic and treatment centre. In total, three health facilities were selected, one from each district because they provide TB diagnostic and treatment services under the National Tuberculosis Control Programme and serve as key TB service delivery points in their respective districts.

Study population and eligibility criteria

The study population comprised adult TB patients aged 18 years and above who had received TB treatment in the selected districts. Eligible participants were individuals who had received TB treatment during the period January 2023 to March 2025 in Bole, Ho, and Achiase. Patients who were critically ill, had severe conditions that prevented them to participate, or declined to provide informed consent were excluded.

Sample size determination

The sample size was calculated using OpenEpi version 3, based on a national TB prevalence of 20.7% for Ghana, as reported by the Country Coordinating Mechanism (CCM) of the Global Fund in its national TB epidemiological profile [22]. A finite population of 330 TB patients across the three districts, a 95% confidence level (Z = 1.96), a precision of 5%, and a design effect of 2 were applied. Using these parameters, the minimum required sample size was 287.

To account for a potential non-response rate of 10%, the sample size was adjusted to 316 participants. The use of TB prevalence in estimating the sample size was due to the limited availability of reliable prior estimates of successful TB contact tracing from the patient perspective in Ghana at the time of study design. With the use of this estimate, the sample was therefore considered adequate for estimating contact tracing-related outcomes among the finite population of registered TB patients in the selected districts.

Using probability proportional to size, the sample was allocated across the three districts as follows: 105 participants from Bole District, 154 from Ho Municipal, and 57 from Achiase District.

Sampling procedure

A systematic sampling technique was employed using TB treatment registers at the selected TB diagnostic and treatment facility in each district as the sampling frame. Since only one facility was selected per district, no facility-level sample allocation or merging of multiple facility registers was required. For each district, the total number of eligible TB patients was divided by the required sample size to obtain a sampling interval. A random starting point within the interval was selected, after which every nth patient was recruited. Selected patients were contacted through the selected facility using available contact information in the TB treatment register or during routine clinic visits. When a selected patient was unavailable after reasonable contact attempts, could not be reached, was ineligible, or declined participation, the next eligible patient on the register was approached. This replacement procedure was used to achieve the required sample size while maintaining the systematic approach to recruitment. This approach was applied independently in Bole District, Ho Municipality, and Achiase District, using the district-specific sample allocations of 105, 154, and 57 participants, respectively.

Data collection

Data was collected using a structured questionnaire developed from similar studies [23, 24] and also in consultation with TB coordinators in the study districts to align with the study objectives and local TB control context. The instrument consisted of four sections: socio-demographic characteristics; individual-level factors; health system factors; and sociocultural influences. The questionnaire was pre-tested among TB patients in the Hohoe municipality, and revisions were made to improve clarity, internal consistency and contextual appropriateness. Trained public health officers conducted face-to-face interviews in private settings at the health facilities or where appropriate, at participants’ residences. Responses were captured electronically using the Kobo Collect tool, with daily checks for completeness and consistency. Written informed consent was obtained prior to data collection.

Several explanatory variables, including perceived stigma, perceived personal health beliefs, perceived community attitudes, perceived cultural beliefs, and perceived religious beliefs, were measured as patient-reported perceptions rather than as standardized psychometric constructs. During interviewer-administered data collection, trained public health officers explained each item to ensure respondents understood the distinction between related sociocultural domains. Cultural beliefs referred to traditional or customary beliefs about TB and its management; community attitudes referred to perceived reactions, acceptance, or support from people in the respondent’s community; and religious beliefs referred to faith-related beliefs or practices that respondents felt influenced participation in TB contact tracing.

Outcome variable

The primary outcome variable was patient-reported successful TB contact tracing. This was defined as the participant reporting that at least one disclosed contact had been traced by health workers. The outcome was coded as binary: 0 = no contact successfully traced and 1 = at least one contact successfully traced. The outcome was measured from the patient’s perspective and was based on participant report rather than independent verification from contact tracing registers.

Explanatory variables

Independent variables were grouped into four domains:

Socio-demographic characteristics: age, sex, education, occupation, marital status, income, and district of residence.

Individual-level factors: willingness to participate in contact tracing, disclosure practices, perceived influence of social stigma, and perceived influence of personal health beliefs. Willingness to participate was measured as whether the respondent reported being willing to participate in contact tracing. Perceived influence of social stigma was measured as whether the respondent reported that stigma influenced contact tracing. Perceived personal health beliefs were measured as whether the respondent reported that personal health beliefs influenced contact tracing.

Health system factors: perceived accessibility of health facilities, being asked about contacts by health workers, perceived provider explanation of contact tracing, perceived supportive attitude of healthcare workers and perceived availability of diagnostic services. Perceived accessibility of health facilities was measured as whether the respondent reported that the health facility was accessible to them. perceived provider explanation of contact tracing was measured as whether the respondent reported that healthcare workers explained contact tracing to them. Perceived availability of diagnostic services was measured as whether the respondent reported that diagnostic services were readily available.

Sociocultural factors: perceived family support during TB treatment, perceived influence of community attitudes on cooperation, perceived influence of cultural beliefs on participation, and perceived influence of religious beliefs on participation. Perceived family support during TB treatment was measured as whether the respondent reported they had family support during TB treatment, perceived influence of community attitudes on cooperation was measured as whether the respondent reported they received support or acceptance from their community, perceived influence of cultural beliefs was measured by whether a respondent reported that traditional or customary beliefs influenced their participation in contact tracing and perceived influence of religious beliefs was measured as whether respondents reported that religious beliefs or faith-related practices influenced their participation in contact tracing.

All these variables were measured using binary response options coded as “No” or “Yes.” These measures were intended to capture respondents’ perceptions of factors that influenced contact tracing, rather than the intensity, multidimensional structure, or objective presence of the underlying constructs.

Data analysis

Data was exported into STATA version 17 for cleaning and analysis. Descriptive statistics (frequencies, percentages, median, interquartile range) summarized respondent characteristics. Logistic regression was used to assess associations between each independent variable and successful contact tracing. At the bivariable analysis level, variables with log-likelihood p-value ≤ 0.2 were included in the multivariable model. This was done to avoid excluding variables that may not be significant in univariable analysis but could become significant or act as confounders once other variables were controlled for in the multivariable model [25]. Before fitting the final multivariable logistic regression model, multicollinearity among explanatory variables was assessed using variance inflation factors (VIF) and tolerance values. A VIF value below 10 and a tolerance value above 0.10 were considered to indicate the absence of serious multicollinearity [26]. At the multivariable level, a p-value of < 0.05 was considered significant at a 95% confidence interval.

Results

Table 1 presents the socio-demographic characteristics of the study respondents. The study was conducted among 316 TB patients. The majority of them 94 (29.8%), were aged above 55 years with a median age of 45 years (IQR = 35–56). The study found that 237 (75%) of the respondents were male, 216 (68%) were married, 263 (83%) were Christians, and 101 (32%) had up to senior high school education.

Table 1.

Sociodemographic characteristics of TB patients

Variable	Frequency (N = 316)	Percentage (%)
Median age (IQR)	45 (35–56)
Age in years
< 24	21	6.7
25–34	56	17.7
35–44	77	24.4
45–54	68	21.5
> 55	94	29.8
Sex
Female	79	25.0
Male	237	75.0
Marital status
Single	58	18.4
Married	216	68.4
Divorced/Separated	30	9.5
Widowed	12	3.8
Religion
Christianity	263	83.2
Islamic	51	16.1
Traditionalist	2	0.6
Level of education
No formal education	84	26.6
Primary education	33	10.4
Junior high education	60	19.0
Senior high education/ Vocational training	101	32.0
Tertiary education	38	12.0
Occupation
Unemployed	80	25.3
Informal worker	157	49.7
Formal worker (Private)	27	8.5
Formal worker (Government)	35	11.1
Others	17	5.4
District
Achiase	57	18.0
Bole	105	33.2
Ho Municipal	154	48.7

Table 2 presents the contact tracing practices as described by the study respondents. About one third (36.1% and 34.2%) of participants disclosed 1–2 and 3–4 contacts respectively to a healthcare worker. However, 40 (12.7%) did not disclose any contact to a healthcare worker.

Table 2.

Contact tracing activities reported by TB patients

Variable	Frequency (N = 316)	Percentage (%)
Contacts disclosed
0	40	12.7
1–2	114	36.1
3–4	108	34.2
≥ 5	54	17.1
Contacts successfully traced
0	94	29.7
1–2	99	31.3
3–4	86	27.2
≥ 5	37	11.7
Contacts tested for TB
0	99	31.3
1–2	113	35.8
3–4	75	23.7
≥ 5	29	9.2

Figure 1 presents the overall contact tracing as reported by TB patients. The results revealed that 222 (70.3%) of the TB patients had their contacts successfully traced, while 94 (29.7%) contacts were not traced.

Fig. 1.

Overall contact tracing reported by TB patients in three selected districts

Figure 2 presents the distribution of the overall contacts tracing by the various study districts. The study found that 127 (57.2%) who were successfully traced were in the Ho Municipality while Bole had the least 38 (17.1%) contacts of TB patients traced.

Fig. 2.

Contact tracing reported by TB patients by district

Table 3 shows individual-level factors disclosed by TB patients on contact tracing. The results revealed that, 303 (95.9%) of the respondents were comfortable with discussing their TB status. Similarly, 311 (98.4%) of the respondents had disclosed their contacts while majority 306 (96.8%) of the respondents revealed that their personal health belief did not affect contact tracing.

Table 3.

Individual-level factors disclosed by TB patients on contact tracing

Variable	Frequency (N = 316)	Percentage (%)
Comfortably discussed TB status
No	13	4.1
Yes	303	95.9
Perceived social stigma affected contact tracing
No	86	27.2
Yes	230	72.8
Perceived personal health beliefs affected contact tracing
No	306	96.8
Yes	10	3.2

Table 4 shows health system related factors on contact tracing from the perspective of TB patients. A majority of respondents, 293 (92.7%) disclosed that health facilities were accessible to them. In addition, 281 (88.9%) reported that they were asked about their contacts at the health facility, and majority 274 (86.7%) reported that healthcare workers explained contact tracing to them. Furthermore, 286 (90.5%) disclosed that diagnostic services were available.

Table 4.

Health system factors disclosed by TB patients on contact tracing

Variable	Frequency (N = 316)	Percentage (%)
Perceived health facility accessibility
No	23	7.3
Yes	293	92.7
Asked about contacts
No	35	11.1
Yes	281	88.9
Perceived provider explanation of contact tracing
No	42	13.3
Yes	274	86.7
Healthcare workers’ attitude supportive
No	5	1.6
Yes	311	98.4
Perceived diagnostic services readily available
No	30	9.5
Yes	286	90.5

Table 5 shows sociocultural factors reported by TB patients on contact tracing. Majority 311 (98.4%) of the respondents disclosed perceived family support during TB treatment and almost all 313 (99.1) perceived community attitudes did not influence their cooperation with the contact tracing process. Also, 315 (99.5%) revealed that perceived cultural beliefs did not influence their participation in the contact tracing process, however, 144 (45.6%) indicated perceived religious beliefs influenced their participation in contact tracing.

Table 5.

Sociocultural factors disclosed by TB patients on contact tracing

Variable	Frequency (N = 316)	Percentage (%)
Perceived family support for TB treatment
No	5	1.6
Yes	311	98.4
Perceived community attitudes influenced cooperation
No	313	99.1
Yes	3	1.0
Perceived cultural beliefs influenced participation
No	315	99.7
Yes	1	0.3
Perceived religious beliefs influenced participation
No	172	54.4
Yes	144	45.6

Factors associated with TB contact tracing from the perspective of TB patients

Before fitting the multivariable logistic regression model to determine factors associated with TB contact tracing from the perspective of the TB patient, multicollinearity diagnostics were performed among the explanatory variables. The results showed no evidence of serious collinearity. All VIF values were below 10, and all tolerance values were above 0.10. The mean VIF was 1.67 for socio-demographic variables, 1.00 for individual-level variables, and 2.09 for health system-related variables. The highest VIF values were observed for healthcare workers explaining contact tracing (VIF = 4.23; tolerance = 0.2361) and being asked about contacts (VIF = 4.18; tolerance = 0.2394), but these remained within acceptable limits. Therefore, multicollinearity was not considered a concern in the analysis (Supplementary Table 1).

As shown in Table 6, the study found that having up to secondary education, living in the Ho or Bole district influenced TB contact tracing. The study found that TB patients who had up to secondary education were four times more likely to have their contacts traced (aOR = 4.04; [95%CI = 1.78–9.20], p-value = 0.001) compared to patients who had no formal education. Patients who lived in the Bole district and Ho Municipality were 99% (aOR = 0.01; [95%CI = 0.00-0.10], p-value < 0.001) and 96% less likely (aOR = 0.04; [95%CI = 0.00-0.67], p-value = 0.025) respectively to have their contacts traced compared to participants in the Achiase District. However, these should be interpreted cautiously as the district variable showed evidence of sparse data and possible quasi-complete separation because all respondents from Achiase District reported successful contact tracing. The odds ratios maybe unstable.

Table 6.

Socio-demographic characteristics, individual level, health system related and sociocultural factors associated with contact tracing

Variable	Contact tracing		cOR (95% CI), p-value	aOR (95% CI), p-value
	Not traced N = 94 (%)	Traced N = 222 (%)
Age in years			0.247*
< 24	5 (5.32)	16 (7.21)	Ref.
25–34	19 (20.21)	37 (16.67)	0.64 (0.21–1.94), 0.432
35–44	27 (28.72)	50 (22.52)	0.61 (0.21–1.79), 0.369
45–54	23 (24.47)	45 (20.27)	0.65 (0.22–1.91), 0.429
> 55	20 (21.28)	74 (33.33)	1.21 (0.41–3.57), 0.728
Sex			0.129 *
Female	18 (19.15)	61 (27.48)	Ref.	Ref
Male	76 (80.85)	161 (72.52)	0.64 (0.35–1.14), 0.129	0.85 (0.42–1.75), 0.668
Marital status			0.833*
Single	16 (17.02)	42 (18.92)	Ref.
Married	67 (71.28)	149 (67.12)	0.86 (0.45–1.63), 0.642
Divorced/Separated	7 (7.45)	23 (10.36)	1.22 (0.45–3.30), 0.701
Widowed	4 (4.26)	8 (3.60)	0.73 (0.20–2.63), 0.634
Religion			0.025 *
Christianity	70(74.47)	193(83.23)	Ref.	Ref
Islamic	23(24.47)	28(12.61)	0.44 (0.24–0.81), 0.009	1.95 (0.88–4.31), 0.101
Traditionalist	1(1.06)	1(0.45)	0.36 (0.04–3.56), 0.385	0.05 (0.00-6.98), 0.230
Level of education			< 0.001 *
No formal education	48 (51.06)	36 (16.22)	Ref.	Ref
Primary education	9 (9.57)	24 (10.81)	3.43 (1.45–8.13), 0.005	1.60 (0.53–4.84), 0.402
Junior high education	14 (14.89)	46 (20.72)	4.26 (2.05–8.84), < 0.001	1.62 (0.66–3.96), 0.295
Senior high education	13 (13.83)	88 (39.64)	8.71 (4.26–17.82), < 0.001	4.04 (1.78–9.20), 0.001
Tertiary education	10 (10.64)	28 (12.61)	3.61 (1.58–8.25), 0.002	1.19(0.39–3.57), 0.754
Occupation			0.003 *
Unemployed	36 (38.30)	44 (19.82)	Ref.
Informal worker	37 (39.36)	120 (54.05)	2.64 (1.49–4.66), 0.001	1.41(0.69–2.89), 0.345
Formal worker	21 (22.34)	58 (26.13)	2.23 (1.15–4.32), 0.017	1.19(0.48–2.92), 0.709
District			< 0.001 *
Achiase	0 (0.00)	57 (25.68)	Ref.	Ref
Bole	67 (71.28)	38 (17.12)	0.005(0.0003–0.083), < 0.001	0.01 (0.00-0.10), < 0.001
Ho Municipal	27 (28.72)	127 (57.21)	0.04 (0.002–0.67), 0.025	0.04 (0.00-0.67), 0.025
Individual level factors associated with contact tracing
Comfortably discussed TB status			0.582*
No	3 (3.19)	10 (4.50)	Ref.
Yes	91 (96.81)	212 (95.50)	0.70 (0.19–2.60), 0.593
Perceived social stigma affected contact tracing			< 0.001 *
No	41 (43.62)	45 (20.27)	Ref.	Ref
Yes	53 (56.38)	177 (79.73)	3.04 (1.80–5.13), < 0.001	3.14 (1.84–5.34), < 0.001
Willing to participate in contact tracing			0.007 *
No	6 (6.38)	2 (0.90)	Ref.	Ref
Yes	88 (93.62)	220 (99.10)	7.50 (1.49–37.87), 0.015	8.14 (1.55–42.74), 0.013
Perceived personal health beliefs affected contact tracing			0.477*
No	92 (97.87)	214 (96.40)	Ref.
Yes	2 (2.13)	8 (3.60)	1.72 (0.36–8.25), 0.498
Health system factors associated with contact tracing
Perceived health facility accessibility			< 0.001 *
No	16 (17.02)	7 (3.15)	Ref.	Ref
Yes	78 (82.98)	215 (96.85)	6.30 (2.50–15.89), < 0.001	4.01 (1.32–12.17), 0.014
Asked about contacts			< 0.001 *
No	34 (36.17)	1 (0.45)	Ref.
Yes	60 (63.83)	221 (99.55)	125.2 (16.80–933.67), < 0.001	9.48 (0.79–113.10), 0.075
Perceived provider explanation of contact tracing			< 0.001 *
No	40 (42.55)	2 (0.90)	Ref.	Ref
Yes	54 (57.45)	220 (99.10)	81.48 (19.09–347.71), < 0.001	17.37 (2.94–102.83), 0.002
Supportive healthcare workers attitude			0.623*
No	2 (2.13)	3 (1.35)	Ref.
Yes	92 (97.87)	219 (98.65)	1.59 (0.26–9.66), 0.616
Perceived readily available diagnostic services			< 0.001 *
No	22 (23.40)	8 (3.60)	Ref.	Ref
Yes	72 (76.60)	214 (96.40)	8.17 (3.49–19.16), < 0.001	3.57 (1.23–10.38), 0.02
Socio-cultural factors associated contact tracing
Perceived family support for TB treatment			0.146 *
No	3 (3.19)	2 (0.90)	Ref.	Ref
Yes	91 (96.81)	220 (99.10)	3.37 (0.65–17.41), 0.146	2.75 (0.41–18.58), 0.300
Perceived community attitude influenced cooperation			0.738*
No	93 (98.94)	220 (99.10)	Ref.
Yes	1 (1.06)	2 (0.90)	0.71 (0.09–5.43), 0.739
Perceived cultural beliefs influenced participation			0.880*
No	94 (100.00)	221 (99.55)	Ref.
Yes	0 (0.00)	1 (0.45)	1.28 (0.05–31.70), 0.880
Perceived religious beliefs influenced participation			< 0.001 *
No	76 (80.85)	96 (43.24)	Ref.	Ref
Yes	18 (19.15)	126 (56.76)	5.42 (3.06–9.61), < 0.001	5.47 (3.06–9.76), < 0.001

(*) = log likelihood probability value; significant at < 0.2

With regards to individual level factors, the study results revealed that TB patients who perceived that social stigma affects contact tracing were three times more likely to have their contacts traced (aOR = 3.14; [95%CI = 1.84–5.34], p-value < 0.001) compared to patients who did not perceive social stigma affects contact tracing. Also, patients who were willing to participate in contact tracing were eight times more likely to have their contacts traced (aOR = 8.14; [1.55–42.74], p-value = 0.013) compared to patients who were not willing to participate in contact tracing.

Concerning health-system related factors, the study results revealed that patients with perceived access to a health facility were four times more likely to have their contacts traced (aOR = 4.01; [95%CI = 1.32–12.17], p-value = 0.014) compared to patients with no perceived access to a health facility. Also, patients who said contact tracing was explained to them by a healthcare worker were seventeen times more likely to have their contacts traced (aOR = 17.37; [95%CI = 2.94–102.83], p-value = 0.002) compared to patients who did not get explanation on contact tracing by a healthcare worker. Again, patients who perceived diagnostic services were available at the health facility were three times more likely to have their contacts traced (aOR = 3.57; [95%CI = 1.23–10.38], p-value = 0.02) compared to those who did not perceive diagnostic services were available in their health facility.

Regarding sociocultural factors, the study found that patients who perceived that religious beliefs affect participation in contact tracing were three times more likely to have their contacts traced (aOR = 5.47; [3.06–9.76], p-value < 0.001) compared to patients who did not perceive religious beliefs influence contact tracing.

Discussion

This study examined patient-reported factors associated with successful TB contact tracing in three selected districts in Ghana. Successful contact tracing was reported by 70.3% of respondents. Several factors, including educational level, district of residence, willingness to participate in contact tracing, perceived influence of stigma, perceived access to health facilities, perceived provider explanation of contact tracing, perceived availability of diagnostic services, and perceived influence of religious beliefs, were associated with patient-reported successful contact tracing. However, these findings should be interpreted as exploratory rather than causal, because several explanatory variables were measured as patient-reported perceptions using binary items, and some adjusted estimates showed evidence of imprecision.

Socio-demographic and individual characteristics associated with contact tracing

Educational attainment was significantly associated with contact tracing success, with respondents who had secondary education more likely to have contacts traced compared to those without formal education. This is consistent with evidence that higher education correlates with improved health literacy, better understanding of TB transmission and greater engagement with health systems, which collectively facilitate participation in contact investigation activities [3, 27]. Low literacy and inadequate knowledge have been identified as barriers to TB case finding and contact engagement in Sub-Saharan Africa, reinforcing the importance of tailored education interventions in low-literacy populations [28, 29]. These findings suggest that TB programs should adopt simplified, and low-literacy communication strategies, including visual aids and local-language counselling, to improve participation among less educated populations.

Variation in patient-reported successful contact tracing was observed across districts, but this finding should be interpreted cautiously because of the possibility of sparse-data bias or quasi-complete separation, which can produce unstable odds ratio estimates Therefore, the lower odds observed in Ho Municipality and Bole District compared with Achiase District should be viewed as exploratory rather than as precise evidence of district-level programme differences. Further studies using larger samples and verified programme data are needed to determine whether true district-level differences exist and what health system or implementation factors may explain them.

Our study found a positive association between perceived influence of social stigma and successful contact tracing but this finding should be interpreted with caution. This finding does not imply that stigma improves contact tracing. In this study, stigma was measured as a binary patient-reported perception and not with a validated multidimensional stigma scale. Several explanations may account for this counterintuitive association. Patients who perceived stigma may have received more counselling or follow-up from health workers; they may have been more aware of the social consequences of TB and therefore more likely to report stigma as influential; or the association may reflect residual confounding, reverse or measurement error. Therefore, this finding should be viewed as hypothesis-generating and should be examined in future studies using validated stigma scales and qualitative methods.

Similarly, willingness to participate in contact tracing was strongly associated with contact tracing success. This aligns with findings from other studies demonstrating that patient motivation, commitment and engagement with health education significantly enhance uptake of contact investigation and related preventive services in low-resource settings [30]. Effective contact tracing depends not only on biomedical infrastructure but also on patients’ active cooperation, underscoring the value of incorporating motivational counselling and patient-centred communication into routine TB care. TB programs should integrate brief motivational counselling at diagnosis, emphasize confidentiality and benefits of contact tracing, and build trust between patients and providers. Strengthening patient motivation is likely to yield substantial improvements in contact tracing outcomes even in resource-constrained settings.

Health system factors associated with contact tracing

Health system-related perceptions were associated with contact tracing outcomes. Respondents who reported that health facilities were accessible to them were more likely to have their contacts traced. However, because accessibility was measured through patient self-report rather than objective measures such as distance, travel time or transport cost, this finding should be interpreted as reflecting perceived access to care. Similarly, reported availability of diagnostic services reflects respondents’ perceptions rather than an independent assessment of diagnostic capacity. These findings suggest that patients’ perceptions of access and service availability may influence their engagement with contact tracing processes, but further studies using objective health facility and geographic measures are needed to clarify the specific mechanisms.

Respondents who reported that healthcare workers explained contact tracing to them were more likely to have their contacts traced. However, this association should be interpreted with caution because the adjusted estimate was imprecise, as reflected by the wide confidence interval. While the finding suggests that provider explanation may support patient engagement in contact tracing, the magnitude of the association should not be overemphasized. Clear communication by healthcare workers may help patients understand the purpose of contact tracing, address concerns, and encourage cooperation. Studies have shown that provider engagement and clear explanation of procedures, expectations, and follow-up plans improve patient understanding and compliance with contact tracing protocols [31]. Clear explanations of contact tracing procedures can reduce fear, correct misconceptions and foster patient cooperation.

Socio-cultural factors associated with contact tracing

Perceived influence of religious beliefs was positively associated with patient-reported successful TB contact tracing. However, because the variable measured only whether religious beliefs influenced participation, and not whether that influence was supportive or discouraging, the direction and mechanism of the influence should be interpreted cautiously. Rather, the association suggests that faith-related beliefs or practices may be relevant to how some patients understand and engage with contact tracing. Evidence from West African contexts indicates that religious and traditional belief systems may either support or hinder TB-related health behaviours, depending on how illness is understood and how community or faith leaders shape engagement with biomedical services [32, 33]. Future studies using qualitative methods or more detailed measures are needed to clarify the direction and mechanisms of this relationship.

Conclusion

This study found that patient-reported successful TB contact tracing was moderately high but remained suboptimal in the selected districts. Successful contact tracing was associated with educational level, district of residence and willingness to participate in contact tracing. Individual level factor associated with successful contact tracing was perceived social stigma; health system factors included perceived access to health facilities and socio-cultural factor was perceived religious beliefs. Because of the reliance on self-reported measures, these findings should be interpreted cautiously as exploratory associations rather than causal determinants. Future studies using validated measures and verified programme data are needed to better understand factors influencing TB contact tracing in Ghana.

Strengths and limitations

This cross-sectional study provides valuable, policy-relevant insights into factors influencing tuberculosis contact tracing by using primary data collected directly from confirmed TB patients across multiple districts with varying levels of contact tracing performance and by applying a comprehensive analytical framework that examined associations between socio-demographic, individual, health system, socio-cultural determinants and successful contact tracing. Although the study assessed several factors commonly understood as multidimensional constructs, including stigma, health beliefs, community attitudes, cultural beliefs, and religious beliefs, these were measured as patient-reported perceptions using binary questionnaire items rather than validated psychometric scales. As a result, the findings should be interpreted as reflecting respondents’ perceived influence of these factors on contact tracing, rather than comprehensive measurement of the constructs themselves. Future studies should consider using validated multidimensional tools to examine these constructs in greater depth. Using patient-reported perceptions may have introduced outcome misclassification if participants incorrectly recalled or reported whether their contacts had been traced. Also, the use of retrospective patient-reported measurement may have introduced recall bias. Social desirability bias may have also affected responses, especially for sensitive issues. Despite these limitations, the inclusion of diverse district contexts provides confidence that the findings offer meaningful and contextually relevant evidence to inform TB contact tracing interventions in similar settings.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

TB

Tuberculosis

WHO

World Health Organization

HIV

Human Immunodeficiency Virus

NTCP

National Tuberculosis Control Programme

CHPS

Community-based Health Planning and services

CCM

Country Coordinating Mechanism

UHAS-REC

University of Health and Allied Sciences Research Ethics Committee

IQR

Inter-Quartile Range

cOR

Crude Odds Ratio

aOR

Adjusted Odds Ratio

Author contributions

JBD, WG, SS, and GA conceptualised and designed the study. WG, SS and GA were involved in data acquisition. JBD and FOK analysed and interpreted the data. JBD supervised the research process. FOK, CMA and PAND drafted the manuscript and JBD, WG, SS and GA critically reviewed it. All authors read and approved the final manuscript.

Funding

This research did not receive funding.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Ethical approval was obtained from the University of Health and Allied Sciences Research Ethics Committee (UHAS-REC B.10 [151] 24–25) in Volta region, Ghana. Administrative permissions were granted by the respective district health directorates and facility heads. All participants received detailed information about the study and provided written informed consent prior to participation. Consent was documented either by a signature or thumbprint. Participation was entirely voluntary, and respondents were informed of their right to withdraw at any time without any consequences. No personally identifiable information was collected. All data were stored securely in a password-protected database accessible only to the research team. The study was conducted according to the declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.