Tuberculosis Preventive Behaviors and their Determining Factors among Household Contacts of Tuberculosis Patients in Thailand: A Cross-sectional Study

2.1. Study Settings

The Ministry of Public Health of Thailand has organized the 76 provinces of the country into 13 health service regional networks known as “Health Region” to enhance the effective management and allocation of healthcare resources [21]. Phichit Province, located in the lower northern part of Thailand, is one of five provinces consolidated into Health Region 3. There has been a consistent annual increase in the number of pulmonary TB cases in Phichit Province, with the figures for 2019, 2020, and 2021 standing at 488, 587, and 544, respectively. The TB treatment success rates for these years were 77.05%, 73.94%, and 75.92%, respectively, falling short of the key performance indicator, as TB treatment success rates should ideally exceed 88%. Similarly, the mortality rates during the same period were significantly higher than the target performance indicator of <5%, reaching 12.50%, 16.87%, and 15.26%, respectively [22]. This information reflects the challenging TB situation in Phichit Province.

2.2. Study Design and Study Population

In this cross-sectional study, data were collected on HHCs of TB patients. The study population comprised a total of 544 HHCs primarily tasked with the care of index cases diagnosed with pulmonary TB by a qualified medical practitioner and subsequently registered in 2021 for treatment at TB clinics in public hospitals within Phichit Province [22]. The G*Power program was used to calculate the sample size. To ensure adequate data quality for a multiple regression analysis, a power value of 0.99, an effect size value of 0.15, and an alpha value of 0.05 were considered [23]. Based on 10 independent variables, the G*Power software recommended a sample size of 226. After including a non-response rate of 10%, 249 was the required sample size for this study.

The HHCs were selected using a systematic random sampling approach. Initially, we acquired a list with the names of all 544 pulmonary TB index cases in Phichit Province from the National Tuberculosis Information Program [22]. To begin the sampling process, we randomly selected the first HHC of an index case using a simple random sampling technique. Subsequently, we determined the sampling interval by dividing the total number of HHCs by the required sample size for this study. This sampling interval was then used to sequentially select the subsequent index cases of HHCs for inclusion in the study. The inclusion criteria for HHCs in this study included having shouldered the primary responsibility for caring for an index case for at least two months, being at least 20 years of age, and expressing willingness to participate. The exclusion criterion was being diagnosed with TB.

2.3. Assessment Instruments

The questionnaire employed in this study has three sections. The first section comprises six closed-ended items designed to elicit sociodemographic information: gender, age, educational attainment, income adequacy, medical conditions, and relationship to a TB patient. The second section was constructed based on the theoretical framework of PMT [19, 20], and includes four subsections addressing four principal domains. The first domain pertains to the perceived severity of TB infection, and its subsection includes 10 items. The second domain examined concerns perceived vulnerability to TB infection, with its subsection featuring 13 items. The third domain has perceived self-efficacy at TB prevention as its focus, addressed by 14 items. The fourth domain is the response efficacy of TB prevention, and its subsection consists of 12 items. The third section comprises 20 items focused on TB preventive behaviors categorized into host, agent, and environment domains [15]. These behaviors encompass practices related to healthy self-care behaviors (host), behaviors performed to contain and prevent the spread of TB infection from TB patients (agent), and environmental management behaviors (environment) [14, 15].

The second section of the questionnaire assesses the levels of participants’ agreement using a five-point Likert scale, ranging from 1 (for strongly disagree) to 5 (for strongly agree). A five-point Likert scale ranging from 1 (for never practiced) to 5 (for practiced regularly) was also employed in the third section to measure the frequency of participants’ behaviors. To categorize mean scores into three equally sized levels [24], mean scores within the range of 1.00 to 2.33 were classified as low, while those in the 2.34 to 3.67 range were categorized as moderate, and those falling between 3.68 and 5.00 were considered high.

The content validity of the questionnaire was evaluated in consultation with three experts from different backgrounds. One expert is an academic specializing in research on the prevention of communicable diseases, another is a physician practicing at a public hospital, and the third is a public health officer at the Department of Emerging and Re-emerging Infectious Diseases of the Provincial Public Health Office. To assess content validity, the experts examined the questionnaire items and calculated the item–objective congruence values, which ranged from 0.67 to 1.00. The item–objective congruence values greater than 0.5 were considered acceptable and indicative of a high level of agreement among the experts [25]. After achieving consensus among the experts, a pilot study was conducted using a sample of 30 HHCs who live in Nakhon Sawan Province, which is in Health Region 3, the same region as Phichit Province. The reliability of the questionnaire was assessed during the pilot study. The Cronbach’s alpha coefficients were as follows: 0.739 for perceived severity of TB infection, 0.919 for perceived vulnerability to TB infection, 0.867 for perceived self-efficacy at TB prevention, 0.784 for response efficacy of TB prevention, and 0.863 for TB preventive behaviors. The coefficient values were all greater than 0.700 [26], indicating that the questionnaire possesses strong internal consistency and reliability for this study.

2.4. Data Collection

This study received ethical approval from the Human Research Ethics Committee of Naresuan University, granted under certificate of approval (COA) No. 062/2021 and institutional review board (IRB) clearance No. P3-0152/2563. Data were collected from February 2022 to July 2022 using the questionnaire. The researchers sought and obtained permission from the director of the Phichit Public Health Provincial Office, and from the administrators of each TB clinic in the public hospitals within Phichit Province. To ensure the accuracy and consistency of the data, the researchers conducted a half-day training session for five research assistants involved in the study: one nurse and four public health officers. This training aimed to acquaint the research assistants with the content of the questionnaire, the proper procedures for data collection, and the rights and privacy of the study participants. The research assistants contacted each prospective participant via telephone, provided a comprehensive explanation of the research objectives, and requested their decision on whether to participate. Those who agreed were subsequently consulted to determine their preferred time and location for data collection. Each participant expressed their consent by signing informed consent forms. Subsequently, the research assistants retrieved the questionnaires from a total of 249 HHCs and forwarded them to the researchers for data analysis. The researchers excluded four questionnaires from the data analysis due to incompleteness. Consequently, the final set of questionnaires utilized in the data analysis numbered 245.

2.5. Data Analysis

The independent variables included six sociodemographic variables and four variables related to perceptions surrounding TB infection and prevention. The dependent variable was the TB preventive behaviors of HHCs. Descriptive statistics, including frequency, percentage, mean, and standard deviation (SD), were used for both independent and dependent variables. Multivariate linear regression using the stepwise method was employed to examine the impact of the independent variables on the dependent variable. The underlying assumptions of multiple linear regression analysis were tested [27] before conducting the analysis, and all assumptions were met. The normality of the residuals was assessed using a probability–probability (P–P) plot, and it showed no significant deviations from the expected straight-line pattern of normality. The Durbin–Watson value was 1.661, which is within the 1.50–2.50 range, indicating no autocorrelation between the residuals. Tolerance values exceeded 0.2, and variance inflation factor (VIF) values were below 10.0, indicating that there was no multicollinearity between the independent variables. Scatter plots revealed a symmetrical distribution, indicating that there was no homoscedasticity in the residuals. Data analysis was conducted using the IBM Statistical Package for the Social Sciences (SPSS) software, version 22. Statistical significance was established at a significance level of <0.05.

4.1. TB Preventive Behaviors among HHCs

The results of this study shed light on various TB preventive behaviors among HHCs. These behaviors are categorized into three domains: healthy self-care behaviors, behaviors performed to contain and prevent the spread of TB infection from TB patients, and environmental management behaviors [14, 15]. The collective average score for all TB preventive measures was notably high, with a mean of 3.84 ± 0.692. These findings reflect the effectiveness of TB policies and awareness initiatives aimed at TB prevention among HHCs in Thailand [28] and align with those of a prior study conducted in Bangladesh, in which approximately 43.6% of the participants demonstrated a high level of TB preventive behaviors [18].

Notably, the mean score (X̅ = 3.45 ± 0.752) for the healthy self-care behaviors domain was moderate. In particular, the practice of undergoing TB screening every six months was the least commonly performed behavior (X̅ = 2.93 ± 1.359). One plausible explanation for this observation in our study is that a substantial proportion of the participants (49.0%) were aged 50 or older, and nearly half (47.3%) reported having insufficient income to cover their living expenses. Therefore, accessing TB screening services at healthcare facilities might be challenging due to their age and financial constraints. In view of this, a prior study conducted in Thailand proposes that providing financial support to HHCs for transportation should lead to higher TB detection rates [29]. A study conducted in Vietnam found that the contact screening of HHCs is primarily dependent on self-referral or passive methods. Their recommendation is to explore active contact screening with scheduled follow-ups [30]. Another study investigated TB screening coverage among HHCs in Southern Thailand, and their findings reveal a screening coverage rate of 46.6% among their 174 HHC participants. Interestingly, about 20% of these HHCs did not receive any TB screening recommendations [31]. Additionally, HHCs were more likely to undergo TB screening when advised by a healthcare professional than by someone outside that profession [31]. Therefore, the challenges surrounding TB screening among HHCs may require highly targeted interventions and the involvement of healthcare professionals.

4.2. Factors Determining TB Preventive Behaviors among HHCs

In this study, we found that the mean scores for the variables on HHCs’ perceptions regarding TB infection and prevention were high across all four domains: perceived severity of TB infection, perceived vulnerability to TB infection, perceived self-efficacy at TB prevention, and response efficacy of TB prevention. This finding aligns with the findings of another study, which found that a significant majority of TB patient attendants reported a high level of perceived susceptibility (76.1%), a high level of perceived severity of the disease (42.6%), a high level of perceived benefit from TB prevention behaviors (55.3%), and a high level of perceived barriers to performing these behaviors (40.1%) [18]. Furthermore, we found that TB preventive behaviors are determined to a statistically significant degree by the response efficacy of TB prevention (b = 0.396, P-value < 0.001) and perceived self-efficacy at TB prevention (b = 0.260, P-value = 0.01). This indicates that when HHCs have an in-depth understanding of the advantages of TB prevention behaviors and confidence in their ability to prevent TB infection, the frequency of their TB preventive behaviors tends to increase. This finding is consistent with that of a previous study conducted in Bangladesh, which found that response efficacy or perceived benefit is positively correlated with preventive behavior (r = 0.179, P-value < 0.05) [18]. Furthermore, this finding is consistent with the findings of a study conducted in Thailand, which indicates that self-efficacy in TB prevention is statistically predictive of TB-prevention behavior among village health volunteers in the community (β = .244, P-value < .05) [17]. A previous meta-analysis research on PMT argues that perceived self-efficacy and response efficacy are vital components of the coping appraisal process, which assesses an individual’s ability to handle and prevent potential threats [32], such as TB infection of HHCs in our study. Response efficacy refers to the belief that an adaptive response will yield positive results, indicating that taking protective actions will effectively safeguard oneself or others. On the other hand, perceived self-efficacy refers to an individual’s perception of their capability to execute an adaptive response [19, 20, 32]. Self-efficacy is recognized as a crucial element of any health behavior theory, as it plays a pivotal role in regulating motivational, cognitive, and emotional processes [32]. Hence, based on the results of our study, we recommend that healthcare institutions and relevant agencies responsible for equipping and training HHCs should facilitate HHCs recognizing their abilities and competence at TB prevention.

This study found that male HHCs were more likely than female HHCs to engage in risky TB preventive behaviors that could increase the likelihood of TB infection. This finding is consistent with those of previous studies conducted in diverse TB burden settings [5, 33-35]. TB infection rates are much higher among men than among women, as indicated by a male-to-female ratio of 1.7 for worldwide case reports [33]. A study on the risk factors for TB among HHCs in Vietnam reports that males are nearly 1.5 times more likely than females to develop TB [5]. It is unclear and widely discussed in the literature why gender-specific susceptibility exists and the reason is probably highly complex. Sociocultural roles and behaviors likely contribute to higher TB rates in men than in women [33, 36]. For instance, studies conducted in high-burden countries reveal correlations between smoking and the chances of contracting TB—and men are more likely to smoke than women [37, 38]. Similarly, alcohol consumption is linked to an elevated risk of TB, with men being more likely to drink than women [4]. Through a systematic review and meta-analysis, a study has highlighted that men in low- and middle-income countries are disadvantaged and face challenges in their endeavors to seek and obtain TB care in diverse settings [35]. Another study, which investigated risk factors for testing and evaluating close contacts of TB, found that males were significantly more likely than females to not be appropriately evaluated [39]. Therefore, the role of gender in susceptibility to TB may be attributable to gender-based social interactions and behaviors that may enhance exposure to TB and thus raise the risk of contracting the disease [40].