Development and validation of self-management scale for tuberculosis patients

Tuberculosis (TB) is one of the most serious problems threatening global health today [1]. In 2020, approximately 10 million new cases and 1.5 million deaths due to TB were reported globally [1]. To prevent the development of drug resistance, compliance with anti-TB treatment is essential. However, patient compliance is a challenge, as treatment is typically long-term (at least 6 months). The World Health Organization (WHO) recommends directly observed therapy, short-course (DOTS) to control TB [2], where patients ingest medications under the direct observation of medical workers as a case management strategy to improve compliance.

According to WHO, China’s National Tuberculosis Control Program is one of the most successful DOTS programs in the world [3]. However, the results of a survey found more than one-third of TB patients in Chongqing were never supervised by any healthcare workers [4]. A meta-analysis [5] indicated that only 20% of patients with TB were covered by Directly observed therapy (DOT) in China, which was much lower than reported by official statistics. DOT requires that patients either go to a clinic or have DOT workers come to their homes to observe the ingestion of medication [6]. In this way, DOT remains time-consuming, resource-intensive, difficult to achieve over distances, and represents the largest single cost of TB treatment [7]. Furthermore, DOT is controversial due to challenges associated with autonomy and leaving patients as passive recipients of treatment [8]. Individual studies have described DOT as intrusive and disempowering for patients [7, 9]. Therefore, reasons for patients’ failure to attend DOT mainly include a lack of primary medical personnel, poor transport access, and patients’ concerns about privacy, autonomy, dignity, and integrity [10‐12]. Therefore, WHO recommended the people-centered care strategies in 2014, which encourage patient self-management and community engagement when medically feasible [13, 14]. Patient-centered care and supervision must be carried out in a context-specific and patient-sensitive manner to treatment interruption.

There is no standard definition of self-management. Barlow et al. [15] depicted self-management as “An individual’s ability to manage the symptoms, treatment, physical, psychosocial consequences, and lifestyle changes inherent in living with a chronic disease”. Lorig and Holman [16] defined self-management as “learning and practicing skills necessary to maintain an active and emotionally satisfying life in the face of a chronic disease”. In this study, we defined self-management as the ability to actively manage disease in daily life, change lifestyles closely related to the disease, and effectively cope with the physical and mental effects of the disease. Effective patient-centered TB case management emphasizes the patient needs [17] and requires individually tailored management strategies among TB patients, rather than universally applying DOT. Furthermore, self-management of symptoms, treatment, and physical and psychological problems [15, 18] can significantly improve medication adherence in patients with chronic illness [19, 20], help establish healthy behaviors [21, 22], and promote treatment outcomes [23, 24]. More than half of TB or multidrug-resistant TB (MDR-TB) patients self-administer treatments [25, 26]. Therefore, self-management of TB patients (SMTP) is important and could improve treatment outcomes if sufficient knowledge and skills are promoted [27, 28]. Moreover, SMTP may be a useful supplement for patients who are reluctant to accept or with limited access to DOT. However, little research has been conducted on SMTP, with only a few studies have discussed the definition of [29, 30], factors associated with [31‐33], and interventions using SMTP [34, 35]. However, understanding and developing an instrument to measure SMTP before intervention is crucial.

SMTP measurement is the foundation for decision-making regarding SMTP intervention delivery and evaluation. Two studies explored tools related to measuring the behaviors of patients with TB, one of which explored the scale of health self-management assessments for elderly patients with TB. The study primarily evaluated the knowledge, attitude, and behaviors (health responsibility, psychology, physical activities, and nutrition) of those patients [36]. The second study developed a self-discipline scale for TB patients treated at home, which evaluated factors including access to information, compliance, isolation control, and daily life management [37]. However, neither study proposed an instrument for evaluating the key behaviors related to SMTP, such as prevention and treatment. Therefore, to our knowledge, no study has yet developed an instrument to measure SMTP.

To address these gaps in knowledge, this study aimed to develop and validate a self-management scale for TB patients (SMSTP), which could be used to measure SMTP and evaluate the effectiveness of interventions using SMTP.

In this study, we developed SMSTP to measure SMTP and evaluate the effectiveness of interventions. The SMSTP consisted of 17 items in three dimensions, which focused primarily on the behavioral aspects of SMTP. Previous studies reported psychological dimension in instrument of self-management which evaluated the associations between psychological dimension and partnership with healthcare professionals, and self-management [57‐59]. Even though, psychological and partnership with healthcare professionals are important factors for SMTP behavior’s adaption or change, the SMSTP focused on evaluating the level or change of behaviors related to SMTP. According to the results, the SMSTP demonstrated good reliability and validity and could be used to assess and develop SMTP interventions. Future research should consider other dimensions of self-management, such as psychological dimension and the dimension of the relationship with professionals.

Although there is no uniform definition, self-management is considered an important part of managing chronic disease, with the aim of illness prevention and promoting health [60]. Health Promotion is defined as the process of enabling people to increase control over and improve their health [61]. Thus, self-management can be considered an important goal of health promotion. In our previous study, we constructed an indicator framework for TB health promotion. The framework contains not only treatment related behaviors but also health-related behaviors that contribute curing TB, such as improving nutrition, avoiding smoking, and prohibiting alcohol consumption. In addition, from the perspective of public health, the framework also includes indicators of behaviors that prevent TB. If TB patients possess these abilities, they will have the ability to manage TB and promote their health, which is self-management. Therefore, in this study, we developed the SMSTP based on the framework.

It can be concluded based on factor analysis that neither Item 1 nor Item 3 in the initial scale reached statistical significance. In terms of clinical significance, Item 1 (Medication-taking record card filled), Item 3 (Taking medications when going out), and Item 6 (Taking medications consistently), all investigated a patient's adherence to medication. Item 1 was an external means to help a patient adhere to medication, Item 3 was a demonstration of medication adherence, and Item 6 could directly assess whether a patient adhered to medication. Considering the results of statistical analysis and clinical significance, we finally deleted Items 1 and 3.

After two-round EFA, we developed the SMSTP with 17 items to assess the self-management levels of patients with TB. Three dimensions were extracted from the 17 items that accounted for 76.60% of the total variance. Furthermore, later internal consistency evaluations showed that Cronbach's α exceeded 0.9 for the total scale and each of the dimensions, indicating that the scale had high internal consistency and that no additional adjustments to the SMSTP were required.

Notably, the most influential dimension was “transmission prevention behavior” followed by “supportive behavior” and “adherence to treatment behavior.” This finding was inconsistent with existing empirical evidence [62, 63]. Both the DOTS strategy recommended by the WHO [64] and the vast majority of studies [25, 65, 66] suggest that treatment adherence is an important determinant of TB control. TB can spread by air, and each patient with active TB can infect an average of 10–15 individuals each year [67]. Accordingly, transmission prevention behaviors among patients with TB may be one of the key measures in reducing community transmission. Moreover, the National Health Commission of China’s Technical Specifications for Tuberculosis Prevention and Control [68] highlighted the importance of prevention. Our results also showed that “transmission prevention behavior” could more likely predict improved SMTP than “adherence to treatment behavior,” which may suggest that TB health promotion should focus more on these behaviors.

Several studies have demonstrated that self-management performance is primarily influenced by a patient's characteristics and lifestyle choices [69]. The EFA results revealed that “supportive behavior,” accounting for 69% of the total variance of TB self-management, was also higher than “adherence to treatment behavior.” Smoking, alcohol consumption, and poor nutritional status are known to be associated with poor treatment outcomes in patients with TB [70, 71]. Furthermore, lifestyle interventions have become increasingly popular strategies for chronic disease management [72]. Hence, lifestyle modifications should be considered an important component of TB self-management interventions.

Cronbach’s α values of the SMSTP were greater than 0.9 for all dimensions and the total scale, indicating a high level of internal consistency and reliability. Test–retest reliability showed that the same rater's scores were consistently measured over time. Typically, the ideal interval between the test and retest is 2–4 weeks [73‐75]. Considering that test–retest reliability with a time interval greater than 2 weeks may be affected by the fluctuation process of the behavior itself, the test–retest reliability was tested for a time interval of two weeks. The ICC values of the three dimensions in the SMSTP were all > 0.7, greater than the standard threshold [76]. This indicated that the measurement time had little influence on the reliability of this scale, and that it had strong time flexibility and stability in assessing SMTP. In conclusion, our findings indicate that the SMSTP had good overall reliability.

The SMSTP was found to have excellent content validity in terms of questionable relevance. Correlation analysis showed high correlations (> 0.8) with the dimensions and only moderate or weak correlations in the other dimensions. In addition, correlations between the dimensions were weak. These findings imply that close associations between items and dimensions could clarify the practical sense of the dimension to which they belonged. Each dimension and the items they contained were also distinguishable from the other dimensions. Therefore, the three dimensions could be used separately to assess the different components of self-management depending on the different needs of assessment.

EFA results revealed that the KMO test (KMO = 0.970) and Bartlett’s test of sphericity were ideal for factor analysis. Moreover, the cumulative contribution rate of this study was over 76%, which indicated that the SMSTP could adequately measure the level of SMTP and confirmed good construct validity of the questionnaire. CFA was also used to verify the construct validity of the scale. All CFI, TLI, NFI, IFI, and GFI values were above 0.9, and the χ²/df values were within the ranges suggested by Wheaton [77]. The RMSEA was less than 0.08, and the RMR was less than 0.05, both meeting the thresholds recommended by Browne and Cudeck [78]. Only one index, AGFI, was 0.889 and did not reach the standards that would indicate an acceptable model fit. This index, however, is likely to be underestimated when the sample size is less than 300 [79]. Overall, the CFA model fit indices of the SMSTP were acceptable, but less than perfect.

Additionally, convergent and discriminant validity were assessed to evaluate construct validity [80]. A high level of convergent validity can assist researchers in understanding how the constructs of the three measures are interconnected both theoretically and practically [81]. To determine convergent validity, we needed to know how well the latent construct could explain the variance of each indicator, which was assessed through AVE [82]. AVE represented the mean value of the commonality of the indicators of a certain construct [83]. Our results indicated that the AVE of all three dimensions of the SMSTP was greater than the critical value of 0.5, demonstrating that the amount of variance between dimensions and corresponding indicators exceeded that caused by measurement errors [83]. Thus, the convergent validity of the SMSTP was established.

Furthermore, we assessed discriminant validity using the Fornell–Larcker criterion and HTMT. First, the square root of the AVE for each dimension was greater than the correlation between it and the other dimensions, which met the Fornell–Larcker criterion [84]. Second, the HTMT values of the correlations were used to assess the discriminant validity. The HTMT is determined by comparing the average correlations of indicators across constructs, which measure different constructs to those indicative of the same construct [85]. According to the criteria of Henseler et al. [86], the HTMT values should be less than 0.85. Results from this study revealed that all HTMT values were far below the conservative 0.85 upper bound, illustrating that the SMSTP constructs were separable from each other. Overall, the results of both the Fornell–Larcker criterion and HTMT implied that the scale had good discriminative validity.

This study developed and validated SMSTP which consisted of 17 items in three dimensions: adherence to treatment behavior (5 items), supportive behavior (6 items), and transmission prevention behavior (6 items). Each item was rated on a 5-point Likert scale, and where higher scores indicated a higher self-management. SMSTPS will be used to design intervention strategies for SMTP and evaluate the effectiveness of interventions on SMTP.