INTRODUCTION

Tuberculosis (TB) is a chronic infectious disease that necessitates long-term treatment, thus presenting challenges in sustained medication adherence and patient management. Factors such as low income, unemployment, lack of healthcare access, alcohol addiction, and homelessness contribute to an increased risk of adverse outcomes [1]. Moreover, the causal relationship between poverty and TB is bidirectional; poverty may lead to TB, or TB may result in poverty due to decreased productivity and loss of income [2]. Recognizing these dynamics, the World Health Organization (WHO) has emphasized the importance of patient-centered care and social protection as fundamental elements of its End TB Strategy [3]. The WHO advocates for a shift from a medical service provider-centric model to one that recognizes patients as partners, calling for medical, public health, and social support tailored to individual patient needs [4].

Korea has an intermediate TB burden with a gradually decreasing incidence. Historically, the national TB control system has operated through public health centers. However, since the integration of national health insurance across all private medical institutions in 1989, the number of patients with TB seeking treatment at these private facilities has increased [5]. In such institutions, which were previously not included in the national TB control system, issues such as low treatment success rates and inadequate patient care have emerged [6]. To address these problems, the Korean government has implemented policies to strengthen patient care by assigning TB specialist nurses to private medical institutions that frequently serve patients with the disease. This initiative began with the public-private mix (PPM) pilot program in 2007 [5]. The PPM program succeeded in improving treatment success rates within private medical institutions; however, PPM coverage varied, and patient care remained suboptimal in certain types of medical facilities [7]. The need was evident for a vulnerable population approach that offers additional support to those at risk, aiming to mitigate health inequities in patient care and the management of infectious diseases [8]. In response, the Korean government launched an enhanced community-based care and management (ECCM) program as a pilot initiative, intended to focus on vulnerable patients with TB, in addition to the medical institution-centered PPM program. Since the introduction of the PPM program in Korea, limited consideration has been given to strengthening the capacity of TB care and management at the community level [9]. This study was conducted to determine whether the vulnerability assessment tool developed for the ECCM program could adequately predict final treatment outcomes for patients with TB and to evaluate whether ECCM, tailored to the patient’s vulnerability level, could improve these outcomes.

METHODS

Study Sites and Design

The study included 2 cities and was conducted in 2 stages. Daejeon (DJ), a city in Korea, has a population of approximately 1.5 million. As of 2018, the year before the present study began, DJ had a TB notification rate of 50.6 per 100 000 residents, with 751 TB cases diagnosed annually. Daegu (DG), another Korean city, has a population of about 2.47 million. In 2018, DG had a TB notification rate of 60.9 per 100 000 residents, with 1491 TB diagnoses per year. In the first stage of this study, the ECCM program was implemented for the residents of DG, which served as the intervention region, while DJ was considered the control region. The second stage utilized a crossover design; the ECCM program was introduced to the residents of DJ, now the intervention region, and DG became the control region. A survey was administered to all patients who received a TB diagnosis notification in both regions during the study period. The survey incorporated a vulnerability assessment questionnaire specifically developed for this research. Based on the results of the assessment, patients were categorized into low vulnerability and high vulnerability groups. Those in the intervention group received active case management from a case manager, while the control group received no such intervention. TB treatment outcomes were compared as the outcome variables. This study was registered with the Clinical Trial Registry of Korea (http://cris.nih.go.kr, KCT0006622).

Vulnerability assessment and intervention

The vulnerability assessment tool was developed to evaluate 3 dimensions of vulnerability in patients with TB: clinical, socioeconomic, and TB-related vulnerability. This tool comprises 20 items, which were derived by referencing the British Enhanced Case Management (ECM) system and items from the Japanese directly observed therapy-short course (DOTS) conference, in addition to further literature reviews and expert consultation [10,11]. The items and their weighted scores are detailed in Supplemental Material 1. The potential weighted values for the vulnerability assessment tool were 1 point, 2 points, 3 points, and 5 points. This categorization was based on the weighted scores used for the Japanese DOTS conference items, which vary by region (1, 2, 3, 10, and 15 points), and the 5 British ECM levels, which range from 0 to 4. The present tool, with a score range of 0 to 63, was evaluated for validity through a survey of 6 TB specialists, following the methodology proposed previously [12]. Individual (I-CVI) and overall, or scale (S-CVI), content validity indices (CVIs) were calculated. Items with an I-CVI of less than 0.78 were revised to attain a final S-CVI of 0.85 [13]. A pilot survey was conducted among 27 patients with TB aged 65 years and older who were living alone and participating in the directly observed therapy (DOT) program for older adults. Based on the results of the pilot survey and discussions among the researchers, the cut-off for vulnerability level was established at 17 points (the median score). The Cronbach alpha, calculated to confirm internal consistency, was 0.635.

A trained case manager visited the participants with high vulnerability scores (≥17) to conduct case investigations. This occurred after explaining the study details and obtaining the patient’s consent. The case investigation, similar to the vulnerability assessment tool, evaluated the patient’s clinical, socioeconomic, and TB-related needs. Based on these findings, the research team discussed the necessity for support measures. Patient support included DOT, and participants received compensation of approximately US$180 (200 000 Korean won) per month for the 2 months of treatment. Additionally, administrative assistance was provided by connecting patients with various social welfare systems. The frequency and method of DOT implementation (personal visit vs. mobile visit) were selected according to the patient’s living situation. The incentive was individualized based on socioeconomic needs. For patients eligible for social benefits, Medical Aid, and/or long-term care benefits who were not currently receiving them due to a lack of knowledge about the application process, we provided administrative procedural assistance. With this support, we aimed to ensure their long-term and stable integration into the social security system.

Statistical Analysis

The outcome variables were categorized as either favorable (cure and treatment completion) or unfavorable (failure, loss to follow-up [LTFU], and death). Treatment outcomes were determined using the values from the National Tuberculosis Surveillance System, which collectively defines cure and completion as treatment success, in line with the WHO definition [14]. Patients whose treatment outcomes were recorded as transfer-out were classified as LTFU if the interval between records was 2 months or longer. If the interval was less than 2 months, the patients were considered to have undergone a single treatment process, and the treatment outcome noted in the final record was used for analysis. LTFU, treatment failure, TB-related death, and other death were collectively categorized as unfavorable outcomes. The following covariates were included in the analysis: age (grouped as <39, 40-49, 50-59, 60-69, 70-79, and ≥80 years), gender, study region, bacteriological classification of TB (either bacteriologically confirmed drug-susceptible or clinically diagnosed TB with unknown bacteriological status), location of the TB lesion (pulmonary, extrapulmonary, or mixed TB), TB history (new or previously treated case), type of healthcare facility (general hospital, hospital, clinic, or community health center), smoking history (non-, ex-, or current smoker), chest X-ray results (normal, abnormal, or unknown/missing), acid-fast bacillus sputum smear results (negative, positive, or unknown), and sputum culture results (negative/unknown or positive).

The sensitivity and specificity of the vulnerability assessment tool were evaluated to determine its predictive ability for unfavorable outcomes. As the outcomes in the intervention group may have been influenced by the intervention, only the vulnerability assessment data from the patients in the control group were used for receiver operating characteristic curve analysis. Area under the curve (AUC) values and 95% confidence intervals (CIs) were calculated based on the distribution of sensitivity and 1−specificity. To assess the impact of ECCM, the study examined patients with TB who had reported final treatment outcomes. The Pearson chi-square test was used to identify differences in treatment outcomes and the distribution of baseline characteristics between the intervention and control groups. An intervention (ECCM)-stratified multivariate logistic regression model was developed to evaluate the influence of vulnerability level on unfavorable outcomes and to examine the effect modification of the ECCM intervention. Odds ratios (ORs) and 95% CIs were determined according to the level of vulnerability and the application of the ECCM intervention, while additive interaction was measured using the relative excess risk due to interaction (RERI) and the proportion of disease attributable to interaction (AP). Additionally, a sensitivity analysis was conducted, which excluded cases of death from other causes and applied the same models as the main analysis. To calculate the required sample size, we assumed that the highest recently reported unfavorable rate (29.4%) would be reduced to the recent 5-year average unfavorable outcome rate of the TB surveillance system (12.8%), as previously reported [7]. Consequently, a minimum of 94 patients with TB and high vulnerability were required in each group (alpha=0.05, 1−beta=0.8). All statistical analyses were performed using Stata/SE version 15 (StataCorp., College Station, TX, USA). The p-values of less than 0.05 were considered to indicate statistical significance.

Ethics Statement

This study was approved by the Institutional Review Board of the Korean National Tuberculosis Association in Seoul, Korea (2019-KNTA-IRB-05). All procedures were conducted following ethical approval, and written informed consent was obtained from all participants. No deviations or violations occurred during the study period.

RESULTS

The study was conducted from November 2019 to September 2020, during which 2765 patients were notified of a TB diagnosis in the 2 Korean cities. After excluding 1409 patients who did not participate in the vulnerability assessment, the remaining patients were divided into the intervention (n =700) and control (n=656) groups. Of the 1356 patients who completed the vulnerability assessment, 639 were included in the final analysis; 714 were excluded because they were still undergoing treatment at the time of analysis, and 3 were excluded due to multidrug-resistant TB (Figure 1). Both intervention and control groups had a high proportion of participants aged 65 years and older, with more men than women in each group. Most patients received treatment in a general hospital. The retreatment rates were 13.9% in the control group and 13.1% in the intervention group. The prevalence of high vulnerability scores was similar in both groups, with 13.7% in the intervention and 11.2% in the control group. Due to funding-related administrative constraints, including limitations on the budget and study duration, the recruitment of patients with high vulnerability fell short of the initial target. No significant differences in baseline characteristics were observed between the 2 groups (Table 1).

The overall rate of treatment success (that is, favorable outcome) was 72.2%. Success rates for the low vulnerability and high vulnerability groups were 76.7% and 40.0%, respectively. The AUC for the vulnerability assessment score’s capacity to predict unfavorable treatment outcomes was 0.70 (95% CI, 0.63 to 0.77). The sensitivity and specificity for the prediction of unfavorable outcomes were 27.4% and 95.6%, respectively. Given an calculated to be 70.6% and 77.2%, respectively (Figure 2).

The treatment success rates of the highly vulnerable patients in the intervention and control groups were 47.9% and 28.1%, respectively. The rate of LTFU was 8.4 percentage points (%p) lower in the intervention group compared to the control group. The TB-related death rate and the rate of death from other causes were 7.3%p and 4.2%p lower, respectively, in the intervention group than in the control group. No significant differences were observed in treatment outcomes between the groups for patients with low vulnerability (Figure 3).

Within the vulnerability assessment tool, most items in the clinical dimension, as well as multiple in the socioeconomic dimension, demonstrated significant associations with unfavorable outcomes. In contrast, no items in the TB-related dimension displayed such relationships (Supplemental Material 2). Furthermore, the vulnerability scores for the clinical and socioeconomic dimensions varied according to treatment outcomes in both the intervention and control groups (Supplemental Material 3). In the multivariate regression model, patients with low vulnerability from the intervention group served as the reference category. The odds of experiencing unfavorable outcomes were not significantly higher for patients with low vulnerability in the control group (OR, 1.16; 95% CI, 0.73 to 1.84). In contrast, patients with high vulnerability in the intervention group (OR, 3.57; 95% CI, 1.74 to 7.34) and those with high vulnerability in the control group (OR, 10.55; 95% CI, 4.13 to 26.95) faced significantly elevated risks of unfavorable outcomes. The probability of such outcomes escalated with an increasing level of vulnerability (intervention group: OR, 4.41; 95% CI, 1.99 to 9.78; control group: OR, 10.12; 95% CI, 3.87 to 26.47). Among patients with the same level of vulnerability, the likelihood of unfavorable outcomes was greater in the control group than in the intervention group, although these differences were not statistically significant (Table 2). A potential synergistic effect between ECCM intervention and the level of vulnerability with regard to unfavorable outcomes was observed (RERI, 6.86; 95% CI, -2.89 to 16.61; AP, 0.65; 95% CI, 0.27 to 1.03), although statistical significance was noted only for AP. Sensitivity analysis showed similar results (Supplemental Material 4).

DISCUSSION

This study explored the impact of tailored social support based on individual vulnerability on treatment outcomes in patients with TB. The examined vulnerability assessment tool demonstrated moderate predictive ability regarding final TB treatment outcomes. ECCM appeared to decrease the likelihood of unfavorable outcomes; however, its effectiveness remains inconclusive due to the limited number of enrolled patients with high vulnerability. The interaction between vulnerability level and ECCM played a role in influencing unfavorable treatment outcomes.

The capacity of social support to improve TB indicators has been emphasized in cross-country comparisons [15]. Research involving a meta-analysis of the impact of cash transfers on TB treatment outcomes in low-income and middle-income countries has also yielded positive findings [16]. Similarly, a study from Brazil demonstrated that a social protection program aimed at alleviating poverty led to an increase in the TB treatment success rate and a decrease in the LTFU rate [17]. Furthermore, an English study that implemented an ECM program tailored to patient vulnerabilities found that vulnerability level is a powerful predictor of final treatment outcomes [18]. A Japanese study also indicated that case management tailored to patient vulnerability levels improved the treatment success rate and lowered the LTFU rate, specifically during the later part of the program compared to the earlier stage [19].

The present findings partially reproduce those of previous research. Specifically, our results support the notion that high vulnerability at the time of TB diagnosis is a strong predictor of unfavorable outcomes. Additionally, this study suggests that a social support program such as ECCM can reduce disparities in treatment outcomes based on the patient’s vulnerability level, although the observed differences did not reach statistical significance. The absence of statistical significance in this study may be attributed to the limited number of patients with high vulnerability, which could have impacted the study’s statistical power, particularly in identifying associations between high vulnerability and treatment outcomes. Future research should address these limitations by securing sufficient resources to include a more representative sample of highly vulnerable populations. Despite this limitation, our findings offer valuable preliminary insights into the role of vulnerability in TB case management. However, caution should be exercised when attempting to generalize these results to broader populations.

Another notable feature of the present study is that the social support provided through the ECCM program led to a reduction in the TB-related death rate, which contrasts with findings from previous research. Although the difference was inconclusive due to the range of uncertainty, TB-related deaths were 7.3%p lower among highly vulnerable patients in the intervention group compared to those in the control group. Previous studies have indicated that social support decreases the likelihood of LTFU and increases the treatment success rate [16,17]. In Korea, a study found that a social protection program, including housing support for homeless patients with TB, reduced the mortality rate by approximately 15%p compared to a historical control group [20]. Moreover, public social support policies in Japan have shown promise in mitigating the adverse effects of working conditions and are recognized as a key strategy for improving TB survival rates [21]. These findings underscore the importance of considering the varied needs of patients with TB, such as employment status, when designing social support programs. Thus, social support not only helps sustain treatment to lower the LTFU rate but also provides a stable treatment environment for patients who lack essential material resources, thus reducing mortality. Throughout the ECCM program in the present study, we offered various types of support based on patient needs. Most of the services provided were focused on helping patients gain access to social services, including the National Basic Livelihood Security System and Elderly Long-term Care Insurance. In some cases, we assisted with applications for rehabilitation procedures. Most importantly, our case manager established a rapport with the patients, similar to friendship, which may have provided emotional support, helped alleviate feelings of isolation, and encouraged treatment adherence.

This research has several limitations, and the results should be interpreted with caution. First, the study was not a randomized controlled trial. Although baseline characteristics were evenly distributed between the intervention and control groups, suggesting minimal influence from measured confounders, the results may still have been affected by unmeasured confounders. Second, the selection of the vulnerability cut-off level was not based on robust evidence. Instead, the median value was chosen for this pilot program due to the inclusion of older patients (≥65 years) living alone, many of whom were socioeconomically vulnerable. This approach lacks scientific rigor and does not represent a data-driven estimation. Nevertheless, a post-hoc analysis using a cut-off of 17 points displayed low sensitivity but high specificity (≥95%) in predicting unfavorable outcomes. Future research should aim to establish an appropriate policy-oriented cut-off point, considering the resources available. Third, the study did not enroll the appropriate number of patients with TB with high vulnerability as determined by the calculated sample size. This was due to several factors, including a pre-fixed study end date of September 2020 due to budget constraints, as well as the direct impact of the coronavirus disease 2019 pandemic. Consequently, the rates of LTFU were relatively high (4.1, 2.1, and 13.8% in the total, low vulnerability, and high vulnerability groups, respectively) compared to the national surveillance rate of 2.2% in 2019 [22]. This discrepancy may be related to the exclusion of patients still undergoing treatment, who are likely to complete treatment successfully without experiencing LTFU. Future investigations should take these administrative limitations into account to improve the ECCM program. Finally, because the study included patients with TB from 2 metropolitan cities, the results likely do not adequately reflect rural areas, where older, more vulnerable populations are likely to be found. As age was the strongest predictor of unfavorable outcomes, further research in rural settings is warranted.

In this study, we employed a vulnerable population approach model to screen patients with TB for vulnerability and evaluated the impact of the ECCM program on treatment outcomes. By assessing the prognostic capability for the level of vulnerability, our findings indicate that tailoring social support to a patient’s vulnerability level and needs may improve treatment outcomes for individuals with TB. Consequently, policies should be developed that leverage the capacity of social support to reduce disparities in treatment outcomes.

Supplemental Materials

Notes

Data Availability

The datasets utilized and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Conflict of Interest

The authors have no conflicts of interest associated with the material presented in this paper.

Funding

The study was supported by the Korea Disease Control and Prevention Agency (KDCA) (grant No. 2019-E3701-00), the National Research Foundation of Korea (NRF) grant funded by the Korean government (Ministry of Science and ICT; grant No. RS-2023-00240537), and a Korea University Grant (K2406681).

Acknowledgements

The authors would like to express their gratitude to all study participants and to the hospital staff who provided care and management for patients with tuberculosis.

Author Contributions

Conceptualization: Choi H, Han J, Kwon Y, Shim E. Data curation: Seo J, Jeong D, Lee IH. Formal analysis: Jeong D, Choi H. Funding acquisition: Choi H. Methodology: Jeong D, Choi H. Project administration: Seo J, Lee IH. Visualization: Choi H. Writing – original draft: Seo J, Choi H. Writing – review & editing: Seo J, Jeong D, Lee IH, Han J, Kwon Y, Shim E, Choi H.

Figure. 1.

Study flowchart. MDR-TB, multidrug-resistant tuberculosis.

Figure. 2.

Sensitivity and specificity of the prediction of unfavorable outcomes by the vulnerability assessment tool (n=287). The area under the receiver operating characteristic curve was 0.70 (95% confidence interval, 0.63 to 0.77; standard error, 0.04).

Figure. 3.

Treatment outcomes of the intervention and control groups based on level of vulnerability (A) high, (B) low. LTFU, lost to follow-up; TB, tuberculosis.

REFERENCES

• 1. Lönnroth K, Jaramillo E, Williams BG, Dye C, Raviglione M. Drivers of tuberculosis epidemics: the role of risk factors and social determinants. Soc Sci Med 2009;68(12):2240-2246. https://doi.org/10.1016/j.socscimed.2009.03.041ArticlePubMed
• 2. Ortblad KF, Salomon JA, Bärnighausen T, Atun R. Stopping tuberculosis: a biosocial model for sustainable development. Lancet 2015;386(10010):2354-2362. https://doi.org/10.1016/S0140-6736(15)00324-4ArticlePubMedPMC
• 3. World Health Organization. Implementing the end TB strategy: the essentials; 2015 [cited 2025 May 19]. Available from: https://iris.who.int/handle/10665/206499
• 4. Odone A, Roberts B, Dara M, van den Boom M, Kluge H, McKee M. People- and patient-centred care for tuberculosis: models of care for tuberculosis. Int J Tuberc Lung Dis 2018;22(2):133-138. https://doi.org/10.5588/ijtld.17.0608ArticlePubMed
• 5. Go U, Park M, Kim UN, Lee S, Han S, Lee J, et al. Tuberculosis prevention and care in Korea: evolution of policy and practice. J Clin Tuberc Other Mycobact Dis 2018;11: 28-36. https://doi.org/10.1016/j.jctube.2018.04.006ArticlePubMedPMC
• 6. Hong YP, Kim SJ, Lee EG, Lew WJ, Bai JY. Treatment of bacillary pulmonary tuberculosis at the chest clinics in the private sector in Korea, 1993. Int J Tuberc Lung Dis 1999;3(8):695-702PubMed
• 7. Son H, Mok J, Lee M, Park W, Kim S, Lee J, et al. Status and determinants of treatment outcomes among new tuberculosis patients in South Korea: a retrospective cohort study. Asia Pac J Public Health 2021;33(8):907-913. https://doi.org/10.1177/10105395211000529ArticlePubMed
• 8. Semenza JC. Strategies to intervene on social determinants of infectious diseases. Euro Surveill 2010;15(27):32-39. https://doi.org/10.2807/ese.15.27.19611-enArticlePubMed
• 9. Yu S, Sohn H, Kim HY, Kim H, Oh KH, Kim HJ, et al. Evaluating the impact of the nationwide public-private mix (PPM) program for tuberculosis under National Health Insurance in South Korea: a difference in differences analysis. PLoS Med 2021;18(7):e1003717. https://doi.org/10.1371/journal.pmed.1003717ArticlePubMedPMC
• 10. Mori T, Kobayashi N. Tuberculosis treatment in Japan: problems and perspectives-how to expand the Japanese version of DOTS. Jpn Med Assoc J 2009;52(2):112-116
• 11. National Institute for Health and Care Excellence (UK). Tuberculosis: prevention, diagnosis, management and service organization; 2016 [cited 2025 May 19]. Available from: https://www.nice.org.uk/guidance/ng33/resources/tuberculosis-prevention-diagnosis-management-and-service-organisation-pdf-2910851744965
• 12. Lynn MR. Determination and quantification of content validity. Nurs Res 1986;35(6):382-385ArticlePubMed
• 13. Polit DF, Beck CT, Owen SV. Is the CVI an acceptable indicator of content validity? Appraisal and recommendations. Res Nurs Health 2007;30(4):459-467. https://doi.org/10.1002/nur.20199ArticlePubMed
• 14. World Health Organization. Definitions and reporting framework for tuberculosis – 2013 revision: updated December 2014 and January 2020; 2013 [cited 2025 May 19]. Available from: https://iris.who.int/handle/10665/79199
• 15. Siroka A, Ponce NA, Lönnroth K. Association between spending on social protection and tuberculosis burden: a global analysis. Lancet Infect Dis 2016;16(4):473-479. https://doi.org/10.1016/S1473-3099(15)00401-6ArticlePubMed
• 16. Richterman A, Steer-Massaro J, Jarolimova J, Luong Nguyen LB, Werdenberg J, Ivers LC. Cash interventions to improve clinical outcomes for pulmonary tuberculosis: systematic review and meta-analysis. Bull World Health Organ 2018;96(7):471-483. https://doi.org/10.2471/BLT.18.208959ArticlePubMedPMC
• 17. Oliosi JG, Reis-Santos B, Locatelli RL, Sales CM, da Silva Filho WG, da Silva KC, et al. Effect of the Bolsa Familia Programme on the outcome of tuberculosis treatment: a prospective cohort study. Lancet Glob Health 2019;7(2):e219-e226. https://doi.org/10.1016/S2214-109X(18)30478-9ArticlePubMed
• 18. Tucker A, Mithoo J, Cleary P, Woodhead M, MacPherson P, Wingfield T, et al. Quantifying the need for enhanced case management for TB patients as part of TB cohort audit in the North West of England: a descriptive study. BMC Public Health 2017;17(1):881. https://doi.org/10.1186/s12889-017-4892-5ArticlePubMedPMC
• 19. Katsuda N, Hirosawa T, Reyer JA, Hamajima N. Roles of public health centers (Hokenjo) in tuberculosis control in Japan. Nagoya J Med Sci 2015;77(1-2):19-28PubMedPMC
• 20. Kim H, Choi H, Yu S, Lee AY, Kim HO, Joh JS, et al. Impact of housing provision package on treatment outcome among homeless tuberculosis patients in South Korea. Asia Pac J Public Health 2019;31(7):603-611. https://doi.org/10.1177/1010539519871779ArticlePubMed
• 21. Uchimura K, Ngamvithayapong-Yanai J, Kawatsu L, Ohkado A, Yoshiyama T, Ito K, et al. Permanent employment or public assistance may increase tuberculosis survival among working-age patients in Japan. Int J Tuberc Lung Dis 2015;19(3):312-318. https://doi.org/10.5588/ijtld.14.0104ArticlePubMed
• 22. Kim J, Han J, Kwon Y, Choi H. Treatment outcomes among drug-susceptible patients with tuberculosis in the Republic of Korea, 2019-2020. Public Health Wkly Rep 2023;16(12):329-353. (Korean). https://doi.org/10.56786/PHWR.2023.16.12.1Article

Figure & Data

References

Citations

Citations to this article as recorded by