People-centered tuberculosis care versus standard directly observed therapy: study protocol for a cluster randomized controlled trial

Development and assessment of novel machine learning models to predict medication non-adherence risks in type 2 diabetics

Background

Medication adherence is the main determinant of effective management of type 2 diabetes, yet there is no gold standard method available to screen patients with high-risk non-adherence. Developing machine learning models to predict high-risk non-adherence in patients with T2D could optimize management.

Methods

This cross-sectional study was carried out on patients with T2D at the Sichuan Provincial People's Hospital from April 2018 to December 2019 who were examined for HbA1c on the day of the survey. Demographic and clinical characteristics were extracted from the questionnaire and electronic medical records. The sample was randomly divided into a training dataset and a test dataset with a radio of 8:2 after data preprocessing. Four imputing methods, five sampling methods, three screening methods, and 18 machine learning algorithms were used to groom data and develop and validate models. Bootstrapping was performed to generate the validation set for external validation and univariate analysis. Models were compared on the basis of predictive performance metrics. Finally, we validated the sample size on the best model.

Results

This study included 980 patients with T2D, of whom 184 (18.8%) were defined as medication non-adherence. The results indicated that the model used modified random forest as the imputation method, random under sampler as the sampling method, Boruta as the feature screening method and the ensemble algorithms and had the best performance. The area under the receiver operating characteristic curve (AUC), F1 score, and area under the precision-recall curve (AUPRC) of the best model, among a total of 1,080 trained models, were 0.8369, 0.7912, and 0.9574, respectively. Age, present fasting blood glucose (FBG) values, present HbA1c values, present random blood glucose (RBG) values, and body mass index (BMI) were the most significant contributors associated with risks of medication adherence.

Conclusion

We found that machine learning methods could be used to predict the risk of non-adherence in patients with T2D. The proposed model was well performed to identify patients with T2D with non-adherence and could help improve individualized T2D management.

Preparedness of outpatient health facilities for ambulatory treatment with all-oral short DR-TB treatment regimens in Zhytomyr, Ukraine: a cross-sectional study

BACKGROUND

Ukraine has a high burden of drug-resistant tuberculosis (DR-TB). Mental health problems, including alcohol use disorder, are common co-morbidities. One in five DR-TB patients has human immunodeficiency virus (HIV). As part of health reform, the country is moving from inpatient care to ambulatory primary care for tuberculosis (TB). In Zhytomyr oblast, Médecins Sans Frontières (MSF) is supporting care for DR-TB patients on all-oral short DR-TB regimens. This study describes the preparedness of ambulatory care facilities in Zhytomyr oblast, Ukraine, to provide good quality ambulatory care.

METHODS

This is a retrospective analysis of routinely collected programme data. Before discharge of every patient from the hospital, MSF teams assess services available at outpatient facilities using a standardised questionnaire. The assessment evaluates access, human resources, availability of medicines, infection control measures, laboratory and diagnostic services, and psychosocial support.

RESULTS

We visited 68 outpatient facilities in 22 districts between June 2018 and September 2019. Twenty-seven health posts, 24 TB-units, 13 ambulatories, two family doctors and one polyclinic, serving 30% of DR-TB patients in the oblast by September 2019, were included. All facilities provided directly observed treatment, but only seven (10%) provided weekend-services. All facilities had at least one medical staff member, but TB-training was insufficient and mostly limited to TB-doctors. TB-treatment and adequate storage space were available in all facilities, but only five (8%) had ancillary medicines. HIV-positive patients had to visit a separate facility to access HIV-care. Personal protective equipment was unavailable in 32 (55%) facilities. Basic laboratory services were available in TB-units, but only four (17%) performed audiometry. Only ten (42%) TB-units had psychosocial support available, and nine (38%) offered psychiatric support.

CONCLUSION

Outpatient facilities in Zhytomyr oblast are not yet prepared to provide comprehensive care for DR-TB patients. Capacity of all facilities needs strengthening with trainings, infection control measures and infrastructure. Integration of psychosocial services, treatment of co-morbidities and adverse events at the same facility are essential for successful decentralisation. The health reform is an opportunity to establish quality, patient-centred care.

People-centred care versus clinic-based DOT for continuation phase TB treatment in Armenia: a cluster randomized trial

Background

WHO’s directly observed therapy (DOT) strategy for tuberculosis (TB) treatment depends upon a well–organized healthcare system. This study sought to evaluate the effectiveness of self-administered drug intake supported by a family member versus in-clinic DOT.

Methods

This open–label, nationally-representative stratified cluster randomized controlled non–inferiority trial with two parallel equal arms involved drug–susceptible pulmonary TB patients in the continuation treatment phase. We randomly assigned outpatient–TB–centres (52 clusters) to intervention and control arms. The intervention included an educational/counseling session to enhance treatment adherence; weekly visits to outpatient–TB–centres to receive medication, and daily SMS medication reminders and phone calls to track adherence and record side effects. Controls followed clinical DOT at Outpatient–TB–centres. Both groups participated in baseline and 4–5 months follow–up surveys. The trial’s non–inferiority comparisons include: treatment success as the clinical (primary) outcome and medication adherence (self–reported), knowledge, depressive symptoms, stigma, quality of life, and social support as non–clinical (secondary) outcomes.

Results

Per–protocol analysis showed that the intervention (n = 187) and control (n = 198) arms achieved successful treatment outcome of 92.0 and 92.9%, respectively, indicating that the treatment success in the intervention group was non–inferior to DOT. Knowledge, depression, stigma, quality of life, and social support also showed non–inferiority, demonstrating substantial improvement over time for knowledge (change in the intervention = 1.05: 95%CL (0.49, 1.60); change in the control = 1.09: 95%CL (0.56, 1.64)), depression score (change in the intervention = − 3.56: 95%CL (− 4.99, − 2.13); change in the control = − 1.88: 95% CL (− 3.26, − 0.49)) and quality of life (change in the intervention = 5.01: 95%CL (− 0.64, 10.66); change in the control = 7.29: 95%CL (1.77, 12.81)). The intervention resulted in improved treatment adherence.

Conclusions

This socially empowering alternative strategy might be a preferable alternative to DOT available to patients in Armenia and in other countries. Further research evaluating cost effectiveness of the intervention and generalizability of the results is warranted.

Trial registration

Clinicaltrials.gov: NCT02082340, March 10, 2014.

ImPROving TB outcomes by modifying LIFE-style behaviours through a brief motivational intervention followed by short text messages (ProLife): study protocol for a randomised controlled trial

Background

South Africa is among the seven highest tuberculosis (TB) burden countries. Harmful lifestyle behaviours, such as smoking and alcohol, and poor adherence to medication can affect clinical outcomes. Modification of these behaviours is likely to improve TB treatment outcomes and has proven possible using motivational interviewing (MI) techniques or use of short message service (SMS) text messaging. There have been no studies assessing the effect of combined MI and SMS interventions on multiple lifestyle factors and TB treatment outcomes.

Methods

This is a prospective, multicentre, two-arm individual randomised controlled trial looking at the effectiveness and cost-effectiveness of a complex behavioural intervention (the ProLife programme) on improving TB and lifestyle-related outcomes in three provinces of South Africa. The ProLife programme consists of an MI counselling strategy, delivered by lay health workers, augmented with subsequent SMS. We aim to recruit 696 adult participants (aged 18 years and over) with drug-sensitive pulmonary TB who are current smokers and/or report harmful or hazardous alcohol use. Patients will be consecutively enrolled at 27 clinics in three different health districts in South Africa. Participants randomised individually to the intervention arm will receive three MI counselling sessions one month apart. Each MI session will be followed by twice-weekly SMS messages targeting treatment adherence, alcohol use and tobacco smoking, as appropriate. We will assess the effect on TB treatment success, using standard World Health Organization (WHO) treatment outcome definitions (primary outcome), as well as on a range of secondary outcomes including smoking cessation, reduction in alcohol use, and TB medication and anti-retroviral therapy adherence. Secondary outcomes will be measured at the three-month and six-month follow-ups.

Discussion

This trial aligns with the WHO agenda of integrating TB care with the care for chronic diseases of lifestyle, such as provision of smoking cessation treatments, and with the use of digital technologies. If the ProLife programme is found to be effective and cost-effective, the programme could have significant implications for TB treatment globally and could be successfully implemented in a wide range of TB treatment settings.

Trial registration

ISRCTN Registry, ISRCTN62728852. Registered on 13 April 2018.

Electronic supplementary material

The online version of this article (10.1186/s13063-019-3551-9) contains supplementary material, which is available to authorized users.

A systematic review of non-pharmacological interventions to improve therapeutic adherence in tuberculosis

BACKGROUND

Reviews examining non-pharmacological interventions to improve therapeutic adherence in tuberculosis have several limitations (design, quality assessment…). Consequently, for clinical practice, it is important to generate a review containing all the information to improve patient adherence, solving the previous issues.

OBJECTIVES

To examine non-pharmacological interventions to improve therapeutic adherence in tuberculosis through clinical trials.

METHODS

A systematic review in MEDLINE/EMBASE was performed.

RESULTS

Thirty seven papers were analysed. The disease treatment interventions were disparate, grouped into: education, psychological interventions, new technologies, directly observed treatment, incentives and improved access to health services. In the treatment of latent infection, the majority of studies were conducted in the marginal population (drug addicts, homeless individuals and prisoners) and were based mainly on the provision of incentives. Study quality was generally low.

CONCLUSIONS

Great variability exists in the studies comparing strategies for identifying interventions, objectives and effects. The designs carried out generally have methodological deficits.

Toward Developing a Standardized Core Set of Outcome Measures in Mobile Health Interventions for Tuberculosis Management: Systematic Review

Background

Tuberculosis (TB) management can be challenging in low- and middle-income countries (LMICs) not only because of its high burden but also the prolonged treatment period involving multiple drugs. With rapid development in mobile technology, mobile health (mHealth) interventions or using a mobile device for TB management has gained popularity. Despite the potential usefulness of mHealth interventions for TB, few studies have quantitatively synthesized evidence on its effectiveness, presumably because of variability in outcome measures reported in the literature.

Objective

The aim of this systematic review was to evaluate the outcome measures reported in TB mHealth literature in LMICs.

Methods

MEDLINE, EMBASE, and the Cochrane Database of Systematic Reviews were searched to identify mHealth intervention studies for TB (published up to May 2018) that reported any type of outcome measures. The extracted information included the study setting, types of mHealth technology used, target population, study design, and categories of outcome measures. Outcomes were classified into 13 categories including treatment outcome, adherence, process measure, perception, technical outcome, and so on. The qualitative synthesis of evidence focused on the categories of outcome measures reported by the type of mHealth interventions.

Results

A total of 27 studies were included for the qualitative synthesis of evidence. The study designs varied widely, ranging from randomized controlled trials to economic evaluations. A total of 12 studies adopted short message service (SMS), whereas 5 studies used SMS in combination with additional technologies or mobile apps. The study populations were also diverse, including patients with TB, patients with TB/HIV, health care workers, and general patients attending a clinic. There was a wide range of variations in the definition of outcome measures across the studies. Among the diverse categories of outcome measures, treatment outcomes have been reported in 14 studies, but only 6 of them measured the outcome according to the standard TB treatment definitions by the World Health Organization.

Conclusions

This critical evaluation of outcomes reported in mHealth studies for TB management suggests that substantial variability exists in reporting outcome measures. To overcome the challenges in evidence synthesis for mHealth interventions, this study can provide insights into the development of a core set of outcome measures by intervention type and study design.

Nursing in adherence to treatment of tuberculosis and health technologies in the context of primary care

Abstract Objective: To describe and analyze the relationships between adherence to tuberculosis treatment and health technologies in the context of nursing action in Primary Care. Method: Contextual reflection, using the theoretical reference of Hinds, Chaves and Cypress. Results: They were categorized according to the conceptual perspective of each context, in an immediate, specific, general and metacontext, respectively: nursing actions in Primary Care and adherence to treatment of tuberculosis; nurses' performance through personal and environmental factors, health technologies and adherence to treatment; influence of nurses' beliefs about adherence to treatment and overall health status, and use of health technologies and shared views of patients and nurses on adherence to treatment as a responsibility for preventive action. Conclusion and Implications for practice: Adherence to the treatment of tuberculosis is directly linked to the substantial performance of the nurse, which presents potentialities to contribute to greater articulation between actions necessary to the success of the treatment, reducing the weaknesses in its operationalization. The technologies in the context of nursing action can favor praxis, especially in the incentive to join, and can subsidize new strategies appropriate to the reality of services.

To End TB, First-Ever High-Level Meeting on Tuberculosis Must Address Stigma

World leaders gather to consolidate their commitment to ending tuberculosis (TB). Vital to the success of renewed efforts is an overdue recognition of the pervasive and pernicious influence of TB stigma. TB stigma is sustained in structures, policies, traditions, and norms. Innovative modifications to infection control, drug dispensing, and surveillance practices are required to increase demand for TB screening and effective therapeutic alliances among those diagnosed. The authors argue that reducing TB stigma requires a scientific and inclusive process, with prominent roles for TB survivors and a willingness to integrate and learn from other stigmatized conditions.

Treatment of tuberculosis infection in children

INTRODUCTION

Identifying and treating children with tuberculosis (TB) infection in both low and high-TB burden settings will decrease the incidence of TB disease worldwide. Areas covered: This review covers each of the available TB infection treatment options for children based on effectiveness, safety, tolerability and treatment completion rates. Six to 9 months of daily administered isoniazid is no longer the treatment of choice for many children with TB infection. Shorter, rifamycin based, TB infection treatment regimens are effective, safe and easier for children to complete. Fluroquinolone-based regimens are recommended for the treatment of children infected by a source case with drug-resistant TB. Directly observed therapy (DOT) programs improve childhood TB infection treatment completion rates. Expert commentary: As shorter, rifamycin-based, TB infection treatment regimens offer superior treatment success rate in both adults and children; the widespread use of these regimens has huge potential to decrease the burden of TB disease worldwide. The implementation of these programs will involve improving patient access to the medications, decreasing their cost to the patient, and the use of novel electronic methods to document patient treatment completion.

Evaluation of a medication monitor-based treatment strategy for drug-sensitive tuberculosis patients in China: study protocol for a cluster randomised controlled trial

BACKGROUND

Treatment for drug-sensitive tuberculosis (TB) is taken for at least 6 months and problems with adherence are common. Therefore, there is substantial interest in the possible use of eHealth interventions to support patients to take their treatment. Electronic medication monitors have been shown to improve adherence to TB medication, but the impact on clinical outcomes is unknown. We aim to evaluate the impact of a medication monitor-based treatment strategy for drug-sensitive TB patients on a composite poor outcome measured over 18 months from start of TB treatment.

METHODS/DESIGN

We will conduct an open, pragmatic, cluster randomised superiority trial, with 24 counties/districts in three provinces in China, randomised 1:1 to implement the intervention or standard of care. Adults (aged ≥ 18 years) with a new episode of GeneXpert-positive and rifampicin-sensitive pulmonary TB, who plan to be in the study area for the next 18 months, and will receive daily fixed-dose combination tablets for 6 months of treatment are eligible. The intervention is centred around a medication monitor that holds a 1-month supply of medication and has three key functions: as an audio and visual reminder for patients to take their daily medication; reminds patients of upcoming monthly visit; and records date and time whenever the box is opened. At the monthly follow-up visit, the doctor downloads these data to generate a graphical display of the last month's adherence record for discussion with the patient and potentially to switch the patient to more intensive management. The primary outcome is a composite poor outcome measured over 18 months from start of TB treatment, defined as either of poor outcome at the end of treatment (death, treatment failure, or loss to follow-up) or subsequent recurrence (culture positive for TB at 12 or 18 months or re-starting TB treatment in the follow-up period). An economic evaluation will also be conducted as part of this study.

DISCUSSION

This trial will assess whether a medication monitor-based treatment strategy can improve clinical outcomes for TB patients. Several trials of other eHealth interventions for TB treatment are ongoing and are summarised in this paper. This trial will provide an important part of the emerging evidence base for the potential of eHealth to improve TB treatment outcomes.

TRIAL REGISTRATION

This trial was registered with Current Controlled Trials (identifier: ISRCTN35812455 ). Registered on May 19, 2016.