Non-adherence to treatment: a methodological approach to compensating for its impact in chronic infections, using tuberculosis as a model disease.

Imperfectly designed treatment regimens for chronic (i.e. long-lasting) diseases result in continued, and possibly worsened, symptoms. When a disease can be passed from person-to-person we additionally risk the development and spread of drug resistance, which is when a bacterium or virus becomes untreatable by normal medicines as it has genetically changed. This endangers both individual- and population-level health. If someone incompletely 'adheres' to their treatment (does not take all their medicine as recommended) this can result in the same problems, but the impact of this depends on the drugs being used, due to their differing properties. Poor adherence to treatment costs hundreds of billions of dollars globally each year.

For many chronic diseases, we do not know how much medicine it is safe for patients not to take and still be cured. Taking at least eight of every 10 doses across the entire length of treatment is often said to be important, but this lacks good evidence and is highly simplistic, especially given how much drugs vary from disease to disease.

This study aims to use different methodologies and real-life, detailed, adherence and treatment outcome data to decide how normal patient behaviour should influence our approach to treatment. Tuberculosis (TB) has been chosen as a model disease due to its global importance (10.4 million new patients, 1.4 million related deaths in 2015) and the need for better treatments. The Objectives are:
1) Describe how, when and why poor adherence occurs during the six months of daily treatment, and how adherence is related to the outcome of treatment across different places.
2) Determine how adherence patterns influence the development of drug resistance over time.
3) Adapt a theoretical model to examine the impact of adherence on the best way to give treatment (e.g. slow release tablets).
4) Draw together these strands of evidence to determine how adherence should alter our approach to treatment and how new regimens should be designed.
This work will not only guide the best way of treating TB globally, but the methods used will also help to interpret how common adherence patterns should influence the use of drugs for other conditions in the real world. The best ways of dosing TB drugs can be included in new clinical trials to improve health services for patients and value for money. The way in which drug trial data is interpreted as reflecting real life by the people writing clinical guidance will be refined. The findings can also inform trials of tools to promote adherence.

Non-adherence to treatment costs hundreds of billions of dollars globally per year and is particularly problematic in chronic diseases, which require lengthy treatment. In addition to persistence of symptoms and increased disease severity, poor adherence in communicable chronic diseases risks the development of drug resistance, jeopardising both individual and population-level health.

Adherence guidelines frequently tacitly assume that all doses must be taken for successful treatment. Clinical trials often set simple thresholds- e.g. 80% of doses taken- to define adherence. The reality is complex; granular analyses of the relationship between adherence (percentage of overall duration taken, patterns of missed doses, percentage of doses taken) and treatment outcomes are critical for effective translation of trials into clinical guidance.

This study aims to utilise different methodological approaches to assess the impact of evidence-based adherence patterns on treatment outcomes and inform how this should influence our approach to treatment. A model disease (tuberculosis) has been chosen on the basis of its global importance and the need for improved regimens. The Objectives and methods are:
1) Describe statistically how, when and why non-adherence occurs and the relationship between adherence and outcomes across different settings.
2) Determine how adherence patterns influence the temporal development of drug resistance.
3) Use an adaptable reaction kinetic model to assess the impact of treatment adherence on optimal dosing.
4) Draw together these strands of evidence to determine how adherence should influence our approach to treatment.

This work will advance the scientific community's thinking when designing future clinical trials. The disease-adaptable approach to determine how adherence influences a drug's 'operational efficacy' will change how guidance is written. Its methodologies will be usable for a wide array of chronic diseases.

The key beneficiaries of this work will be: tuberculosis (TB) patients and members of the public, the study participants and their broader communities, health workers, policymakers, other researchers, and global economies.

ACADEMIC SECTOR: in addition to furthering knowledge and building international collaborations, within the five year duration of the project the CDA will produce a Research Fellow with epidemiological and statistical research and professional skills, a second Research Fellow with additional experience in bioinformatics, and a research nurse in the Republic of Korea with expertise in working on an international study, thus benefitting the academic and research sectors. In addition, as the Chief Investigator, I will end the study as a recognised academic within my research niche.

ENHANCED COMMUNITY HEALTH: study participants and the patient communities from which they arise will be key stakeholders in the project. As well as the anticipated direct benefits to the included patients in terms of the debrief at the end of the project and thus greater awareness of the importance of treatment adherence (for all diseases, knowledge of which they will take back to their communities), patient and public involvement from the communities of participants will empower individuals and provide informed citizens to the benefit of society (end of study).
The results of the study, once adopted into practise (10 years) will reduce morbidity and mortality in TB (which has a substantial burden globally), the development of secondary drug resistance and the transmission of both drug resistant and drug sensitive strains. This is because policymakers will be able to write better evidence-drive guidance on when it is most important to intervene to promote adherence and in whom. They will also be in an improved position to have the tools to accurately interpret how data arising from clinical trials represents how a treatment regimen will perform in real healthcare settings (operational efficacy), which will influence how old and new regimens are introduced and ranked based on their expected effectiveness and cost-effectiveness.
This work will also help guide future trials of adherence-promoting measures (targeted to the critical stages of treatment and populations most at risk of the worst patterns of non-adherence; 5-15 years) and the design and trial of treatment regimens with greater forgiveness e.g. different dosing strategies (10-15 years).
The methods used in the study can be expanded to aid the treatment of other communicable conditions, initially hepatitis (15 years) and then other diseases, further benefitting global health.

INCREASED EFFECTIVENESS OF HEALTH SERVICES AND POLICY: healthcare workers involved peripherally with the project will benefit from insight into the running of a research study and its influence on national guidance (throughout study). Staff in health-related disciplines working on the project will have their research and associated skills (communication, writing of publications, statistics) developed. These transferable skills will undoubtedly be useful across many fields during future work (from end of study). Ministries of health and international agencies will benefit from data regarding how to prevent non-adherence (end of study), thus reducing medication wastage and arising drug resistance, which can be more expensive to treat. I will seek to engage both groups throughout the grant to support two-way dialogue (throughout study).

FOSTERED GLOBAL ECONOMIC PERFORMANCE: preventable morbidity and mortality, particularly among the young, substantially hinders global economic development. In many nations, TB is a disease of populations of working age, proving a huge burden to developing economies, particularly in countries that cannot yet afford universal healthcare.