The funder and the study team have mutually agreed to terminate the study considering the difficulties encountered in recruitment and consequently the impossibility to perform the planned analysis.

Non-adherence is defined as: "the extent to which a person's behaviour - taking medication, following a diet, and/or executing lifestyle changes, corresponds with agreed recommendations from a health care provider". Non-adherence in chronic cardiometabolic diseases including diabetes is very common and is often the primary reason for treatment failure. This leads to significant excess costs to the health economy through avoidable investigations, treatment escalations, hospital admissions, and disease complications.

Methods to diagnose non-adherence have until recently been poor. We have recently developed an objective and robust chemical adherence test to detect the presence of 160 cardiovascular medications in urine using high performance liquid chromatography-tandem mass spectrometry (LC-MS/MS). Chemical adherence testing has not been utilised in people with diabetes, further its relationship with other measures of adherence is unknown.

The main aim of this observational study is to compare chemical non-adherence with other commonly used measures of non-adherence in people with diabetes. Chemical testing for non-adherence will be performed using urine provided by 600 patients with poorly controlled diabetes attending primary care recruited over a 15-month period.

Participants will also be required to complete a self- reported questionnaire and pharmacy records will be reviewed to ascertain prescription refill rates. The prevalence and metabolic control of non-adherence as diagnosed chemically will be compared with those obtained by pharmacy refill rates and patient self-reported questionnaires. Further, the determinants of non-adherence as ascertained by urine LC-MS/MS analysis will be studied.

It is hoped that this innovative study will lead to further larger intervention studies that will change the management of non-adherence in diabetes.

Non-adherence is defined as: "the extent to which a person's behaviour - taking medication, following a diet, and/or executing lifestyle changes, corresponds with agreed recommendations from a health care provider". Adherence to chronic disease medications remains poor and it is estimated that around one third to half of all patients become nonadherent. Medication non-adherence is a risk factor for adverse clinical outcomes, including hospital admissions, and costs the U.S. health care system over $100 billion annually. In the UK, it is estimated that up to 50% of patients do not take their prescribed medications and its costs the NHS £4 billion annually.

Non-adherence has been measured by various methods including patient self-reported questionnaires and by looking at pharmacy refill records and electronic monitoring devices (EMDs). There was until recently the lack of a clinically useful tool to assess non-adherence in a simple, reliable method. We have recently developed an innovative, first in the world chemical test to objectively ascertain adherence by testing for presence of medications in a spot urine sample. This test uses a high-performance liquid chromatography tandem mass spectrometry (LC-MS/MS) analysis of spot urine samples to detect 160 commonly prescribed cardiovascular medications including those for diabetes. The method is robust and reliable and has been developed from techniques used in Forensic Medicine. We, in Leicester, utilise the test routinely in the Hypertension Clinic and have set up the first National Centre for Adherence Testing (NCAT). We now receive samples from around 40 centres across UK and have analysed more than 64000 samples to date. In our experience, it is well accepted by patients, it provides confidence to the clinician about non-adherence and helps initiate an objective discussion with patients. The test helps patients understand that the absence of medications in their body is linked with the suboptimal control of their illness. The test helps explore the reasons for non-adherence and tailor therapy to patient's needs. The use of the tests is now becoming accepted practice in hypertensive patients.

A meta-analysis that included more than 300,000 patients with diabetes showed that good adherence was associated with reduced all-cause mortality hospitalisation. The studies included in the meta-analysis used pharmacy refill records. Chemical testing for non-adherence has not been used to investigate the extent of non-adherence in diabetes or in patients in primary care.

Therefore, it would of utmost interest to confirm non-adherence rates using this method and compare it with commonly used tools such as pharmacy refill rates and patient self-reported measures of non-adherence. Overcoming this gap of knowledge is the primary aim of the study.

The study is designed as a prospective cohort study. We aim to study non-adherence in patients diagnosed with type 2 diabetes and plan to recruit 600 patients attending primary care in Leicestershire. Patients will be recruited if they have uncontrolled diabetes (HbA1c ≥ 7.5%) and are on at least one oral antidiabetic and one other cardiovascular medication that is part of our medication screen. This excludes aspirin and simvastatin as they are not on our panel.

The patients will be those attending their annual diabetes review in primary care. The urine and/or blood sample provided by patients as a part of their visit will be analysed for adherence testing. The sample will be collected by the patient's direct clinical care team and forwarded to the Department of Metabolic Medicine and Chemical Pathology (DMMCP). It will then be tested for the presence of prescribed medications by high performance liquid chromatography-tandem mass spectrometer (LC-MS/MS). There will be no extra samples obtained from patients.

Patients will also be asked to complete a medication adherence questionnaire which will ask questions about their medication taking behaviour. Patient data including demographic data, list of medications and co-morbidities, details of cardiovascular risk factors will be collected in a coded anonymised fashion. GP medication records will also be accessed by the patient's direct clinical care team. The laboratory analysts will be unaware of the clinical details of the patients. The results will not be sent back to the treating clinician or the patients. The results of the anonymised samples will be sent to the research team.

A patient research ID will be created by the patient's direct clinical care team at the GP practice after patients are consented and then samples then sent to the lab for analysis.

The study will take place over a duration of two years. Screening and recruiting of patients will take place over 9 months. Samples and data collection will be carried out in tandem and will be completed over 12 months. The data will then be analysed and reviewed before being written up as a report, which will take place over 3 months.

Inclusion Criteria:

• Aged 18 years and over
• Confirmed diagnosis of type 2 diabetes
• People with at least one HbA1c ≥ 7.5% (59mmol/mol) measured in the last 6 months
• Patients on at least one OHA and one other oral cardiovascular medication for blood pressure, diabetes or lipid lowering therapy
• Patients able to understand written and verbal English

Exclusion Criteria:

• Type 1 diabetes
• Age less than 18 years
• Patients unable to give informed consent
• Pregnant or planning pregnancy
• Terminally ill
• Patients on simvastatin and or aspirin if they are the only other medication for cardiovascular disease besides the OHAs that the patient is prescribed.
• Patients on insulin or GLP-1 analogues if they are the only other medication for cardiovascular disease that the patient is prescribed.
• Major medical or psychiatric illness
• Patients unable to understand written and verbal English

Male and female (non-pregnant nor breast feeding) subjects with uncontrolled diabetes attending their routine Diabetes review that are capable of giving informed consent, who are above 18 years of age, and have been prescribed at least one oral hypoglycaemic agent and one other medication for cardiovascular disease.