Clinical Use of Urine Biochemical Testing as a Screening Method for Non-adherence in a Hypertension Clinic

MR Pucci1,2, U Martin1,2. 1University of Birmingham, Birmingham, UK, 2University Hospitals Birmingham NHS Foundation Trust, Birmingham, UK

Hypertension is a major contributor to the development of cardiovascular disease. Although effective and well-tolerated antihypertensive medications are available, poor adherence with recommended treatments continues to be one of the main causes of unsatisfactory blood pressure control (1). Measurement of adherence is fraught with difficulties. Indirect methods such as pill counts and patient diaries have been shown to overestimate adherence. Medication event monitoring systems have been validated for use in clinical trials but are not suitable for routine clinical use. Direct methods such as directly observed pill taking are labour intensive and plasma testing has failed to find broad clinical use. However, urine biochemical testing by high performance liquid chromatography-tandem mass spectrometry (HPLC-MS) has recently become easily available and is inexpensive (2).

We present our experience of urine biochemical testing in a specialist hypertension clinic in an NHS tertiary referral centre. Suitably identified patients with uncontrolled hypertension were asked to provide a random urine sample for analysis following consultation with a specialist. Samples were sent to Leicester Royal Infirmary and analysed using HPLC-MS for the presence of antihypertensive drugs or their metabolites. Fifty patients were included in the analysis. The main reasons for testing were suspected non-adherence (33 cases) or to exclude non-adherence in cases of resistant hypertension (17 cases including 4 being worked-up for renal denervation). The mean number of antihypertensive drugs prescribed per patient was 4.46 (range 2-10). The mean number of drugs detected in the urine was 2.54 (range 0-6). Thirteen patients (26%) were completely non-adherent (no antihypertensive drugs detected), 17 patients (34%) were partially non-adherent (some of the antihypertensive drugs detected), with a combined total of 30 patients (60%) being completely or partially non-adherent. Only 20 patients (40%) were fully adherent with their medication.

Upon discussion of their results in clinic, 11 patients denied non-adherence and disputed the result, and 9 patients admitted non-adherence, most commonly due to complaints of adverse drug reactions. Even when patients denied non-adherence, most were open to starting afresh with their treatment, with fewer tablets or fixed-dose drug combinations. In 4 cases, non-adherence was attributed to a language barrier. Reasons for non-adherence were unclear in other cases: 3 patients did not attend follow-up, and in the remainder it was yet to be addressed. Out of the most commonly prescribed drugs, the poorest adherence rates were seen with the diuretics, spironolactone, indapamide (adherence rate for both only 42.9%) and bendroflumethiazide (adherence rate 47.1%).

This observational study confirms that non-adherence in patients with uncontrolled hypertension is common, with 60% of patients in this cohort not taking their antihypertensive medication as prescribed. Urine biochemical testing is an effective method of screening for non-adherence in patients with uncontrolled hypertension. Once identified, treatment approaches can be individualised with emphasis on methods of maintaining good adherence, such as simplification of the drug regimen.

(1) Naderi SH et al. (2012). Am J Med. 125: 882-887.

(2) Tomaszewski et al. (2014). Heart. 100: 855-61.