UCLH to deliver innovative new study using electronic health records

An innovative new study at UCLH is investigating the feasibility of delivering a clinical trial in routine care using the electronic health record system (EHRS).

The ‘THIRST Alert’ study aims to evaluate the effectiveness of restricting the amount of fluid that patients drink if they are being treated for fluid overload. The common causes of fluid overload include heart failure, liver disease and kidney disease. It is uncertain whether fluid restriction is beneficial for these patients or if the practice may lead to increased thirst.

In cases where there is uncertainty about whether fluid restriction will cause benefit or harm, an EHRS alert will be sent to the treating clinician. This will happen at the point when the clinician goes to prescribe a diuretic in response to fluid overload.

The alert will highlight the uncertainty around best practice for treatment of fluid overload, and the clinician will be asked to consider if the patient would be suitable for the THIRST Alert study. If the clinician decides the patient is suitable, the patient will be randomised to one of two ‘arms’ of the study: in one arm, oral fluid will be restricted by 1 litre per day. In the other arm, there will be no fluid restriction.

In this initial feasibility study, the research team led by Dr Yang Chen, Dr Anoop Shah, and Dr Tom Lumbers at the UCLH Clinical and Research Informatics Unit (CRIU) will test whether an automatic clinician-facing alert with verbal consent is an effective means of enrolling patients into the randomised controlled trial (RCT).

Using data collected in routine care, the study will measure the number of patients recruited and the effects of the alert on clinician behaviour and patients’ oral fluid intake. THIRST Alert is a ‘pragmatic’ RCT – where the intervention given is part of routine NHS care rather than within a formal research setting. The study was designed and implemented with support from the CRIU, led by CRIU Director, Prof Folkert Asselbergs.

Traditional RCT designs are challenging to implement in the acute setting in a way that is cost-effective and generalisable. By harnessing the power of the electronic health record system for patient identification, point of care randomisation and collection of study outcomes, THIRST Alert will serve as a pathfinder study exploring ways to overcome barriers to research participation and evidence generation in routine care. 

If successful, THIRST Alert will establish a new approach to enrolling patients in clinical trials and may generate evidence on whether oral fluid restriction is a helpful treatment or not.

Visit the CRIU web pages for updates on the progress of the study and more information.