Creating evidence-based optimal testing strategies for monitoring long-term conditions in primary care

Objectives

Patients with long term conditions (LTC) have regular monitoring appointments including blood tests. These tests aim to monitor disease progression, response to treatment, and detection of complications. The evidence base for current testing recommendations is weak because measuring patient benefits or harms of regular monitoring is challenging and are dependent on what is done in response to the test result. The aim of the Optimal Testing project is to develop a methodology for creating evidence-based testing strategies to monitor people with LTCs and accompanying patient and clinician materials.

Methods

We identified a list of commonly used tests. We defined a series of filtering questions to determine whether there was evidence to support the rational of monitoring, such as ‘Is this patient population at greater risk than the general population to get a complication which can be picked up by a candidate test’, and ‘can the GP do anything in response to an abnormal test result?’. Through a series of rapid reviews we identified evidence to answer each question. The evidence was presented at a consensus meeting where clinicians and patients voted for inclusion, exclusion, or further analysis. A process evaluation was performed alongside this. Further analyses are performed using routinely collected healthcare data, by performing incidence analyses and emulating RCTs. The optimal frequency of monitoring will be determined by disease progression modelling and health economic modelling.

Results 

We tested this methodology on three common LTCs chronic kidney disease (CKD), type 2 diabetes mellitus (T2DM), and hypertension. We found sufficient evidence to include HbA1c and eGFR for monitoring T2DM patients; haemoglobin and eGFR for CKD; and eGFR for hypertension. The consensus panel voted to exclude haematinics, clotting tests, brain natriuretic peptide, c-reactive protein, and erythrocyte sedimentation rate. However, for the majority of tests, there was not enough evidence to include or exclude from the testing panel, which were selected for further analyses. Incidence analyses suggested patients with CKD and T2DM do not have a higher risk of abnormal thyroid function or haemoglobin, respectively, than matched controls and monitoring is therefore not necessary. The emulated RCTs aim to investigate the effect of regular testing with certain tests on patient outcomes among routinely monitored patients (ongoing). The final testing panel will be decided in an upcoming consensus meeting.

Conclusions 

Many tests that are currently done in primary care may not have a strong rationale and may not lead to any patient benefit. The methodology we have developed may be used to optimise disease monitoring of other chronic conditions. We currently developing an intervention package and are planning to run a feasibility trial to test the cost-effectiveness of the evidence-based testing panels. This programme of work has the potential to change how LTCs are monitored in primary care, ultimately improving patient outcomes, and leading to more efficient use of NHS resources.