Liver Function Testing (LFT) in Primary Care: Retrospective Case Note Review of the Management of Patients with Abnormal LFTs

The problem

Liver disease is fifth largest killer in UK and there has been an exponential rise in LFT testing. Abnormal LFTs are common but often do not represent serious liver disease, with 22% of ‘new' LFTs checked locally in a previous published study having ?1 abnormality. There is known widespread variation in laboratory usage and planned changes to laboratory reporting of LFTs and an automatic triggering of new care pathway have been developed.

The approach

In order to inform design of proposed new laboratory reporting and care pathway, this study set out to examine the indications for LFT testing in primary care to assess the prevalence of abnormal results and their current management whilst assessing the feasibility of implementing validated algorithms for calculating risk of death. To do this a retrospective case note review of 100 random adult patients who had LFTs taken in primary care and found to have an abnormal result in 2013 was performed across 2 practices. Data was extracted from records to identify test indication, whether the abnormality was new, what follow-up was performed, and whether it was possible to calculate risk scores for predicted mortality within 1 year and predicted probability of formal liver disease diagnosis within 24 months.

Findings

Of the 100 patients with abnormal LFTs notes studied 28 had these taken for being unwell, tired all the time or abdominal pain, 60 as part of planned monitoring such as diabetes reviews, and 12 for suspected alcohol abuse, abnormal LFTs being repeated, secondary care requests or no reason documented. 13 patients had newly abnormal LFTs. In the general practice notes 17 patients had a suspected diagnosis accounting for LFT abnormalities documented. For follow up 52 were filed as satisfactory, 17 patients were seen in GP consultations, 14 in nurse consultations, 7 had a telephone consultation with a GP, 1 had a telephone consultation with a nurse, 4 were repeated and 5 had a message left via reception. 7 patients had additional liver blood tests performed. Risk scores were generated in 98% of cases for predicted mortality and 100% of the cases for the probability of liver disease diagnosis. 18 cases had greater than 1% chance of being diagnosed with a serious liver disease within 24 months (average 2.3%).

Consequences

Majority of abnormal LFTs in the samples are for planned monitoring, the introduction of a semi-automated laboratory pathway may be of value. It is feasible to implement risk prediction tools based on information in GP record but utility unclear. New semi-automated care pathway introduction needs evaluating if implemented as planned.

Credits

• Christopher Weatherburn
• Bruce Guthrie