Aims With changing indications for performing medical liver biopsies, we aimed to develop a tool to allow pathologists to evaluate the current usefulness, value and impact of their medical liver biopsy service.
Methods We designed and piloted a questionnaire-based clinico-pathological audit for medical liver biopsies.
Results The audit tool was simple to implement and provided useful information about our service. Hepatologists felt that 96% of reports were clinically useful. 56% of biopsies confirmed clinical diagnoses, 46% helped differentiate between diagnoses and 42% were able to exclude possible diagnoses. 74% resulted in a change of management and 27% of liver biopsies resulted in a diagnosis which was not clinically suspected.
Conclusions We demonstrate the usefulness of an audit tool in providing evidence of the value of the liver pathology service in a large UK regional centre.
- QUALITY ASSURANCE
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Liver biopsy has traditionally been a mainstay and gold standard in the clinical evaluation of medical liver diseases.1 There are conflicting data over the number of medical liver biopsies being performed,2–4 but in the authors’ experience, numbers in the UK are increasing. Furthermore, there is debate over the changing indications for this investigation within the context of modern imaging techniques.5 ,6 While there is good evidence that this investigation is of high value,7 in light of changing UK and European guidelines8–10 there is a need for tools which clinicians and pathologists alike can use to evaluate their service locally.
Previous UK audit templates have not addressed the appropriateness of biopsies being performed or demonstrated the value of the reports generated.11–13 As part of an initiative of the UK Royal College of Pathologists (RCPath) to develop a series of clinical audit templates,14 we designed an audit tool to assess the usefulness of medical liver biopsies within an institution. Here we describe the development of this clinical effectiveness audit template and our pilot of it, including implementation, proposed standards and results.
Materials and methods
The audit is a prospective questionnaire-based review of adult medical liver biopsy reporting carried out at the participating institution. The questionnaire consists of nine tick-box (yes or no) questions which are completed by the hepatologists on a case-by-case basis (see table 1). Consecutive cases are advised. Referred cases are excluded to avoid problems coordinating with external hepatologists/pathologists. The questionnaire can be printed out, or used electronically, and is most conveniently filled in at around the time of the multidisciplinary team (MDT) meeting. The pilot took place at the John Radcliffe Hospital in Oxford, where the hepatopathology service receives around 500 liver biopsy cases annually (including referrals and mass lesions) and reports around 200 in-house medical liver biopsies per year. We audited 50 consecutive, new medical liver biopsies which were performed at our institution between September and December of 2014.
We found that the audit template was quick and easy to fill in and needed only minor changes in the wording and layout. We found that the most common reasons for biopsy included the assessment of fibrosis/cirrhosis, investigation of fatty liver disease (mainly non-alcoholic fatty liver disease) and abnormal liver enzymes (see table 2).
A breakdown of the questionnaire results can be seen in table 3. We found that 47/50 (94%) of cases had adequate clinical information provided on the requisition form. The hepatologists felt that 48/50 (96%) of reports addressed the clinical indication for biopsy and provided clinically useful information. The two cases that were deemed not clinically useful (4%) were because the biopsy was inadequate for assessment.
Reports were deemed clinically useful for more than one reason in many of the cases. Most commonly (56% of useful cases), the biopsy was able to confirm a diagnosis suspected clinically; 46% of reports helped decide between differential diagnoses and 42% of reports were able to exclude suspected diagnoses. In 13 cases (27% of useful cases), the biopsy report raised a differential diagnosis which was clinically unanticipated (Q4c in the audit proforma, see table 1), such as unexpected biliary disease and cirrhosis. In 63% of cases, reports provided grading or staging information.
Overall, 74% of reports (37/50) resulted in a change of management. In many of the cases, more than one change of management resulted, but in 15/37 (41%) of cases this was due to a change from the original diagnosis to a final diagnosis (Q8a in the audit profoma, see table 1) based on the biopsy report. The resulting change was usually (84%) in follow-up arrangements, but in 51% of cases treatment plans were changed. All cases had a record of the case discussion in the medical notes.
The indications for medical liver biopsy, and consequently the numbers being performed are changing.1–10 The reasons include the recent availability of non-invasive tests for liver fibrosis,15 ,16 and treatment guidelines for viral hepatitis which are non-biopsy directed.8 In our experience, the consequence is that the number of biopsies taken primarily to assess disease severity is declining, whereas the proportion of liver biopsies requested to investigate the diagnosis where this is not apparent from the non-invasive liver screen, or decide among possible clinical diagnoses is increasing. The interpretation of biopsies in the latter category is more complex and an adequate sample and good careful clinico-pathological correlation are both essential to ensure maximum benefit from the biopsy.
Two medical liver biopsy audit templates were produced by the RCPath following the publication of the Tissue Pathways for Medical Liver Biopsy interpretation.10 These were designed to enable pathologists to evaluate their medical liver biopsy service. The audit piloted here reviews the clinico-pathological contribution of the biopsy to patient management. It aims to develop understanding about when a biopsy does, and doesn't, meet the needs of the clinician and patient. There is a separate audit template relating to the adequacy of the biopsy sample.
The audit design was based on the template for other pathology specialties,14 and the ‘yes/no’ format used to give quantitative data, with explanation and percentage compliance with audit criteria where appropriate. Both histopathologists and hepatologists are able to fill in appropriate parts of the questionnaire and this allows the tool to assess how appropriate the biopsy was and how well the requestor provides information while also assessing the quality of the report provided. The audit can also be filled in contemporaneously by the end user, which is ideal in assessing the real life impact of the biopsy report. The MDT meeting is an ideal time to do this either concurrently with each case review or in bulk at the end of the meeting. On average, around five cases would be suitable per week and these would take around 10 min in total to audit. The audit was welcomed by the hepatology team and the workload impact was minimal. We found the proforma to be fit for the purpose of encouraging dialogue, and appreciation of the information required on the part of both pathologists and clinicians to gain optimal patient benefit from the biopsy. We audited 50 cases but with small numbers of cases at most centres in the UK, the time frame to complete a large number of cases prospectively may be prolonged and to minimise increased workload we suggest 20 cases to be audited annually. Centres with fewer than 20 cases should probably consider the appropriateness of this service, as we have suggested previously.10 ,11 Our initial questionnaire posed a few syntax issues to begin with and we edited the wording to make this easier to understand. For example, we found that the biopsy report may not change management in all situations, such as the exclusion of a differential diagnosis, but that it was still necessary that the biopsy had been taken in order to plan management. The original questionnaire didn't take into account this situation and we adapted the wording of questions 6 and 7 in light of our experience.
There is no pre-existing evidence base on which to set standards. So, we created standards for only two categories, at levels which initially we felt to be sensible and appropriate. Thus, initially we expected that a standard of 100% would be achieved for both providing relevant clinical information and for providing a report which addressed clinical indications. We found that 94% of cases had adequate clinical information provided on the request form. Review of the service highlighted that the hepatologist's request form was different from the request form received in the histopathology department and steps were taken to ensure all relevant clinical information is passed on in future. On reflection, a standard of 100% for this is probably unrealistic and we are now recommending 95% as a reasonable target.
We found that only two reports did not address the clinical indication, but in these instances it was because of an inadequate biopsy. Again, a standard of 100% is not reasonable as a biopsy inadequacy rate will always be present in every centre. In our audit, 4% of biopsies were deemed inadequate for assessment and so an audit standard of 95% is probably reasonable.
Our results show that hepatologists found the histopathology reports to provide clinically useful information in 96% of cases. This supports the appropriateness of requests in our service. We are recommending that a centre should be able to provide clinically useful reports in all cases where the biopsy sample and clinical information are adequate. Interestingly, almost a third of liver biopsies raise a possible differential diagnosis which was not clinically suspected. Furthermore, despite changing diagnostic trends, a large number of cases (63%) still provided useful grading and staging information.6 ,15 ,16
As already mentioned, while not all reports may result in a management change, the exclusion of other diagnoses is important. In this context, in 92% of cases the biopsy was needed before the management plan could be carried out (changed or not). Seventy-four per cent of reports resulted in a change of management. Where changes were made, this was usually in follow-up, but in a significant number of cases (just over half of such cases—19/37), these were treatment changes. These numbers are high and again this highlights the relevance of medical liver biopsy, demonstrating the high-quality local service, although implementing a standard for these values we felt was probably not appropriate.
In conclusion, we have developed an audit tool which can be used primarily to give insight and feedback into the medical liver biopsy histopathology service. We found that this audit provided us with useful information about our local liver biopsy service, which has resulted in improvements and given us assurance that our reports match clinical needs. We encourage other departments to use the template and recommend that all centres providing medical liver histopathology services audit their service periodically to ensure the service remains fit for purpose during this time of changing role of the biopsy.2 ,3 ,8 ,9 We expect that the audit template will be published on the Royal College of Pathologists website14 in due course.
Take home messages
A new medical liver biopsy clinico-pathological audit tool for the UK has been developed and successfully piloted.
The audit tool resulted in valuable insights in the role of medical liver biopsy in patient management and encouraged good practice in clinico-pathological diagnosis.
Clinicians found liver biopsy helpful, contributing to patient management in all of the cases (96%) where biopsy was adequate size.
Almost a third of liver biopsies result in a clinically unanticipated diagnosis.
Around three-quarters of liver biopsies result in a change of management.
Handling editor Cheok Soon Lee
Contributors SH and JW designed the audit template. RC and EF carried out the audit. JeC, JaC and RC completed questionnaires. RC drafted the manuscript. KF oversaw the project. All authors contributed to the interpretation of the results and the editing of the final manuscript.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Further data can be obtained from the corresponding author.