The National Joint Registry may fail to collect accurate, validated anaesthetic data

The National Joint Registry (NJR) was established by the Department of Health in England and the Assembly Government in Wales to collect information on all hip and knee replacement operations and monitor the performance of joint prostheses [1]. A levy is paid by prosthesis manufacturers for data to be collected locally and recorded nationally. There are four procedure-specific forms, all of which request details of anaesthetic type as part of the minimal dataset. Anaesthetic technique classifications are general anaesthetic, nerve block, epidural or spinal, with the option to select more than one. The American Society of Anesthesiologists (ASA) grade is recorded. A registered staff member uploads the completed forms to the NJR database, usually on a weekly basis. The NJR recommends that the anaesthetist records the anaesthetic and ASA grade on the form. The ASA grading is published in the annual reports as a measure of pre-operative health status. Anaesthetic technique data for 2005 was published in the 3rd NJR Annual Clinical Report, but it was not published in the NJR 4th Annual Report (comprising April 2006 to March 2007, includes the clinical report). The NJR compares the submitted volumes of data with hospital episode statistics in England, and the patient episode database Wales. This is the only form of systematic validation (I. Mulcahy, NJR, personal communication). There is no validation of anaesthesia data. As part of the validation process of the 3rd National Audit Project of the Royal College of Anaesthetists (NAP3) [2] one author (TC) requested the dataset for anaesthetic techniques used during the period of NAP3 (1st September 2006–31st August 2007). The information received showed that, during the relevant period, data from 141 280 hip or knee joint arthoplasties (primary and revision surgery) were reported to the NJR, representing 81% of such operations in England and Wales. Details of anaesthetic technique were recorded in 92.5% of operations. We considered it might be useful to publish the data as a bi-national census of anaesthetic techniques for such surgery, and permission for this was given by the NJR. Our initial step was to validate our local data. We prospectively collected details of anaesthetic technique and ASA grade from the case notes of 100 patients in our hospital immediately after knee or hip replacement. Data were collected over a 2-month period (October–November 2008). Six weeks later (to allow for central collection of NJR forms) we compared this data with that recorded on the NJR form for the same patients. We made no attempt to validate the orthopaedic data submitted to the NJR. Of the 100 NJR forms, 16 had no anaesthetic technique recorded and six did not record ASA grade. Overall, in only 48 forms did the NJR data and anaesthetic chart data match entirely. The ASA grade was different to the anaesthetic chart in 16 NJR forms: a higher grade in nine cases and lower grade in seven. Twenty-seven NJR forms recorded a different anaesthetic technique to that recorded on the anaesthetic chart for the same patient. The differences are summarised in the Table 2. Epidural was reported on three NJR forms where spinal was used. Two spinals were not reported when combined spinal-epidural was given. In our hospital NJR forms are filled in by orthopaedic surgeons and their trainees with only occasional informal involvement of the anaesthetist. A member of the hospital audit department collects the NJR forms and attempts to chase up missing forms and data. Anaesthetists are not involved in this stage. As part of this project we explored the data held in the hospital theatres database and found similar discrepancies between that and anaesthetic charts. It is possible that missing anaesthetic data is completed on the NJR form using the inaccurate theatres database. On the basis of our local data, we have concerns that the anaesthetic data (both health status and anaesthetic technique) in the NJR may be inaccurate. It is possible that this is a problem unique to our Trust, but this seems unlikely. If similar large discrepancies exist even in a relatively small proportion of English and Welsh hospitals the NJR anaesthetic information will be invalid. Debate continues as to the optimal anaesthetic technique for elective major joint arthroplasties. Recent studies suggest marginal benefit from neuraxial or peripheral nerve blockade techniques above general anaesthesia [3, 4]. While similar joint registers exist in the USA, Canada, Australia, New Zealand, Norway, Sweden and Finland, not all of these record anaesthetic data. To date, we are not aware of any large observational studies of current anaesthetic practice for large joint replacement. The NJR presents a considerable opportunity to collect and report such data; however that can only be achieved if the entered data is accurate and validated. Involving anaesthetists directly in NJR form completion will increase the accuracy of locally collected data. The NJR might emphasise how such data is collected and perform a validation or ‘cleaning cycle’ in a similar manner to the Intensive Care National Audit and Research Centre. Alternatively, the NJR might wish to stop collecting such data. At present there seems to be significant risk of ‘rubbish in, rubbish out’.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.