Background Robust data collection regarding diagnoses of those referred to supportive and palliative care services is essential. It enables constructive dialogue between provider and stakeholders, supports service planning and delivery, and identifies trends in clinical conditions. With increasing age and multi-morbidity, data collection is increasingly challenging. This audit aimed to look at the accuracy of data collection within a single hospice.

Methods This was a retrospective audit of eighty patient notes. An audit proforma was developed to collect data relating to demographics, diagnosis, and comorbidities. The ICD-10 diagnostic code for each patient was extracted from the electronic data information system. The agreed standard was 100% of ICD-10 diagnostic codes should match primary diagnosis.

Results 61% of notes showed congruence between primary diagnosis and ICD-10 diagnostic code recorded. Of those notes showing discrepancy, 4 had no ICD-10 code recorded; 2 did not easily fit within ICD-10 codes available; 3 had multiple malignancy and immediate life limiting malignancy was incorrectly identified; 5 had the incorrect non-malignant diagnosis selected; and 1 was incorrectly recorded as having a non-malignant diagnosis. 63% of patients had a primary diagnosis of malignancy. The overall mean number of comorbidities was 3 (range of 0–7) with a mean of 3.5 for those with an incorrect code. Most common co-morbidities were hypertension (30%), type two diabetes (22%), ischaemic heart disease/heart failure (17%), atrial fibrillation (17%), and other malignancy (15%).

Conclusions Recommendations to improve clinical coding included: identifying clinical and administrative points during patient journey to review ICD 10 diagnostic code; use of multiple diagnostic codes; and staff education around data collection. The difficulties identified reflect the challenge of accurate clinical coding within the context of an increasingly complex caseload and multiple co-morbidities. It highlights the importance of developing more nuanced approaches to clinical coding within the palliative care setting.