Evaluation of the accuracy of diagnostic coding and clinical documentation for traumatic heterotopic ossification diagnoses in Western Australian hospitals
Publication Details
Foster, N.,
Raby, E.,
Wood, F.,
Fear, M.,
Pavlos, N.,
&
Edgar, D.
(2024).
Evaluation of the accuracy of diagnostic coding and clinical documentation for traumatic heterotopic ossification diagnoses in Western Australian hospitals.
Injury, 55 (3).
https://doi.org/traumatic, heterotopic ossification, spinal cord injury, brain injury, burn injury, orthopaedic injury
Abstract
Background: Traumatic heterotopic ossification (tHO) refers to the pathological formation of ectopic bone in soft tissues that can occur following burn, neurological ororthopaedic trauma. As completeness and accuracy of medical diagnostic coding can vary based on coding practices and depend on the institutional culture of clinical documentation, it is important to assess diagnostic coding in that local context. To the authors’ knowledge, there is no prior study evaluating the accuracy of medical diagnostic coding or specificity of clinical documentation for tHO diagnoses across Western Australia (WA) trauma centres or across the full range of inciting injury and surgical events.
Objective: To evaluate and compare the clinical documentation and the diagnostic accuracy of ICD-10-AM coding for tHO in trauma populations across 4 WA hospitals.
Methods: A retrospective data search of the WA trauma database was conducted to identify patients with tHO admitted to WA hospitals following burn, neurological or orthopaedic trauma. Patient demographic and tHO diagnostic characteristics were assessed for all inpatient and outpatient tHO diagnoses. The frequency and distribution of M61 (HO-specific) and broader, musculoskeletal (non-specific) ICD-10-AM codes were evaluated for tHO cases in each trauma population.
Results: HO-specific M61 ICD-10-AM codes failed to identify more than a third of true tHO cases, with a high prevalence of non-specific HO codes (19.4 %) and cases identified via manual chart review (25.4 %). The sensitivity of M61 codes for correctly diagnosing tHO after burn injury was 50 %. ROC analysis showed that M61 ICD-10-AM codes as a predictor of a true positive tHO diagnosis were a less than favourable method (AUC=0.731, 95 % CI=0.561–0.902, p = 0.012). Marked variability in clinical documentation for tHO was identified across the hospital network.
Conclusion: Coding inaccuracies may, in part, be influenced by insufficiencies in clinical documentation for tHO diagnoses, which may have implications for future research and patient care. Clinicians should consistently
Link to Publisher Version (URL)
