Inter-Rater Reliability & Value of the Hawkins Sign

Category: Trauma; Ankle Introduction/Purpose: Hawkins III talus neck fractures are associated with elevated rates of complications, notably avascular necrosis, which range between 40% to 70%. The Hawkins sign, characterized by subchondral lucency of the talar dome observed on post-injury ankle x-rays at 6-8 weeks post-injury, has traditionally served as a predictive factor for avascular necrosis development within a year. Nevertheless, we propose that the presence of the Hawkins sign exhibits considerable inter-observer variability and may not predict clinically relevant avascular necrosis development as reliably as previously assumed. Methods: A retrospective study was conducted on patients who experienced Hawkins III fractures between 2018 and 2022 at a Level 1 trauma center. Electronic medical records were thoroughly reviewed to ascertain the occurrence of avascular necrosis. An assessment of six-week postoperative ankle x-rays was conducted by a fellowship-trained foot and ankle orthopaedic surgeon to ascertain the presence or absence of the Hawkins sign. Subsequently, five orthopaedic surgeons independently assessed the x-rays to determine the presence of the Hawkins sign. Statistical analysis was then employed to assess the inter-rater reliability of the Hawkins sign and its association with the future development of avascular necrosis over a one-year follow-up period. Results: Seventeen patients with Hawkins III talus neck fractures were identified. Fleiss' kappa was run to determine if there was agreement between five orthopaedic trauma surgeons’ judgement on whether 17 individuals with Hawkins III talar neck fractures had a Hawkins sign present or not on 6-week postoperative x-rays. There was poor agreement between the surgeons’ judgements, κ = .040 (95% CI, -.035 to .115), but was not statistically significant, p = .294. There was complete agreement between surgeons on only 5/17 (29.4%) cases. Of the 6 patients that went on to develop avascular necrosis at 1 year, only 1 had a Hawkins sign (determined by fellowship-trained foot and ankle surgeon) present on their 6-week postoperative x-ray (Table 1). Conclusion: The reliability of the Hawkins sign among orthopaedic surgeons is notably low due to significant inter-observer variability. This variability introduces subjectivity to the interpretation of the Hawkins sign, rendering it an unreliable objective measure. Furthermore, its inability to correlate with the presence of avascular necrosis at the one-year mark raises doubts about its validity as a predictive marker for talar neck fracture outcomes.