Decision-making in the treatment of diaphyseal clavicle fractures: is there agreement among surgeons? Results of a survey on surgeons' treatment preferences.

BACKGROUND AND HYPOTHESIS

Nonoperative treatment is standard for most diaphyseal clavicle fractures, but recent studies have demonstrated improved outcomes with operative treatment of displaced fractures. The objectives of this diagnostic study were to assess agreement of orthopaedic surgeons regarding their treatment preferences for diaphyseal clavicle fractures and to compare them with recent recommendations. Interobserver and intraobserver agreement in treatment decisions were hypothesized to be only slight.

METHODS

Anonymized case vignettes of 50 acute diaphyseal clavicle fractures including medical history, physical examination findings, and radiographs were independently reviewed by 32 orthopaedic surgeons from the United States. Four treatment options were offered and decisions were compared with current treatment recommendations. Interobserver agreement was calculated using Fleiss' kappa coefficient. Average intraobserver agreement for surgeons who completed a retest review (minimum interval of 8 weeks) was calculated.

RESULTS

Thirty-two surgeons completed the first round of reviewing and 27 completed the retest (mean interval, 22 weeks). Interobserver agreement was overall fair (kappa = 0.36) and moderate (kappa = 0.56) when operative options were compared with nonoperative options. Median intraobserver agreement was 74% for the 4 treatment options offered and 84% in deciding on operative vs. nonoperative means. Concordance with recent recommendations for operative vs. nonoperative treatment was seen in 91% of decisions (median).

DISCUSSION AND CONCLUSIONS

Recent recommendations appear to have been adopted by a selected subgroup of U.S. orthopaedic surgeons, showing a surprisingly high median concordance of 91% in this study. However, only fair to moderate interobserver and intraobserver agreement was present, leaving potential for improvement.