طب اورژانس

An effective clinical decision tool for evaluating patients with elbow injury might reduce the need for radiography for fracture. A prospective study of the use of an elbow extension test to assess for elbow fracture showed that this test has a high sensitivity and negative predictive value for elbow fracture.^[1]

The purpose of this study was to assess whether the elbow extension test could be used to rule out fracture in patients with acute elbow injury. This was a prospective validation study in adults and an observational study in children carried out in five emergency departments in England, which entered patients from July 2004 to April 2006. Adults 15 years old and older and children 3 to 15 years old who presented within 72 hours of acute elbow injury were included. Patients presenting more than 72 hours after injury, with previous limited extension, or with no history of trauma were excluded.

The investigators prospectively identified a sensitivity of 99% for the test to be clinically acceptable. Patients were identified during triage and were given standard analgesia. Patients underwent an elbow extension test that consisted of having patients seated with arms supinated flex their shoulders to 90 degrees and fully extend and lock both elbows. Normal and injured sides were compared. Patients with equal extension were considered to have “full extension.” Adults with a negative test result (full extension) were discharged with analgesia and a sling as needed but without radiography. Regardless of the test result, children with acute elbow injury underwent radiography on the decision of the treating physician or nurse. Assessment was done by telephone in 7 to 10 days for patients discharged without radiography. The study included criteria for recall of patients. Patients with inability to extend the elbow fully, continued or worsening pain, or difficulty using the injured arm were recalled for radiography. Patients not requiring recall were considered not to have significant injury.

The study enrolled and tested 960 adults and 780 children. The mean age for adults was 38 years, and mean age for children was 10 years. The most common type of injury in adults was injury to the radial head and in children was supracondylar injury. Fractures were present in 31% (538 of 1740) of adults and children. With regard to the elbow extension test, 33% (311 of 958) of adults were able to extend fully. Among these patients, five patients had fractures, two of which—both olecranon fractures—required surgery. Among adults who could not fully extend, 48% (311 of 647) had confirmed fractures. Among children, 37% (289 of 778) could fully extend. In these patients, there were 12 fractures, none of which required surgery. Among children that could not fully extend, 43% (210 of 491) had fractures.

Overall, sensitivity for detecting fractures was 96.8% (95% CI 95.0-98.2). Specificity was 48.5% (95% CI 45.6-51.4). For adult patients who could fully extend their elbows, the negative predictive value for fracture was 98.4% (95% CI 96.3-99.5). Among children, the negative predictive value was 95.8% (95% CI 92.6-97.8). This translated to a 1.6% chance of fracture among adults with a negative test (95% CI 0.5-3.7) and a 4.2% chance of fracture among children with a negative test (95% CI 2.2-7.4).

The authors conclude that patients who fail the test should receive radiography for evaluation for fracture because these patients have a nearly 50% chance of fracture. They further conclude that patients who are able to extend fully do not require radiography on the following conditions: that olecranon fracture is unlikely, that interpretation of the test is done cautiously in children, and that patients receive sufficient follow-up. The authors note the high sensitivity and negative predictive value of the test but acknowledge that although the test did not meet the target sensitivity of 99%, sensitivity and specificity were similar to established tests for other joints. The authors urge caution in the setting of potential olecranon fractures in adults and supracondylar fractures in children, in which groups the test also carries a higher false-negative rate. Limitations of the study include the fact that follow-up may have missed undetected fractures and that the recall criteria were unvalidated.

An accompanying editorial discusses the level of acceptable risk in clinical examination.^[2] The author notes that clinicians must make conscious decisions about the level of risk of missed diagnoses. He also notes that risk decisions must be explained to patients.

References