A mixed methods study to explore the diagnostic accuracy and acceptability of the tuning fork test in the detection of ankle fractures
Student thesis: Doctoral Thesis
Aim and methodology: Ankle injuries account for 8% of all minor injuries attending emergency departments in the United Kingdom and the Ottawa ankle rules were introduced to assess the need for x-ray in the early 1990s (Stiell et al 1992). Although the rules are said to have reduced the number of ankle x-rays requested the frequency of fractures in the population still receiving x-rays is only 15% nationally. This study aims to assess whether the tuning fork can increase the diagnostic accuracy of the Ottawa ankle rules when used on twisting ankle injuries by multiple operators in multiple emergency care settings. A mixed methods study conducted in two phases was undertaken. Phase one consisted of a diagnostic test study using the Ottawa ankle rules in conjunction with the tuning fork test on patients already screened as being Ottawa positive to the ‘malleolar’ zone and requiring an x-ray of their ankle. Patients aged 12 years or over who had sustained an ankle injury by a twisting mechanism were eligible to take part. Patient age, gender, ethnicity, and previous history of injury or presence of distracting injuries, degree of swelling, and role of operator were all considered potential variables for an accurate tuning fork test, and these were analysed individually and in a multiple logistical regression model to assess for predictor variables of a correct tuning fork test. Phase two of the study included a series of focus group discussions to explore participant and clinician experiences of the tuning fork test. Data was analysed using thematic analysis. Results Data was collected for 2-years and 1313 patients were included in the final analysis. 56% of the study participants were male. Mean age was 34 years (range 12-91). 98% were of white ethnic origin. 210 (16%) were diagnosed with fractures, of which 38 were deemed to be not clinically significant. The tuning fork had a diagnostic accuracy of 56% (95% CI 53-58), NPV 96% (95% CI 94-97), sensitivity 84% (95% CI 78-89) and specificity 51% (95% CI 48-54). X-rays could have been reduced by 47% but this was at the expense of missing 29 ‘clinically significant’ fractures. However, seven of these were managed as soft tissue injuries and in nine the initial assessment of tenderness did not match the site of the fracture. A total of 113 clinicians (nurses & doctors) were involved in performing the tuning fork test independently. Patient age (adjusted OR 1.021, p. <0.001) and role of the operator (adjusted OR 1.595, p. 0.003 for nurse) were the only predictors of an accurate test. Ten patients and ten clinicians attended the focus group discussions in phase two of the study. Patients and clinicians appeared to accept the tuning fork as a method for assessment provided adequate explanation was given. Patients claimed the tuning fork test was not painful but had a similar sensation to that of a ‘Tens’ machine. There were differences in opinion between the two groups as to whether the tuning fork was accurate or not and clinicians held the perception that patients expect an x-ray when they present with an ankle injury, whereas patients disagreed with this. Patients were fully aware of the dangers of x-rays and stated that a reduction in x-rays was one of the main potential benefits of the study. Conclusions This is the largest study to investigate the accuracy of the tuning fork to detect fractures, not only in the size of the study population but the number of clinicians involved. It is also the first to report inconclusive Ottawa ankle rule and tuning fork test results. It is unlikely that the lower sensitivity will be accepted by patients and clinicians. Further research to assess inter-operator reliability is recommended before implementing the tuning fork test into clinical practice.
|Award date||Jan 2012|
2.68 MB, PDF document