The correlation between the suspected clinical diagnoses and the radiographic diagnoses for patients with shoulder pain

Abstract

Background The association between the radiographic and clinical findings of shoulder pain is unclear and it is not fully understood how plain film radiographs of the shoulder influence the suspected clinical diagnosis and conservative management of shoulder pain at the Chiropractic Day Clinic (CDC) at the Durban University of Technology (DUT). Previous research at the CDC reported that plain film radiographs did not significantly influence the suspected clinical diagnosis and conservation management of the lumbar spine and knee conditions. This study was conducted in order to determine if a similar trend was observed for plain film radiographs of the shoulder and the suspected clinical diagnosis and conservative management of shoulder pain.

Objectives Objective 1: To determine the association between the suspected clinical diagnosis and radiographic diagnosis of shoulder pain. Objective 2: To describe the type of incidental radiographic findings in the selected plain film radiographs of the shoulder. Objective 3: To determine the proportion of change in the suspected clinical diagnosis and the conservative management of shoulder after obtaining the radiographic report.

Method The archives of the CDC at the Durban University of Technology were searched for plain film radiographs of the shoulder and corresponding patients’ records from 4 April 1992 to 19 September 2011. These were collected, examined and evaluated, and the relevant data was extracted. Statistical analysis included the use of percentages, mean, standard deviation, range and frequency counts for the descriptive objectives. The suspected clinical diagnoses were categorized into groups (trauma, mechanical conditions, muscular or tendon dysfunction, capsular syndromes, neurological conditions, arthritides and other). These were then constructed using two-by-two tables for the absence or presence of radiographic diagnoses versus the suspected clinical diagnosis. The McNemars chi square test was used to determine any association between the radiographic and suspected clinical diagnosis.

Results The mean age of the patients whose clinical and plain film radiographic records were examined was 43.5 years, with 53.7% male and 46.3% female patients. It was not possible to find an association between the suspected clinical and radiographic diagnoses as the categories were too different for statistical testing to be done. Thirty one (57.4%) plain film radiographs were requested at the first consultation. In 53.7% (29/54 radiographs) of cases, no specific suspected clinical diagnosis was given and “suspected pathology” was the reason given for referral. Of the 54 patients, 21 had a change in the suspected clinical diagnosis; however, in many of these cases (13/21) it was not a direct result of the radiographic findings. A wide range of treatments were used both before and after plain film radiographs, including soft tissue therapy, manipulation and electrotherapy. There was no significant change in the frequency of use of any of the modalities post-radiographs.

Conclusion The reasons provided for ordering plain film radiographs were sometimes vague or even invalid. Although there was a change in 21 of the 54 suspected clinical diagnoses it was not conclusively as a result of the radiographic findings. The management of shoulder complaints did not change appreciatively following plain film radiographs. These findings suggest that the current use of plain film radiographs in the clinical and management context at the CDC needs to be reviewed.