Dataset Information

[Abdominal ultrasound: a diagnostic tool within the reach of general practitioners].

ABSTRACT:

Aim

To assess the impact of training two general practitioners (GPs), on performing low complexity ultrasound examinations of the abdomen with diagnostic competence.

Study design

A non-randomised, prospective, double blind evaluation study of the acquired competence. Ultrasound examinations were performed successively and independently by the GP and the radiologist, and registered on two sheets of an identical case report form.

Setting

Departments of Vic General Hospital and Vic Health Care Centre of the Catalan Health Institute.

Subjects

A total of 120 patients who needed a low complexity abdominal ultrasound from a total amount of 868 ultrasound examinations scheduled.

Main measurements

The kappa index of the primary diagnosis (ultrasound diagnostic conclusion), as well as of the ultrasound findings for each abdominal organ was calculated. A good level of training was considered if a minimum kappa index of 0.80 was attained between general practitioner and radiologist.

Results

After six month running, an "ad interim" analysis showed a kappa index for the primary ultrasound diagnosis of 0.85. At the end of the study, an overall kappa index of 0.89 (95% CI: 0.82-0.98) was achieved. The agreement by organs was also high (pancreas and spleen excluded), as well as sensitivity (>75%) and specificity (>90%).

Conclusions

The results of the study demonstrate the feasibility of the performance, by trained GPs, of ultrasound examinations of low complexity of the abdomen with diagnostic competence.

SUBMITTER: Esquerra M

PROVIDER: S-EPMC7025941 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundDiagnostic errors are a major cause of preventable patient harm. Studies suggest that presenting inaccurate diagnostic suggestions can cause errors in physicians' diagnostic reasoning processes. It is common practice for general practitioners (GPs) to suggest a diagnosis when referring a patient to secondary care. However, it remains unclear via which underlying processes this practice can impact diagnostic performance. This study therefore examined the effect of a diagnostic suggestion in a GP's referral letter to the emergency department on the diagnostic performance of medical interns.MethodsMedical interns diagnosed six clinical cases formatted as GP referral letters in a randomized within-subjects experiment. They diagnosed two referral letters stating a main complaint without a diagnostic suggestion (control), two stating a correct suggestion, and two stating an incorrect suggestion. The referral question and case order were randomized. We analysed the effect of the referral question on interns' diagnostic accuracy, number of differential diagnoses, confidence, and time taken to diagnose.ResultsForty-four medical interns participated. Interns considered more diagnoses in their differential without a suggested diagnosis (M = 1.85, SD = 1.09) than with a suggested diagnosis, independent of whether this suggestion was correct (M = 1.52, SD = 0.96, d = 0.32) or incorrect ((M = 1.42, SD = 0.97, d = 0.41), χ2(2) =7.6, p = 0.022). The diagnostic suggestion did not influence diagnostic accuracy (χ2(2) = 1.446, p = 0.486), confidence, (χ2(2) = 0.058, p = 0.971) or time to diagnose (χ2(2) = 3.128, p = 0.209).ConclusionsA diagnostic suggestion in a GPs referral letter did not influence subsequent diagnostic accuracy, confidence, or time to diagnose for medical interns. However, a correct or incorrect suggestion reduced the number of diagnoses considered. It is important for healthcare providers and teachers to be aware of this phenomenon, as fostering a broad differential could support learning. Future research is necessary to examine whether these findings generalize to other healthcare workers, such as more experienced specialists or triage nurses, whose decisions might affect the diagnostic process later on.Trial registrationThe study protocol was preregistered and is available online at Open Science Framework ( https://osf.io/7de5g ).