The efficiency of a clinical pathway to guide combined applications of interventional pulmonology in undiagnosed pleural effusions

Abstract

Abstract The diagnostic procedure of pleural effusion (PEs) is challenging due to low detection rates and numerous aetiologies. Hence, any attempt to enhance diagnosis is worthwhile. We present a clinical pathway to guide combined application of interventional pulmonology (IP) for tracing causes of undiagnosed PEs. Subjects with undiagnosed PEs were identified in the Hospital Information System of Dalian Municipal Central Hospital from January 1, 2012, to December 31, 2018. Eligible subjects were divided into a group of combined tests and a group of medical thoracoscopy (MT). Optimal and subsequent diagnostic tests were performed depending on the guidance of the clinical pathway by matching profitable chest lesions with the respective adaptation. As the guidance of clinical pathway, common bronchoscopy would be preferentially selected if pulmonary lesions involved or within the central bronchus, EBUS-TBNA was favoured when pulmonary lesions were adjacent to the central bronchus or with the enlarged mediastinal/hilar lymph nodes, guided bronchoscopy would be preferred if pulmonary nodules/masses were larger than 20 mm with discernible bronchus signs, CT-assisted transthoracic core biopsy was preferred if pulmonary nodules were less than 20 mm, image guided cutting needle biopsy was the recommendation if the pleural thickness was larger than 10 mm and pulmonary lesions were miliary. MT was preferred only when undiagnosed PEs was the initial symptom and pulmonary lesions were miliary or absent. A total of 83.57% cases of undiagnosed PEs were eligible for the clinical pathway, and 659 and 216 subjects were included in the combined tests and MT groups, respectively, depending on the optimal recommendation of the clinical pathway. The total diagnostic yields in the combined tests and MT groups were 95.99% and 91.20%, respectively, and the difference in total diagnostic yield was statistically significant (χ2 = 7.510, p = 0.006). Overall, clinical pathway guidance of the combined application of IP is useful for tracing the causes of undiagnosed PEs. The diagnostic yield of undiagnosed PEs is significantly increased compared with that of MT alone.