Chest
Volume 117, Issue 4, April 2000, Pages 1038-1042
Journal home page for Chest

Clinical Investigations
PLEURA
Limited Utility of Chest Radiograph After Thoracentesis

https://doi.org/10.1378/chest.117.4.1038Get rights and content

Study objective

To assess the utility of chest radiograph (CXR) immediately after routine thoracentesis.

Setting

Multispecialty clinic/teaching hospital.

Participants

All outpatients and inpatients undergoing thoracentesis in the procedure area from October 1995 to January 1998.

Measurements

Immediately after thoracentesis, the physician completed a questionnaire assessing the likelihood of a complication. CXRs were obtained at physician discretion. Patient demographics, indications for thoracentesis, use of ultrasound guidance, level of training, radiographic interpretation, and eventual patient outcome were recorded.

Results

Two hundred eighteen patients were enrolled for a total of 278 thoracenteses. Two hundred fifty-one procedures performed on 199 patients could be prospectively evaluated. A complication was suspected in 30 procedures; immediate CXR confirmed such in 9 (30%). There were 221 procedures with no clinical suspicion or indication of a complication. Ninety CXRs were obtained immediately after the procedure; the remaining 131 procedures had no CXR. The complication rates were 3.3% and 2.3%, respectively, for these groups. Four postthoracentesis radiographs demonstrated additional findings regardless of the indication for the radiograph.

Conclusions

In the absence of a clinical indication of a complication, chest radiography is not indicated immediately after routine thoracentesis. Aspiration of air strongly correlates with the occurrence of pneumothorax, whereas pain, hypotension, and dry tap do not. Use of a vacuum bottle to withdraw fluid obscures the appreciation of this finding and was identified as a risk factor for subsequent pneumothorax. Additional radiographic findings are rarely detected and may not contribute to clinical management.

Section snippets

Materials and Methods

Subsequent to approval by the institutional review board, all patients having thoracentesis performed in the treatment area from October 1995 until January 1998 were prospectively enrolled. As this study constituted comparison of two accepted regimens (process research), consent was not required, and the study was transparent to the physicians and patients. Exclusion criteria included thoracentesis outside the treatment room, concomitant tube thoracostomy, or positive-pressure ventilation.

The

Results

A total of 278 thoracenteses were performed on 218 patients. Complete data allowed for evaluation of 251 procedures on 199 patients. Those data were evaluated to assess the utility of the CXR. The physician declined to complete the questionnaire for 27 procedures performed on 25 patients; because of multiple procedures, six patients appear in both groups. The data on the 27 procedures are only included in the determination of overall rates of complications and associated risk factors. No

Discussion

This study represents the first prospective evaluation of the role of immediate CXR after thoracentesis for both inpatients and outpatients. The overall rates of complication were similar to those of previously published series, as was the requirement for subsequent tube thoracostomy.4 The rate of pneumothorax was similar for experienced staff physician (9 of 142 procedures, 6.3%) and the fellows (5 of 99 procedures, 5.0%) but was slightly higher for the residents/medical students (4 of 37

Conclusion

In the absence of suspicion or clinical indication of a complication, chest radiography immediately after thoracentesis is not warranted in the vast majority of cases either for the identification of pneumothorax or detection of new diagnostic information. This is independent of the intention (diagnostic or therapeutic) of the thoracentesis. The use of a vacuum bottle to withdraw fluid negates this finding and increases the risk of complication. The use of such vacuum devices to withdraw

References (14)

There are more references available in the full text version of this article.

Cited by (0)

View full text