Ann Thorac Surg 2002;74:2161-2164
© 2002 The Society of Thoracic Surgeons
Original article: cardiovascular
The use of routine chest x-ray films after chest tube removal in postoperative cardiac patients
James T. McCormick, DOa,
Michael S. O’Mara, MDa,
Pavlos K. Papasavas, MDa,
Philip F. Caushaj, MDa*
a Department of Surgery, Temple University School of Medicine, Clinical Campus, Western Pennsylvania Hospital, Pittsburgh, Pennsylvania, USA
* Address reprint requests to Dr Caushaj, 4800 Friendship Ave, Pittsburgh, PA 15224, USA
e-mail: pcaushaj{at}wpahs.org
Presented at the Poster Session of the Thirty-eighth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 28–30, 2002.
 |
Abstract
|
|---|
BACKGROUND: The use of routine postoperative chest x-ray films (CXRs) for postoperative cardiac patients has been challenged, suggesting that only clinically indicated CXRs be obtained. The removal of chest tubes has been used as an indication for CXRs. Our hypothesis is that routine postoperative chest tube removal CXRs are not indicated in the asymptomatic postoperative cardiac patient.
METHODS: Charts of 1,021 consecutive postoperative median sternotomy patients were reviewed, focusing on postoperative findings of CXRs, clinical evaluations, and interventions. Those who died prior to tube removal were excluded from the study.
RESULTS: Tubes were removed on postoperative days 1 to 7 (average, 1.45 days). The two groups of patients were comparable in age, gender, procedure, and co-morbidity (p > .01). Seven hundred three patients underwent routine postoperative tube removal CXRs. Abnormal findings were present in 282 patients. Resultant therapeutic intervention was undertaken in 13 patients and 9 were symptomatic. No imaging after routine postoperative CXRs was conducted in 283 patients. These patients remained asymptomatic and required no intervention. Fourteen patients had clinically indicated CXRs after chest tube removal. Two of these patients had additional tubes placed, and 1 patient had follow-up films. In total, there was a 1.5% incidence of therapeutic intervention after chest tube removal. All patients were discharged without further sequelae of their tubes.
CONCLUSIONS: Omission of routine postoperative chest tube removal CXRs in postoperative cardiac patients is safe. The removal of chest tubes in these patients is not an indication for CXRs.
|
Introduction
|
|---|
The use of routine postoperative chest x-ray films (CXRs) in postoperative cardiac and thoracic patients has been challenged. [1–3] Silverstein and Coworkers, [4] and others, have concluded that CXRs are not useful in the absence of a clinical indication. The removal of chest tubes is considered by many as an indication for CXRs. However, this indication for CXRs has been refuted in the trauma literature, which specifically studied closed thoracostomy chest tubes [5, 6]. By extrapolation, our hypothesis is that routine postoperative chest tube removal CXRs are not indicated in the asymptomatic postoperative cardiac patient. The purpose of this study was to assess the yield and clinical impact of the CXRs and determine the safety of omitting them.
|
Patients and methods
|
|---|
Charts of 1,021 consecutive patients who underwent coronary artery bypass grafting or a valve operation by median sternotomy between October 1999 and May 2001 were reviewed. Exclusion criteria were death in the cardiovascular recovery unit or death with chest tubes in place. Particular attention was given to postoperative CXR findings, clinical evaluations, interventions, outcomes, and management of the chest tubes. The endpoint of the study was discharge from the hospital; however, follow-up included any readmission or emergency room visit to our institution.
No routine postoperative chest tube removal CXRs were ordered on all patients during a 4-month time period in the year 2000 when James McCormick and Michael O’Mara, working as house staff, were responsible for postoperative patient care on the cardiothoracic service. This was a conscious deviation from the institutional postoperative routine. Patients from multiple surgeons were included.
Patients were broken down into two groups: (1) patients who had routine postoperative chest tube removal CXRs and (2) patients who did not have CXRs. These two groups were compared statistically using Fisher’s exact test, 
2 test, and t test where appropriate. Probability (p) values of less than 0.05 were considered significant. The yield of the CXRs in terms of pathologic image findings and alteration of clinical management was determined, and interventions were analyzed. For purposes of this study the interpretation of the radiologist was considered final.
CXRs were defined as routine if the indication was routine or chest tube removal, or if they were ordered as part of some protocol or postoperative follow-up plan. If none of these criteria were met and no indication was provided, progress notes were reviewed for clinical indications (described as follows). CXRs requested on the basis of clinical indication were considered indicated and therefore were not routine. Clinical indications for CXRs were abnormal arterial blood gases, hemodynamic instability or critical condition, dyspnea, low oxygen saturation (pulse-ox), bleeding, and abnormal findings on chest examination (adapted from Rao and Colleagues) [2].
Interventions were defined as any procedure (ie, chest tube insertion) performed in an attempt to correct an image abnormality or to treat some intrapleural pathology. Changes in medical management (ie, administration of diuretics) and further diagnostics (ie, repeat CXRs) were not considered interventions.
|
Results
|
|---|
Figure 1
illustrates the breakdown of patients into the two groups. Table 1
demonstrates that there were no statistically significant differences between the groups by demographics or by key co-morbidities. The primary procedure was performed with equal frequency in both groups as shown in Table 2.
View larger version (33K):
[in this window]
[in a new window]
|
Fig 1. Breakdown of patients into groups based on presence or absence of routine postoperative chest tube removal chest x-ray films (CXRs).
|
|
Chest tubes were removed on average on day 1.4 (range, 1 to 7), with 292 patients having chest tubes retained after postoperative day 1. The most common reason for retention was high output (266 patients), followed by air leak (15 patients), pneumothorax (7 patients), critical condition (3 patients), and pleural effusion (1 patient). Three of these patients had interventions before original chest tube removal (two chest tubes and one re-exploration).
Routine postoperative chest tube removal CXRs group (n = 703)
Findings on routine postoperative chest tube removal CXRs are presented in Table 3.
Eight patients had chest tubes placed for symptomatic pleural effusions or pneumothoraces. Four additional patients had chest tube placements and another patient had thoracentesis for small asymptomatic pneumothoraces. Findings on routine postoperative chest tube removal CXRs prompted repeat imaging (in some cases multiple imaging) in 27 asymptomatic patients. None of the remaining patients required intervention. Included in this group was 1 patient who was discharged with a known, clinically improving, moderate-sized pleural effusion discovered on routine postoperative chest tube removal CXRs. This patient was readmitted to the hospital with dyspnea and had a chest tube placed. No other patients in this group required intervention during the follow-up period.
No routine postoperative chest tube removal CXRs group (n = 297)
Fourteen patients (5%) proceeded to have clinically indicated CXRs. Three CXRs yielded no pathology. CXRs demonstrated pleural effusion in 6 patients, pneumothorax in 4, and both effusion and pneumothorax in 1 patient. Two of these patients underwent interventions (one small caliber chest tube and one large caliber chest tube). One patient underwent repeat CXRs. The remaining 283 patients who had no routine postoperative chest tube removal imaging remained asymptomatic and required no intervention. There were no identifiable adverse events resulting from omission of the routine postoperative chest tube removal CXRs.
First outpatient visit follow-up was obtained for 231 (82%) of the 283 patients who had no CXRs after chest tubes removal. Average follow-up was 11 days after discharge. Thirty-two (14%) of these patients had CXRs at that time (Table 4).
None of these patients required intervention.
Analysis of subgroups
Eleven of 14 (79%) of indicated CXRs yielded pathology. Whereas, only 281 of 703 (40%) routine CXRs demonstrated pathology (p = 0.005 by Fisher’s exact test).
There were 292 patients who had chest tubes retained beyond postoperative day 1. Abnormal postoperative chest tube removal CXR findings were present in 112, significantly more than those patients whose chest tubes were removed on day 1 (p = 0.001 by Fisher’s exact test). However, this did not result in significantly more interventions (5 of 292) (p = 0.564 by Fisher’s exact test).
An internal mammary artery was used in 741 patients (80%) who had bypass grafting. In this group there were 10 interventions, which is not more than the overall sample population (p = 0.765 by Fisher’s exact test).
|
Comment
|
|---|
Overall, the incidence of intervention after chest tube removal was 1.4%. The incidence of intervention was 0.6% in the no CXR group and 1.7% in the routine CXR group (p = 0.25 by Fisher’s exact test). Findings on routine postoperative chest tube removal CXRs prompted intervention in 4 asymptomatic patients and repeat CXRs in an additional 27 patients who lacked clinical evidence of disease. These patients may have done well without intervention raising the possibility that the CXRs may have had a deleterious effect on patient care.
CXRs obtained for clinical indications were more likely to yield pathology. However, this difference did not translate into an increased number of interventions.
Dissection of the internal mammary artery did not seem to result in an increased need for intervention. Hurlbut and colleagues [7] reported that dissection of the internal mammary artery resulted in a significantly higher incidence of intervention (4%) compared with patients that had only saphenous vein grafts (n = 100). We report an incidence pleural pathology requiring intervention of 1.3% in this subgroup of patients, which was not significantly higher than the overall sample population.
At our institution, CXRs cost $112.00. Omission of postoperative chest tube removal CXRs for 283 patients saved $31,696. There was the potential to save an additional $78,736 in CXRs alone, not to mention the cost of follow-up CXRs and unnecessary intervention.
This is a retrospective chart analysis and therefore it has inherent bias. With some patients we had to rely on the investigator’s interpretation of the progress notes to determine patient symptoms, clinical findings, and impressions. Therefore, it is possible that some CXRs and procedures may have had clinical indications not elicited from the chart. However, this process occurred in only a few patients because most CXRs were done as part of a standard postoperative follow-up plan, and procedure notes were generally complete and coherent.
The endpoint of the study was discharge from the hospital, but included any re-admission to the hospital or emergency room visit at our institution. Therefore any conclusions must rely on the assumption that patients who developed problems after discharge returned to our institution for treatment. It is possible that patients requiring treatment after discharge went elsewhere and therefore are not be represented by our data. In an effort to negate some of this concern, we were able to accomplish first outpatient visit follow-up on 81% of patients who had no CXRs after chest tube removal. Despite these drawbacks, we believe our data statistically reflects actual outcomes.
At our institution the routine postoperative orders have been changed eliminating routine CXRs.
|
Conclusion
|
|---|
Omission of routine postoperative chest tube removal CXRs is safe, and removal of chest tubes in postoperative cardiac patients is not an indication for CXRs. Rather, liberal use of clinical indications for imaging should be encouraged to avoid missing important and alterable pathology. The need for intervention after chest tube removal in postoperative cardiac patients occurs infrequently. In addition to consuming valuable resources, unnecessary imaging may result in nonindicated interventions. These interventions may be avoided when clinical signs and symptoms guide the imaging.
|
References
|
|---|
- Hornick P.I., Harris P., Cousins C., Taylor K.M., Keogh B.E. Assessment of the value of immediate postoperative chest radiograph after cardiac operation. Ann Thorac Surg 1995;59:1150-1154.[Abstract/Free Full Text]
- Rao P.S.R., Abid Q., Khan K.J., et al. Evaluation of routine CXR in the management of the cardiac surgical patient. Eur J Cardiothor Surg 1997;12:724-729.[Abstract]
- Graham R.J., Meziane M.A., Rice T.W., et al. Postoperative portable CXR: optimum use in thoracic surgery. J Thorac Cardiovasc Surg 1998;115:45-52.[Abstract/Free Full Text]
- Silverstein D.S., Livingston D.H., Elcavage J., Kovar L., Kelly K. The utility of routine daily chest radiography in the surgical intensive care unit. J Trauma 1993;35:643-646.[Medline]
- Pacanowski J.P., Waack M.L., Daley B.J., et al. Is routine roent-genography needed after closed CT thoracostomy removal?. J Trauma 2000;48:684-688.[Medline]
- Palesty J.A., McKelvey A.A., Dudrick S.J. The efficacy of x-rays after CT removal. Am J Surg 2000;179:13-16.[Medline]
- Hurlbut D., Myers M.L., Lefcoe M., Goldbach M. Pleuropulmonary morbidity: internal thoracic artery versus saphenous vein graft. Ann Thorac Surg 1990;50:959-964.[Abstract]
This article has been cited by other articles:
|
|
|
|
 
T. Khan, G. Chawla, R. Daniel, M. Swamy, and W. R. Dimitri
Is routine chest X-ray following mediastinal drain removal after cardiac surgery useful?
Eur. J. Cardiothorac. Surg.,
September 1, 2008;
34(3):
542 - 544.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
G. Hejblum, V. Ioos, J.-F. Vibert, P.-Y. Boelle, L. Chalumeau-Lemoine, C. Chouaid, A.-J. Valleron, and B. Guidet
A Web-Based Delphi Study on the Indications of Chest Radiographs for Patients in ICUs
Chest,
May 1, 2008;
133(5):
1107 - 1112.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
