Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Should chest drains be put on suction or not following pulmonary lobectomy ?

Three Part Question

In [patients undergoing pulmonary procedures] is the use of [suction to the chest drains] of benefit in reducing the incidence of [prolonged air leak]?

Clinical Scenario

You are on your ward round and you see a fit 51 year old gentleman 3 days post right upper lobectomy. He has an air leak. He asks why he is not allowed to go to the toilet or go for a walk with the physiotherapist like everyone else on the ward. You tell him that the only way to resolve the air leak is to use suction. He mumbles that it is probably making it worse. You resolve to search the literature to see if he is right.

Search Strategy

Medline 1996-Jan 2006 using the OVID interface
[lobectom$.mp/OR pulmonary OR lung OR pulmonary OR OR exp Thoracic surgery, Video-assisted/OR Thoracoscopy/OR] AND [exp suction/OR OR exp Chest tubes/OR exp Drainage/ OR chest tube$.mp OR chest drain$.mp OR water] AND [air leak$.mp OR exp 'length of stay'/OR hospital OR OR exp Pneumothorax/]

Search Outcome

A total of 391 abstracts were found of which 6 were directly relevant.These are presented in the table. In addition 2 papers were of interest although not tabulated.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
Alphonso et al.
239 patients undergoing lobectomy or wedge resection were randomized into two groups from Feb 2002 to Feb 2004 Grp1-No suction from theatre onwards (N=123) Grp2- Suction at 2kPa (N=116) Randomization by Minimization technique after drain insertion.PRCT (level 1b)Persistent air leak (>6days)No suction 13/123(10.1%). Suction 9/116(7.8%) P=0.6215 patients lost to follow up
Persistent air leak at 2 days20% in each group
Recurrent PneumothoraxNo suction 3pts. Suction 2pts. P=NS
Brunelli et al.
145 patients undergoing lobectomy and with an air leak on the first post-operative day, were randomized into two groups: Grp1-drains taken off suction and placed on Water seal on first Post op day (N=72) Grp2- Suction continued at -20cmH2O (N=73) Drains removed after 24hr clamping trialPRCT (level 1b)Prolonged air leak (>7days)No suction 20/72 (27.8%). Suction 22/73 (30.1%) P=0.8Period of suction overnight 80% of the study group had a pleural tent procedure. Patients without an air leak on day 1 had no suction. No leak meters used.
Air leak duration (days)No Suction 6.5 days. Suction 6.3 days. P=0.9
Postoperative complicationsNo Suction 23/72 (31.9%). Suction 13/73 (17.8%) P=0.056
Ayed KA.
100 patients undergoing thoracoscopy for primary spontaneous pneumothorax were randomized into two groups on the ward after a CXR Grp1-Water seal (N=50) Grp2-Suction (N=50)PRCT (level 1b)Prolonged air leak (>5days)No suction 1/50 (2%). Suction 7/50 (14%). P=0.03Period of suction of 2 hours. Randomization method not given
Chest tube duration (days)No suction 2.7 +/- 1.1 days. Suction 3.8 +/- 2.1 days. P=0.004
Hospital stayNo suction 3.7 +/- 1.1 days. Suction 4.8 +/- 2.1 days P=0.004
Marshall et al.
68 patients that underwent wedge resection, segmentectomy and lobectomy were randomized to two groups after transfer to the recovery room and a CXR showing <25% pneumothorax. : Grp1-Water seal after a short period of suction (N=34) Grp2- Suction at 20cmH2O (N=34)PRCT (level 1b)Air leak (days)No suction 1.50+/- 0.32 days. Suction 3.27 +/- 0.80 days. P=0.05Period of suction of 5 -30mins.
Chest tube removal (days)No suction 3.33 +/- 0.35 days. Suction 5.47 +/- 0.98 days P=0.06
Hospital stayNo suction 4.67 +/- 0.37 days. Suction 11.13+/-4.58 days. P=0.18
Cerfolio et al.
33 patients that underwent pulmonary resections were randomized into two groups on the second post operative day. Grp1- Water seal (N=18) Grp2- Suction (N=15) Leak meter used to grade leak from 1 (least) to 7 ( largest)PRCT (level 1b)Prolonged air leak (>3days)No suction 6/18 (33%). Suction 14/15 (93%) P=0.001

NB 13 of 14 patients with persistent air leak on suction resolved once changed to water seal.
Period of suction till postoperative day 2. 140 patients randomized but only 33 had an air leak on POD 2 when the study commenced.
Air leak Classification. 7 point scale tested for reproducibility17% Disagreement by 1 point. No disagreement by more than 1 point.
Pneumothorax7/32 on water seal developed a pneumothorax, all had a 4/7 or more air leak on the scale.
Antanavicius et al.
Retrospective review of 109 consecutive patients who underwent lobectomy or segmentectomy. For the analysis, patients were divided into those that developed air leak (A=31) and no air leak group (NA=78) immediately after the operation and whether these patients received suction 1-Air leak on suction grp (leak and suction)=24 Air leak on water seal grp (leak no suction)=7 2- No air leak on suction grp (no leak suction)=46 No air leak on water seal grp (no leak no suction)=32Retrospective cohort study (level 3b)Chest tube duration (days)No suction groups 3.19+/-0.24 days

Suction groups 4.52 +/- 0.40 days

No randomization Retrospective review of current surgical practice
Hospital discharge (days)No suction groups 5.13 +/- 0.61 days

Suction groups 6.74 +/- 0.5



5 randomized controlled trials and 1 single center cohort study were identified which investigated the effect of using chest drain suction or water seal drainage in patients following pulmonary procedures. Alphonso, Treasure et al in 2005 performed a prospectively randomized 239 patients either to suction or water seal drainage following pulmonary procedures (mainly lobectomy or wedge resection). There was persistent air leak (lasting more than six days) in 7.8% of patients in the suction group and 10.1% of patients in the non-suction group. This difference was not significant. In addition there was no difference in air leak persistence on any of the previous 6 days. Of note this is the only study of the 5 that randomized patients immediately after the operation. As a result of this study, a policy of underwater seal drainage only has been adopted at Guy's Hospital in London. Brunelli et al in 2004 randomized 145 patients post-lobectomy who developed an air leak on the first postoperative day. The chest tubes were placed either on water seal or on 20cmH2O continuous suction. No significant differences were found between the two groups in terms of duration of air leak (6.5days versus 6.3days) and the incidence of prolonged air leak (27.8% versus 30.1%). There was an almost significant increase in all postoperative complications when water seal was compared to the suction group (31.9% versus 17.8%, p=0.056). Obscurely, as a result of this study, the authors decided to institute a policy of 10cmH2O suction at night and no suction in the day for their institution. Ayed in 2003 performed a randomized trial in Kuwait on 100 consecutive patients undergoing thoracoscopy for primary spontaneous pneumothorax. Patients were assigned into two groups to receive suction or water seal to their chest tubes after two hours of suction. 14% of patients in the suction group had prolonged air leak compared to 2% of those in the water seal group. Also, the duration of chest tube days in-situ and hospital stay was lower in the water seal group (2.7days) compared to the suction group (3.8days). These results were found to be statistically significant. Marshall et al in 2002 randomized 68 patients who underwent wedge resection, segmentectomy or lobectomy into a water seal group or suction group after return to the recovery room and a CXR while on suction. When corrected for staple lines, the duration of air leak (0.08days versus 0.17days) and mean times to removal of chest tubes (0.17days versus 0.32days) were significantly lower in the water seal group compared to the suction group. Hospital stay was also significantly shorter in the water seal group. Cerfolio et al have published multiple papers addressing the issue of air leaks in thoracic patients, including the production of an algorithm and an air leak classification system. In their prospective randomized controlled trial, 33 patients who developed an air leak following pulmonary resection were assigned to either suction or water seal 48 hours post surgery. Prolonged air leak (lasting more than 3 days) was found to be significantly reduced in the water seal only group with 12 of 18 stopping compared to only 1 out of 15 in the suction group. In addition they described a 7 point air leak scale and in their study, all 6 persistent air leak patients had a score of 4 or more. Antanavicius et al in 2005 retrospectively reviewed the duration of chest drains and hospital stay in 109 consecutive patients that underwent pulmonary procedures in their hospital in Pittsburgh, USA. Most patients with a post-operative air leak received suction but if no air leak was present, half of patients tended to have no suction. (Decision was based entirely on surgeon's preference and was not randomized). Chest tube and hospital stay duration were significantly lower in the no suction groups.

Clinical Bottom Line

Of the 6 studies presented, no studies found in favour of suction to reduce the incidence of air leak, 2 studies found no difference between the two strategies, and 4 studies found evidence that water seal drainage without suction reduced the incidence of air leak. 5 of the 6 studies used a short period of suction in the immediate post-operative period and the one study looking at immediate water seal drainage found no differences in outcome. Exceptions to the water seal strategy may be patients with a large air leak, or a large pneumothorax on CXR.


  1. Alphonso N, Tan C, Utley M, Cameron R, Dussek J, Lang-Lazdunski L, Treasure T. A Prospective Randomized Controlled Trial of Suction Versus Non-Suction to the Under-Water Seal Drains Following Lung Resection. European Journal of Cardio-Thoracic Surgery 27(3):391-4, 2005.
  2. Brunelli A, Monteverde M, Borri A, Salati M, Marasco RD, Al Refai M, Fianchini A. Comparison of Water Seal and Suction After Pulmonary Lobectomy: a Prospective, Randomized Trial. Annals of Thoracic Surgery 77(6):1932-7; discussion 1937.
  3. Ayed AK. Suction Versus Water Seal After Thoracoscopy for Primary Spontaneous Pneumothorax: Prospective Randomized Study. Annals of Thoracic Surgery 75(5):1593-6, 2003.
  4. Marshall MB, Deeb ME, Bleier JI, Kucharczuk JC, Friedberg JS, Kaiser LR, Shrager JB. Suction Vs Water Seal After Pulmonary Resection: a Randomized Prospective Study. Chest 121(3):831-5, 2002.
  5. Cerfolio RJ, Bass C, Katholi CR. Prospective Randomized Trial Compares Suction Versus Water Seal for Air Leaks. Annals of Thoracic Surgery 71(5):1613-7, 2001.
  6. Cerfolio R.J., Bryant AS, Singh S, Bass CS, Bartolucci AA. The Management of Chest Tubes in Patients With a Pneumothorax and an Air Leak After Pulmonary Resection. Chest 2005; 128(2): 816-20.
  7. Cerfolio R.J. Recent Advances in the Treatment of Air Leaks. Current Opinion in Pulmonary Medicine 2005; 11(4): 319-23.
  8. Cerfolio R.J., Tummala RP, Holman WL. A Prospective Algorithm for the Management of Air Leaks After Pulmonary Resection. Annals of Thoracic Surgery 1998; 66: 1726-31.
  9. Antanavicius G, Lamb J, Papasavas P, Caushaj P. Initial Chest Tube Management After Pulmonary Resection. American Surgeon 71(5):416-9, 2005.