• 
• 

Three Part Question

In [patients undergoing median sternotomy] is [lung deflation or direction of incision] important in the reduction of [inadvertent pleurotomy].

Clinical Scenario

You are performing an aortic valve replacement in a 78-year-old lady, with poor lung function, and who was smoking up until the day of the operation. You are keen to keep the pleura intact for this operation to optimise her post-operative recovery. During the sternotomy you ask the anaesthetist to deflate the lungs and you perform the sternotomy from the sternal notch to the xiphisternum. You are disappointed to find that despite these manoeuvres, you have widely opened the right pleura with the saw. The anaesthetist comments that deflating the lungs makes no difference and that you should have gone the other way with the saw as a colleague does this and 'never' has this problem. You resolve to search for the evidence for these comments.

Search Strategy

Medline1966 to Jan 2005, Embase 1980 to Jan 2005 and CINAHL 1982 to Jan 2005 using the OVID interface. The Cochrane database of systematic reviews was also searched.
[exp sternotomy/OR sternotomy.mp] AND [exp pleura/OR pleura.mp OR pleural.mp OR pleurotomy.mp]

Search Outcome

One hundred and four papers were found in Medline, 64 papers in Embase, and 2 papers in CINAHL. Cochrane reviews were searched using the term sternotomy or pleurotomy but yielded no relevant papers. Of these papers only 4 provided evidence to answer the question

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Pick et al, 1998, Australia	95 consecutive patients at a single centre undergoing cardiac surgery Group I (n=49) sternotomy from sternal notch downwards Group II (n=46) Sternotomy from Xiphoid upwards Lungs were deflated	Prospective cohort study (level 2b)	Accidental pleurotomy	Group I (Sternum down) 3/49 (7%). Group II (Xiphoid up) 11/46 (24%) P=0.02. (93% of accidental pleurotomies were the right pleura)	Not a randomised study One surgeon always sawed downwards and a second surgeon commenced the study sawing from the xiphoid then changed to the opposite technique halfway
			Mean days in hospitals	No accidental pleurotomy (n=81) 6.9 days. Accidental pleurotomy (n=14) 7.8 days P=NS.
Lichtenstein et al, 1994, Canada	126 cardiac patients in a single centre Anaesthetist opened envelopes to reveal patient group, and disconnected lungs without any audible change to surgeon Deflated lung group (n=66) 3 excluded due to protocol violation (2 redos and one misread envelope)	Single blinded PRCT (level 2b)	Inadvertent pleurotomy	Deflated lungs 9/60 (15%). Inflated lungs 6/66 (9%) P=0.455.	Authors should have used Fisher's exact test not Chi squared test to assess significance due to low cell count in groups for all cases.
			Inadvertent pleurotomy by sternotomy direction (non-randomized)	Sternal notch down 12/57 (21%). Xiphoid process up 3/69 (4%). P=0.009 by Chi squared test.
Ronday et al, 1993, Holland	666 patients undergoing first time sternotomy for cardiac surgery	PRCT (level 1b)	Incidence of accidental pleurotomy	Lungs deflated 51/330 (15.5%). Lungs inflated 47/336 (14%) P=NS	Flawed randomisation technique
			Risk factors for accidental pleurotomy	COPD, use of positive pressure ventilation, age and sex were not risk factors. Incidence of pleurotomy varied from 5.9% to 24% among 6 surgeons (P<0.001)
Stock et al, 1986, USA	38 patients undergoing valve or vein only CABG Post-operative functional residual capacity, FEV1 and FVC measured	Cohort study (level 4)	Incidence of Accidental Pleurotomy Functional Residual Capacity post-op	31/38 (82%) patients underwent accidental pleurotomy. Pleurotomy 1,470±640 ml. No pleurotomy 1,410±420 ml. P=NS	Non-randomised study Very small numbers, and no sample size to justify this No patients undergoing CABG with LIMA harvest used No indication as to how the pleura was damaged
			Forced Vital Capacity (FVC)	Pleurotomy 1,500±700 ml. No pleurotomy 1,1,590±610 ml. P=NS

Comment(s)

While several studies presented evidence discussing the effects of an accidental pleurotomy on lung function or outcome, only 4 papers studied the impact of either disconnecting the lungs or direction of sternotomy, and thus only these 4 papers were reviewed. Ronday et al performed the largest PRCT in this area. 666 patients were randomized to either receiving pleurotomy with the lungs disconnected from the ventilator, or randomized to continued ventilation. There were 98 accidental pleurotomies and the incidence was 15.5% in the lungs deflated group and 14% in the lungs inflated group. In addition they could find no risk factors for predicting accidental pleurotomy, and COPD, use of positive pressure ventilation age and sex had no impact. Only the operating surgeon influenced the likelihood of accidental pleurotomy. In 1998 Pick et al performed a prospective cohort study into the incidence of accidental pleurotomy according to the direction of sternotomy. One of the surgeons performed all his sternotomies from the Sternal notch downwards. A second surgeon initially performed all his ster-notomies from the Xiphiod process upwards. Halfway through the study he changed his technique to follow that of the first surgeon. Both surgeons required the lungs to be deflated and both surgeons performed a digital displacement of the retro-sternal or manubrial structures at the start of the incision. There were 11 accidental pleurotomies in the Xiphoid upwards group (24%), but only 3 accidental pleurotomies in the sternal notch downwards group. This study is unfortunately significantly flawed by its lack of randomisation, small number of surgeons and the clear bias of the second surgeon who was obviously convinced that his original technique was inferior to his new technique, and thus changed technique halfway through the study. In 1994 Lichtenstein et al. reported their findings of a single blinded randomized controlled trial assessing whether lung deflation reduced the incidence of accidental pleurotomy. The anaesthetist opened an envelope and according to randomization either left the lungs ventilating normally or disconnected the ET tube while insuring that there was no indication to the surgeon that this had been done. Of 126 patients randomized into the study there was a 15% accidental pleurotomy rate in the lungs deflated group, and a 9% accidental pleurotomy rate in the lungs inflated group. Thus there was no difference found whether the lungs were inflated or not. As a sub-analysis, surgeons were allowed to use their discretion in the direction in which they performed the sternotomy. There was a 21% incidence of accidental pleurotomy in the sternal notch down group but only a 4% incidence when sawing from the Xiphoid process upwards. It should be remembered that this part of the study was not randomized and therefore may be open to considerable surgeon related operator bias. The only other study found was by Stock et al in 1986. They performed a cohort study and found that 82% of patients undergoing valve surgery or CABG using vein grafts only, had their pleura accidentally opened. However, they did not record which direction the sternotomy was performed and all patients had the ventilation continued for sternotomy. This was a study mainly looking at the significance of accidental pleurotomy, which is an issue which we specifically did not address as it has been covered in a previous Best Evidence Topic. These authors could find no significant difference but this is not surprising as there were only 7 patients in their 'no=pleurotomy' group. Therefore in the 2 papers that assess the impact of deflating the lungs prior to sternotomy, neither support its continued practise. Of the 2 papers that looked at the direction of sternotomy, Pick et al supported sternotomy from the sternal notch, from a non-randomized study of 95 patients and Ronday supported the use of sternotomy form the Xiphoid process upwards from a non-randomised study of 666 patients. Thus no convincing evidence for either method can be supported and the only clear message was that the incidence of accidental pleurotomy was highly surgeon dependent.

Clinical Bottom Line

Disconnection of the ventilator prior to sternotomy cannot be supported as a strategy to reduce the incidence of accidental pleurotomy. In addition there is little evidence to support sternotomy from the Xiphoid process upwards over sternal notch downwards.

References

1. Pick A, Dearani J, Odell J. Effect of sternotomy on the incidence of inadvertent pleurotomy. J Cardiovasc Surg 1998;39:673–676.
2. Lichtenstein SV, Abel JG, Miyagishima RT, Ling H, Warriner CB, Stilwell ME, Thompson CR. Effect of lung inflation and sternotomy direction on pleural space violation. Ann Thorac Surg 1994;58:1734–1737.
3. Ronday M, Damen J, Van der Tweel I. Disconnection of the ventilatory system does not prevent pleural lesions during sternotomy. J Cardiothorac Vasc Anesth 1993;7:535–537.
4. Stock MC, Downes JB, Weaver D, Lebenson IM, Cleveland J, McSweeney TD. Effect of pleurotomy on pulmonary function after median sternotomy. Ann Thorac Surg 1986;42:441–444.