Secondary Logo

Journal Logo

Original Article

Arterial oxygenation changes in valvular heart disease patients with cardiomegaly in different recumbent positions

Puri, G. D.1; Dutta, A.1; Chinnan, N. K.1; Thingnam, S. K. S.1; Sharma, S. K.2; Chari, P.1

Author Information
European Journal of Anaesthesiology: November 2005 - Volume 22 - Issue 11 - p 834-838
doi: 10.1017/S0265021505001407



Posture affects pulmonary gas exchange. In comparison to the erect position, a recumbent posture decreases lung volumes and increases closing volumes, thereby increasing airway resistance and reducing the airflow to cause hypoventilation of the dependent lung. Meanwhile, the dependent lung perfusion increases under the influence of gravity. This ventilation-perfusion mismatch may possibly cause hypoxaemia [1]. These changes are accentuated by obesity, cardiopulmonary pathology, breathing at low lung volumes and anaesthesia [2].

Studies in spontaneously breathing patients with unilateral lung disease have shown that arterial oxygenation improves with healthy lung in the dependent position [3-6]. In patients with bilateral lung disease arterial oxygenation is best in the right lateral decubitus position. This may be due to the higher anatomic volume of the right lung and the minimal compression of lungs by heart in this posture [6].

A few studies have documented the impact of an enlarged heart on regional ventilation and gas exchange. An enlarged heart can compress the left lower lobe in the supine position. This may become exaggerated in the left lateral position with a resultant decrease in dependent lung ventilation at a time when it is receiving predominant perfusion [7-10]. In infants with congenital heart disease, an enlarged left atrium has been shown to compress the bronchial tree [10]. In adults also, cardiomegaly decreases ventilation due to lung compression, predominantly of the left lower lobe [7-9]. At times, an enlarged cardiac chamber can directly compress the branches of the pulmonary artery also [11]. Thus, cardiac chamber enlargement can affect arterial oxygenation by altering either ventilation or regional perfusion of the lung.

In postoperative cardiac surgery patients, Banasik and colleagues have shown significantly higher arterial oxygen tension (PaO2) in the supine and right lateral postures when compared to the left lateral posture [12]. Since the change in PaO2 brought about by posture was small, they suggested that potential risks of lateral positioning in the early postoperative period outweigh the clinical benefits.

To date no prospective study has evaluated the effect of different recumbent positions on arterial oxygenation in patients awaiting open-heart surgery. To know whether a particular posture in bed during the preoperative period may benefit valvular heart disease patients with cardiomegaly, we studied their arterial blood gases in supine, left and right lateral decubitus postures.


This prospective study was carried out in the cardiothoracic operating theatre of the Post Graduate Institute of Medical Education and Research at Chandigarh, India. We studied the effects of different recumbent positions on arterial oxygenation in valvular heart disease patients with cardiomegaly, and related the postural changes in blood gas values to cardiothoracic ratio and left ventricular end-diastolic diameter (LVEDD).

Following institutional Ethics Committee approval and informed consent, 42 valvular heart disease patients of either sex, aged 18-60 yr and having cardiomegaly (cardiothoracic ratio ≥0.5 in pulmonary artery (PA) view plain chest X-ray) were included in the study [13]. Patients with localized lung pathology, severe obstructive or restrictive lung disease on pulmonary function testing, New York Heart Association Class-IV dyspnoea or SpO2 <90% on room air were excluded. LVEDD was measured in all patients from the preoperative echocardiogram. To avoid reporting bias and changes brought about by drugs, both cardiothoracic ratio and LVEDD were reported by the same physicians 1 day prior to surgery.

All patients were premedicated with 0.04 mg kg−1 lorazepam 2 h before the procedure. They received oxygen supplementation by ventimask at a fixed FiO2 of 0.35 during the study. Under strict asepsis and local anaesthesia, a 20 G radial artery cannula was inserted. Patients were asked to assume three different recumbent positions randomly, that is, supine, left and right lateral for 15 min each, and at the end of the period arterial blood gas samples were drawn and analysed immediately (Ciba-Corning 288 blood gas system; Medfield, MA).

Comparison of blood gas values in different positions was performed using paired ‘t’ test with Bonferroni correction. To determine whether posture brought about significant changes in arterial oxygenation, repeated measures of analysis of variance (ANOVA) was used. LVEDD and cardiothoracic ratio were used as covariates in ANOVA to analyse whether the postural change in arterial oxygenation were influenced by the extent of cardiomegaly. Pearson's correlation test and linear regression analysis were used to determine the relationship of postural changes in arterial oxygenation with cardiothoracic ratio and LVEDD. Data are shown as mean ± SD.


Patient characteristics data, preoperative diagnosis and planned surgical procedure are shown in Table 1. Both the PaO2 and haemoglobin saturation (SaO2) were significantly higher in the right lateral position (PaO2 = 120.6 ± 29.5 mmHg, SaO2 = 98.1 ± 1.4%) than in supine (PaO2 = 111.0 ± 30.6 mmHg, SaO2 = 97.6 ± 2.2%) and left lateral positions (PaO2 = 109.7 ± 32.0 mmHg, SaO2 = 97.6 ± 1.7%; P < 0.01). The difference between left lateral and supine positions was not significant (P > 0.05, paired ‘t’ test; Table 2).

Table 1
Table 1:
Patient characteristics data.
Table 2
Table 2:
Arterial oxygenation in different positions.

Repeated measures of ANOVA with LVEDD as a covariate showed a statistically significant change in PaO2 with posture (P = 0.011). PaO2 showed an inverse relationship with LVEDD in the three different recumbent postures (Pearson correlation coefficient, r = −0.36, P < 0.05 in supine, r = −0.57, P < 0.01 in left lateral and r = −0.32, P < 0.05 in right lateral decubitus, respectively).

The difference in oxygen tension (PaO2 R-L) and saturation (SaO2 R-L) with change of posture from right to left lateral showed a significant relationship with LVEDD (r = 0.50 and 0.60, respectively; P < 0.01). PaO2 R-L did not have a significant relationship with cardiothoracic ratio (r = 0.28, P > 0.05). Though SaO2 R-L showed a significant relationship with cardiothoracic ratio (r = 0.40, P < 0.05) it was not as strong as with LVEDD (r = 0.60, P < 0.01).

When the data from patients with isolated mitral stenosis (12 patients) were analysed separately, oxygenation in the supine position (PaO2 = 120.8 ± 31.6 mmHg, SaO2 = 98.1 ± 1.1%) was lower than in both left lateral (PaO2 = 130.3 ± 29.4 mmHg, SaO2 = 98.4 ± 1.0%) and right lateral positions (PaO2 = 127.6 ± 32.8 mmHg, SaO2 = 98.2 ± 1.2%). Change of posture from supine to left lateral brought about a significant improvement in oxygenation in these patients (P < 0.05, paired ‘t’ test).


The effect of posture on regional lung function has long interested pulmonary physiologists. Gravity causes a vertical gradient in the distribution of pulmonary blood flow such that dependent parts of the lung receive maximal perfusion. In the lateral decubitus position, the dependent lung receives greater blood flow than the non-dependent lung. In awake spontaneously breathing patients, the dome of the lower diaphragm is pushed higher into the chest allowing it to contract more efficiently. Thus, there is no significant ventilation-perfusion mismatch in these patients as both ventilation and perfusion are better in the dependent lung [14]. However, in patients with cardiomegaly, the enlarged cardiac chamber may compress the dependent lung in the left lateral position and negate the advantage of the higher position of the dependent dome of the diaphragm [7-9].

In our study, we found that change of posture caused a significant change of PaO2 and SaO2. This maybe either due to the enlarged cardiac chamber compressing the left main stem bronchus, minor bronchi and lung parenchyma to cause a simple airspace volume reduction or generation of more low ventilation-perfusion areas on a congestive basis with an increase in the post capillary pulmonary vascular resistance or both [7-9]. In supine and left lateral positions, these effects are exaggerated. Thus, PaO2 and SaO2 showed an improvement in the right lateral position.

Previous studies have also found avoidance of left lateral decubitus position during sleep [15] and improvement in level of comfort on assuming right lateral decubitus position in patients with congestive cardiac failure and cardiomegaly [16]. This was attributed to the lower left ventricular preload [16] and decreased cardiac sympathetic activity [17] in the right lateral decubitus position. Our study suggests an improvement in the arterial oxygenation in the right lateral decubitus position could have also contributed to the increased patient comfort in that posture.

PaO2 R-L had a significant relationship with LVEDD (r = 0.50, P < 0.01), but not with cardiothoracic ratio (r = 0.26, P > 0.05; Fig. 1). The relationship of SaO2 R-L with LVEDD (r = 0.60) was stronger than with cardiothoracic ratio (r = 0.40, Fig. 2). This may be due to the fact that some patients with cardiothoracic ratio >0.5 did not have an enlarged left ventricle [18]. The cardiomegaly noted in their chest radiographs was due to other cardiac chamber enlargement.

Figure 1.
Figure 1.:
Relationship between LVEDD and change in PaO2 with change of posture from right to left lateral (PaO2 R-L). r2 = 0.0256, coefficient of linear regression ( P < 0.05).
Figure 2.
Figure 2.:
Relationship of LVEDD with change in SaO2 with change of posture from right to left lateral (SaO2 R-L). r2 = 0.036, coefficient of linear regression ( P < 0.05)

Though most patients included in our study had an enlarged left ventricle as shown by a LVEDD >45 mm m−2, 12 patients (28%) had predominant mitral stenosis without other valvular lesions. Unlike in mitral or aortic regurgitation patients with an enlarged left ventricle, mitral stenosis patients have a normal sized left ventricle and cardiomegaly on their chest X-ray is probably due to the enlarged right atrium. Patients with severe mitral stenosis with severe pulmonary arterial hypertension showed a lack of improvement or a slight deterioration in arterial oxygenation when placed in the right lateral position. A change of posture from supine to left lateral brought about a significant improvement in these patients. This is possibly due to the posteriorly located enlarged left atrium compressing the lung parenchyma, bronchi or pulmonary vasculature in the supine position. This is in contrast to other valvular heart disease patients with an enlarged left ventricle whose oxygenation actually worsened in the left lateral position.

Our study suggests that cardiac chamber enlargement influences arterial oxygenation in different positions and forms the basis for a larger trial on this topic. One may argue that although these changes are statistically significant, their clinical implications are trivial in our patient population as they all had normal arterial oxygenation. This is because we excluded patients with room air saturation <90% and gave supplemented oxygen throughout the study. Most of our patients had severe pulmonary hypertension and any hypoxaemia during the study period would have resulted in right heart decompensation. Arterial oxygen levels of most of our patients were in the upper flat portion of the oxygen-dissociation curve. The maximum improvement in arterial oxygenation with a change of posture may be expected in those patients whose PaO2 levels lie in the steep portion of the oxygen dissociation curve, that is, in sicker patients than included in our study.

Change in position can increase the blood flow to the dependent lung and bring about changes in systemic and pulmonary haemodynamics and minute ventilation. To minimize the impact of these factors a choice of 15 min was made before recording blood gases after change of position. Another factor that could have possibly affected the results of our study is the increased perfusion of the right lung in the right lateral position (60% of total pulmonary blood flow) with better ventilation-perfusion matching during spontaneous breathing [14]. This may also improve arterial oxygenation in the right lateral position. Despite the limitations, based on the above evidence, we suggest that valvular heart disease patients with an enlarged left ventricle may be given a right lateral posture in bed in the preoperative period as one of the initial therapeutical measures to improve oxygenation.


We are indebted to the patients and their families for co-operating in our study. We also thank our colleagues in Departments of Anaesthesia and Intensive Care, and Cardiothoracic Surgery for their co-operation and encouragement during the course of the study.


1. Dean E. Effect of body position on pulmonary function. Phs Ther 1985; 65: 613-618.
2. Ross J, Dean E. Body positioning. In: Zadaic C, ed. Clinics in Physical Therapy: Pulmonary Management in Physical Therapy. New York, USA: Churchill Livingstone Inc, 1992: 78-98.
3. Remolina C, Khan AU, Santiagotv TV, Edelman NH. Positional hypoxemia in unilateral lung disease. New Engl J Med 1981; 304: 523-525.
4. Sonnenblick M, Melzer E, Rosin AJ. Body positional effect on gas exchange in unilateral pleural effusion. Chest 1983; 83: 784-786.
5. Neagley SR, Zwillich CW. The effect of positional changes on oxygenation in patients with pleural effusions. Chest 1985; 88: 714-717.
6. Zack MB, Pontoppidan H, Kazemi H. The effect of lateral position on gas exchange in pulmonary disease: a prospective evaluation. Am Rev Respir Dis 1974; 110: 49-55.
7. Weiner CM, Mckenna WJ, Lavender JP, Hughes JMB. Left lower lobe ventilation is reduced in patients with cardiomegaly in the supine but not prone position. Am Rev Resp Dis 1990; 14: 1150-1155.
8. Mark SM, Arnot RN, Lavender JP. Left lower lobe ventilation and its relation to cardiomegaly and posture. BMJ 1989; 299: 94.
9. Alexander MS, Peters AM, Cleland JP, Lavender JP. Impaired left lower lobe ventilation in patients with cardiomegaly: an isotope study of mechanisms. Chest 1992; 101: 1189-1193.
10. Corr L, Mccarthy PJ, Lavender JP, Hallidie-Smith KA. Bronchial compression by an enlarged left atrium in infants: a cause of hypervascularity of the left lung. Paed Radiol 1988; 18: 459-463.
11. Beckman CF, Levin DC, Ulreich S. Cardiomegaly as a cause of non-uniform pulmonary artery perfusion. Am J Roentgenol 1977; 129: 661-666.
12. Banasik JL, Emerson RJ. Effect of lateral position on arterial and venous blood gases in postoperative cardiac surgery patients. Am J Crit Care 1996; 5: 121-126.
13. Raphel MJ, Donaldson RM. The normal heart: methods of examination. In: Sutton D, ed. Textbook of Radiology and Imaging, 5th edn. London, UK: Churchill Livingstone Inc, 1998: 541-565.
14. Benumof JL, Alfrey DD. Anaesthesia for thoracic surgery. In: Miller RD, ed. Anaesthesia, 5th edn. Philadelphia, USA: Churchill Livingstone, 2000: 1665-1752.
15. Leung RS, Bowman ME, Parker JD, Newton GE, Bradley TD. Avoidance of the left lateral decubitus position during sleep in patients with heart failure: relationship to cardiac size and function. J Am Coll Cardiol 2003; 41: 227-230.
16. Berensztein CS, Pineiro D, Luis JF, Iavicoli O, Lerman J. Effect of left and right lateral decubitus positions on Doppler mitral flow patterns in patients with severe congestive heart failure. J Am Soc Echocardiogr 1996; 9: 86-90.
17. Miyamoto S, Fujita M, Sekiguchi H et al. Effects of posture on cardiac autonomic nervous activity in patients with congestive heart failure. J Am Coll Cardiol 2001; 37: 1788-1793.
18. Hada Y. Cardiothoracic ratio. J Cardiol 1995; 26: 51-54.


© 2005 European Society of Anaesthesiology