| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
Interact CardioVasc Thorac Surg 2008;7:819-824. doi:10.1510/icvts.2008.176560 © 2008 European Association of Cardio-Thoracic Surgery
Arterial oxygen partial pressure and cardiovascular surgery in elderly patients
a Laboratoire de Physiologie, Faculté de Médecine de Nancy, EA 3450, Centre Hospitalier Universitaire de Nancy, 9, avenue de la Forêt de Haye, B.P. 184, 54505 Vandoeuvre-lès-Nancy, France Received 28 January 2008; received in revised form 3 June 2008; accepted 3 June 2008
*Corresponding author. Tel.: +33 3 83 68 37 45; fax: +33 3 83 68 37 39.
Arterial blood gas assessment is still routinely performed in candidates for a cardiovascular surgery. Whether sampling arterial blood is useful in an elderly patient with a near normal lung function and who meets all other criteria for operability, is unknown. Therefore, it was our purpose to provide reference values for arterial blood gases in these patients and to examine how the level of arterial oxygen partial pressure (PaO2) might influence postoperative outcome. We retrospectively studied arterial blood gases in 201 patients, aged 70–92 years with normal or near normal ventilatory function awaiting a planned cardiovascular surgery. PaO2 averaged 81.6±7.6 mmHg and PaCO2 averaged 37.7±3.2 mmHg. Both were independent of age. Factors associated with mortality according to bivariate analysis were: gender (female), type of surgery (valve replacement), and a low PaO2 with strictly no ventilatory abnormality. In conclusion, PaO2 values in elderly patients with cardiac disease and normal ventilatory function are greater than those obtained by extrapolation from healthy younger subjects. PaO2 measurement should be recommended prior to cardiovascular surgery in elderly patients since a low PaO2 with strictly normal ventilatory function is significantly associated with an increased risk for postoperative mortality.
Key Words: Arterial blood gases; Cardiovascular surgery; Elderly; Mortality risk; Pulmonary function test
With the progressive increasing life expectancy, a growing number of elderly people is liable to undergo a cardiac surgical procedure [1, 2]. The normalcy, or otherwise, of the pulmonary function is crucial to determine in candidates for a cardiovascular surgery. This preoperative evaluation becomes essential in elderly patients since extensive comorbidity presented in this population greatly influences their surgical outcome [3]. This question is even more critical for resting arterial blood gas tension, which remains the cornerstone in the evaluation of respiratory disease severity. However, a rational, data based, preoperative strategy in elderly is still needed [4]. Reference values for arterial oxygen partial pressure (PaO2) in subjects older than 70 years have mostly been obtained by extrapolation of the linear regressions assessed in younger subjects [5]. Such an extrapolation appears to be misleading, since in contrast to the decrease in PaO2 with aging obtained from such a prediction, PaO2 of healthy old people does not differ considerably from that of middle age individuals [4, 6]. However, the possible relevance of such findings is problematic since the concept of ‘healthy’ elderly people is not simple to define and may not be useful when dealing with surgical decision in patients over 80 years old. Indeed, all the studies which have tried to assess ‘normal’ values for arterial blood gases in the elderly have logically excluded comorbidities, such as candidates for a cardiovascular surgery, even if their pulmonary function is not altered [7]. Moreover, numerous preoperative risk scores have been developed and are commonly used to predict mortality after heart surgery but the specific role of a low PaO2 in a sample of elderly patients has not been yet evaluated as an independent preoperative risk factor [8]. Therefore, we analyzed arterial blood gases from a large population of elderly subjects with cardiovascular disease but normal pulmonary function, prior to a planned cardiovascular surgery. We intended to determine if sampling arterial blood in these patients is still relevant in the preoperative strategy, when non-invasive criteria for surgery are already met. Therefore, our aim was two-fold: 1) to determine blood gas values in a preoperative elderly population with cardiovascular disease but for whom surgery was proposed based on the absence of clinical or non-invasive functional respiratory abnormalities, 2) to assess the relationship between preoperative arterial blood gas status, lung function and short-term outcome after surgery (i.e. morbidity and mortality during the first month following surgery).
2.1. Patients We retrospectively studied 201 patients over 70 years old. All patients were examined in our department from January 1999 to June 2003 to assess their preoperative pulmonary function before a planned cardiovascular surgery. The surgical procedures were: aortic valve replacement (n=57), mitral valve replacement (n=19), coronary artery bypass grafting (n=72), aortic aneurysm repair (n=39), femoral artery bypass (n=14). All subjects with symptoms or diagnosis of active respiratory, renal, hepatic, neurologic, hematologic or metabolic diseases; patients with a known chronic respiratory disease or acute pulmonary edema, were excluded. Thus, only patients with no ventilatory abnormalities on their pulmonary function tests defined by pulmonary volumes, forced expiratory volume in one second (FEV1) and forced vital capacity (FVC) higher than 80% of the expected value and a FEV1/forced vital capacity (FVC) ratio >70%, were included.
2.1.1. Pulmonary function testing
2.1.2. Arterial sampling
2.1.3. Blood gas analysis All patients were informed on the risks of the radial artery puncture and gave their verbal consent. The use of the results for research purpose, obtained in these patients, was approved by our Local Ethics Committee. Results are expressed as mean±S.D. Patient characteristics, preoperative arterial blood gas status, lung function relationships with short-term outcome after surgery (morbidity and mortality) were analyzed using Pearson 2, Fisher exact, analysis of variance, and Wilcoxon Kruskal–Wallis. Variables significant at a 0.05 level were subsequently used in multivariate analyses (logistic regression model) of the factors associated with short-term outcome after surgery. Data were recorded on Excel files. Statistical analysis was performed using SAS veriosn 9.1 statistical software.
3.1. Sample description Two hundred and one patients (71 women and 130 men) were studied, aged from 70 to 92 years. The mean age was 77.2±4.7 years and 33.8% of them were aged 80 or over. Only 4% of patients can be classified as morbidly obese with a body mass index of 33 or over. Demographic information is summarized in Table 1.
3.2. Spirometic values Lung volumes are reported in Table 2.
The patients awaiting a valve replacement showed a significantly lower TLC and VC than other groups (from P=0.0007 to P=0.04). According to the criteria of inclusion, FEV1 ranged from 80 to 155% of the predicted value (mean: 105.4±15.3%) and FVC ranged from 80 to 144% (mean: 102.4±15%). FEV1/FVC ranged from 70 to 100% (mean: 78.5±5.3%). Patients with mitral valvulopathy had the lowest, albeit still within the expected range, FEV1 and FVC (Table 2).
Despite a normal FEV1/FVC ratio, 93 patients demonstrated a decrease of more than 20% of the expected value of the Maximal mid-expiratory flow rate (FEF25/75%), leading to a concave shape of the flow/volume curve. This ‘concave flow/volume curve’ group had a significant lower (although within the expected range) value in FVC, FEV1 and FEV1/FVC ratio (Table 2)
3.3. Arterial blood-gases The average PaO2 was 81.6±7.6 mmHg. The frequency distribution is shown on Fig. 1. There was no difference in mean PaO2 between women and men or between the clinical subgroups (Table 4). Table 5 presents reference values for arterial blood gases distinguished by sex in our population.
However, PaO2 of the group with a concave flow/volume curve was found significantly less than in the other groups (80.4±7.5 vs. 82.6±7.6 mmHg, P=0.0397). No significant correlation was found between arterial oxygen tension and age. There were significant correlations between PaO2 and VC (% VC=79.18+0.30 PO2; r2=0.02; P=0.047) and between PaO2 and FEV1 (PaO2=76.40+2.30 FEV1; r2=0.034; P=0.0087). There was a significant negative correlation between PaO2 by BMI: PaO2=94.14–0.48 BMI; r2=0.056 (P=0.008). The arterial dioxide partial pressure remained unchanged with aging, regardless of the cardiovascular disease, sex or aspect of the flow-volume curve (on average 37.7±3.2 mmHg). There were no correlation between PaCO2, spirometric values and BMI. 3.4. Short-term postoperative outcome One hundred and twenty-four patients were surgically treated in our hospital and could be followed (operated group: OP=61.7%) while 28 patients were denied, or refused surgery (non-operated group: NOP=13.9%). The 49 remaining patients were not treated in our hospital and we were not able to assess their postoperative outcome (no follow-up group: NF=24.4%).There was no significant difference in PaO2 between these three groups. In the OP group, 63 patients had no specific postoperative problems (averaged hospital stay's length: 8.9±2.4 days). Forty-six patients presented postoperative complications which led to a significant increase in the length of stay (average: 18±8 days). The postoperative complications consisted of: respiratory infection (22), life-threatening cardiac arrhythmia (6), excessive bleeding (5), persistent low cardiac output (3), sternal dehiscence (3), renal failure requiring dialysis (2), respiratory failure (2), septicemia (1), mediastinitis (1), stroke (1). Fourteen died within the first month after the surgical procedure. The causes of death were: multivisceral failure (7), postoperative valve rupture (3), respiratory distress syndrome (2), myocardial infarction (2). Mean PaO2 was not different between survivors and non-survivors. In the group of survivors, no difference was found in PaO2 between patients with simple and complicated outcomes (82.1±7.4 mmHg for both). Results reveal statistically significant relationships between mortality and the type of surgery (valve replacement, P=0.014), gender (female, P=0.021), and a low PaO2 with strictly no ventilatory abnormality on the flow/volume curve (P=0.0125) (Fig. 2).
This study provides adequate reference values for PaO2 for elderly patients with cardiovascular disease with near normal spirometric values prior to a surgical treatment. These patients represent one of the most common populations of elderly people who are liable for a cardiovascular surgical procedure. 4.1. What reference values for PaO2 in the elderly? The traditional view assumes that physiological PaO2 values decrease with aging [5]. However, recent studies have already pointed out that PaO2 in subjects above 70 years is higher than expected from extrapolation [6, 7]. Therefore, it appears that using linear regression of PaO2 against age to establish reference values in the elderly leads to a systematic underestimation [9]. Our results clearly show that even in a population with a known cardiac disease and no or slight alteration of their ventilatory function, the averaged PaO2 is as high as 80 mmHg. More precisely, 124 patients (61.7%) had a PaO2 higher than 80 mmHg and only 11 patients (5.5%) had a PaO2 lower than 70 mmHg. Thus, the normalcy of PaO2 in the general healthy elderly population should be considered at least equal or superior to this value.4.2. Relationship between spirometric values and PaO2 Despite a normal FEV1 and FEV1/FVC ratio, close to 50% of the patients demonstrated a concave aspect of the flow/volume curve reflected by a reduction in FEF25;b175%. In addition we found: 1) a correlation between normal FEV1 and PaO2 (P=0.0015) as patients with chronic obstructive pulmonary disease [10], and 2) a significant reduction in PaO2 in the patients with a reduction in FEF25–75% (‘concave flow/volume curve’). However, the difference was very small and PaO2 remained high. The relevance of such a link as well as the meaning of this concave pattern observed in some patients is unclear. It could be related to a specific effect of aging i.e. an age-related decrease in elastic recoil [11]. We cannot rule out the role of the cardiac diseases in our patients, as previously shown by Fowler et al. [12], since left ventricular failure and mitral stenosis may have caused airway narrowing and bronchial hyperresponsiveness [13], even without any obvious pulmonary edema [14]. This is supported by the fact that the ‘concave flow/volume curve’ is significantly more pronounced in patients with mitral disease.4.3. Short-term postoperative outcome and PaO2 We report a short-term postoperative mortality of 11.3% for the overall surgical procedures, in agreement with the literature [1, 15]. Our statistical analysis showed that the preoperative risk factors associated with operative mortality were female sex and valve surgery. Age per se and PaO2 were not a predictor but we found that a low PaO2 with no ventilatory abnormalities of the flow/volume curve was a risk factor for postoperative mortality. Thus, the main objective to systematically assess PaO2 prior to planned heart surgery is to diagnose a clear hypoxemia which is not linked to a known respiratory disease, requiring further investigations to assess its origin and ideally its treatment before surgery.In conclusion, PaO2 in elderly people with cardiovascular disease awaiting surgery and with no or slight ventilatory abnormalities have been shown systematically greater than those obtained by classical extrapolation from younger subjects, averaged 80 mmHg. A decrease in PaO2 below 70 mmHg in this population is therefore to be regarded as hypoxemia, which needs further investigations since it seems to be associated with a higher risk of postoperative mortality when isolated.
The authors wish to thank Ms. Massin for the technical assistance.
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
Top
Abstract
1. Introduction
