This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation

Arterial Blood Gas Analysis and Alveolar-Arterial Oxygen Gradient in Diagnosis and Prognosis of Elderly Patients With Suspected Pulmonary Embolism

^a Institute of Internal Medicine and Geriatrics, University of Siena, Italy
^b Department of Quantitative Methods, University of Siena, Italy

L. Masotti, Istituto di Medicina Interna e Geriatria, Policlinico Le Scotte, Viale Bracci 1 53100 Siena, Italy E-mail: images{at}unisi.it.

Decision Editor: John E. Morley, MB, BCh

	Abstract

Top Abstract Methods Results Discussion References

 
Background. Arterial blood gas analysis (BGA) remains a first-step diagnostic approach in patients with suspected pulmonary embolism (PE). The aim of this study was to evaluate BGA parameters in elderly patients with suspected pulmonary embolism for diagnosis and 14-day prognosis.

Methods. We performed a retrospective cohort observational study of 6 years (1994–1999) in a 60-bed acute geriatric ward of University Hospital in Siena, Italy. Room air arterial oxygen partial pressure (pO₂), arterial carbon dioxide partial pressure (pCO₂), pH, arterial oxyhemoglobin saturation (SO₂), and alveolar-arterial oxygen gradient [D(A-a)O₂] were performed on hospital admission of 75 patients with confirmed PE (CPE) and were compared with data from 43 patients with unconfirmed PE (UCPE). The same parameters of 54 CPE surviving patients were compared with 21 CPE nonsurviving patients.

Results. Significantly lower pO₂ and SO_2, and higher DA-aO₂ were found in CPE patients. Respiratory alkalosis was found in one third of the patients in both groups (no significant difference). In the CPE group, there was a significantly lower SO₂ in nonsurviving patients, without significant differences for the other parameters. Metabolic acidosis was significantly more frequent in nonsurviving patients.

Conclusion. More severe hypoxemia, oxyhemoglobin hyposaturation, and higher D(A-a)O₂ are associated with the diagnosis of PE in elderly patients. Respiratory alkalosis is less frequent than in younger patients, and metabolic disorders are negative prognostic indicators.

PULMONARY embolism (PE) remains an underdiagnosed disease, even though it has a high incidence and mortality, especially in the elderly. Arterial blood gas analysis (BGA) should be performed as a first-step diagnostic approach in suspected pulmonary embolism, but its results are nonspecific ⁽¹⁾. It has been reported that the typical picture of BGA in young patients with PE is hypoxemia, hypocapnia, arterial oxyhemoglobin hyposaturation, increased alveolar-arterial oxygen gradient [D(A-a)O₂ = FiO₂(barometric pressure -47 mm Hg) - 1.25pCO₂-pO₂ = 150 - 1.25pCO₂-pO₂] and respiratory alkalosis ⁽²⁾. However, there is consensus that normal arterial oxygen partial pressure (pO₂) and D(A-a)O₂ do not exclude the diagnosis of PE. There have been few studies regarding BGA and D(A-a)O₂ in elderly patients with pulmonary embolism ⁽³⁾. Therefore, the aim of this study was to evaluate these aspects in elderly patients with suspected PE, both for diagnosis and for 14-day prognosis.

	Methods

Top Abstract Methods Results Discussion References

 
For this study, we enrolled 118 patients (69 women and 49 men) with a mean age (±SD) of 77.7 ± 7.1 years. All were at least 65 years old, had been admitted to the acute geriatric ward of University Hospital, Siena, Italy, in the last 6 years (1994–1999), and had undergone a scintigraphic lung scan for suspected pulmonary embolism. We did not exclude patients with pre-existing cardiopulmonary diseases.

We recorded the values of room air arterial BGA parameters [i.e., pO₂, arterial carbonic dioxide partial pressure (pCO₂), pH, arterial oxyhemoglobin saturation (SO₂), and D(A-a)O₂], performed in the first 12 hours after hospital admission and before performing the lung scan. We compared the data of 75 patients (mean age ± SD, 78.5 ± 7.5 years) with a high probability of PE on lung scan according to the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED) criteria ⁽⁴⁾ (confirmed pulmonary embolism [CPE] group) with the data of 43 patients (mean age ± SD, 76.3 ± 6.2 years) with a normal lung scan (unconfirmed pulmonary embolism [UCPE] group). We defined as normal values pO_2, >75 mm Hg; pCO_2, 38–42 mm Hg; pH, 7.38–7.42; SO_2, >90%; and D(A-a)O_2, <25 mm Hg.

We then compared the same BGA parameters between the surviving patients with CPE (54 patients, 72%; age ± SD, 77.9 ± 7.7 years) and the nonsurviving patients with CPE (21 patients, 28%; age ± SD, 80.3 ± 6.9 years).

Blood samples were taken from the radial or femoral artery with a sterile, disposable 1-ml syringe, and measurements were made within 5 minutes after sampling. The samples were analyzed on an automated, computerized blood gas, acid-base, and electrolytes analyzer (BGA/Electrolytes, IL Instrumentation Laboratory, Milan, Italy).

In the statistical analysis, we used a novel technique based on a permutational procedure that combines marginal tests of each considered variable ⁽⁵⁾. Because the technique is based on permutational procedures, the different sizes of the groups have no effect on the inference. With this method, the null hypothesis of homogeneity of the groups with respect to the constant considered variables is decomposed into the intersection of constant marginal hypotheses, each one dealing with the homogeneity of the groups with respect to the relevant marginal variable. Accordingly, constant marginal test statistics can be used to assess the respective marginal hypotheses. Subsequently, the significance of each marginal statistic is obtained by a permuta-tional procedure, and the resulting significance values are combined in order to obtain an overall statistic. In turn, the significance of the final statistic is computed using the same permutations of data. In this way, the dependence structure of the marginal statistics is nonparametrically captured by the permutation procedure. Thus, if the overall hypothesis is rejected, one can determine on which marginal variables the rejection depends. Moreover, because the overall test statistic is merely a function of the significance values of the marginal test statistics, the marginal hypotheses can be assessed using mixed-type test statistics, thus allowing joint analysis of mixed-type variables. A p value <.05 was considered significant.

	Results

Top Abstract Methods Results Discussion References

 
We found significantly lower mean values of pO₂ and SO₂ and significantly higher mean values of D(A-a)O₂ in the CPE group. We found no significant difference in pCO₂ or pH between the two groups (Table 1 ). Table 2 shows the distribution of pO₂, pCO₂, and D(A-a)O₂ in both groups. We found a picture of acute respiratory failure in 72% of CPE patients and in 44% of the UCPE group (p < .05). Hypoxemic-type respiratory failure was present in 45% of the CPE group and in 18.5% of the UCPE group (p < .05); hypoxemic/hypercapnic-type failure was present in 21% of CPE patients and in 14% of UCPE patients (p < .05); and hypercapnic-type failure was present in 5% of CPE patients and in 11.5% of UCPE patients (p < .05). Fig. 1 reports the acid-base status in the two groups. We found respiratory alkalosis in one third of both groups with no significant difference between them. Metabolic alkalosis was significantly more frequent in the UCPE group.

View larger version (17K): [in this window] [in a new window]   Figure 1. Acid-base status in diagnosis of pulmonary embolism. CPE, confirmed pulmonary embolism; UCPE, unconfirmed pulmonary embolism.

View larger version (19K): [in this window] [in a new window]   Figure 2. Acid-base status in short-term prognosis.

	Discussion

Top Abstract Methods Results Discussion References

 
Arterial BGA remains a necessary examination in patients with clinical suspicion of PE, both for evaluation of oxygen and carbon dioxide gas exchange and acid-base status, even though it has been demonstrated that its findings are nonspecific ⁽⁶⁾.

It should be noted that in healthy elderly patients, pO₂ seems to decrease with age (0.33 mm Hg/y) following the equation pO₂ = 109 - (0.43 x age in years) ⁽⁷⁾ with a tendency to plateau after the 75th year ⁽⁸⁾, while the alveolar-arterial oxygen gradient seems to increase (4 + age in years/4) ⁽⁹⁾. In contrast, pCO₂ seems to remain constant with age probably because of decreased body metabolism ⁽¹⁰⁾. However, the frequent multiple pathology in elderly patients could influence their arterial gases and acid-base status. Moreover, aging is associated with reduced compensation mechanisms in acute stress, so the typical picture of arterial BGA of the young patient with PE is not found in the elderly.

In the patients without pre-existing cardiopulmonary diseases enrolled in the PIOPED study (mean age ± SD, 54 ± 17 years), no significant differences were found in mean values of pO₂ and D(A-a)O₂ between subjects with and without pulmonary embolism: 70 ± 16 mm Hg versus 72 ± 18 mm Hg and 37 ± 17 mm Hg versus 35 ± 18 mm Hg, respectively ⁽¹¹⁾. Of these patients, 26% had pO₂ greater than 80 mm Hg. In contrast, in another study of patients with cardiopulmonary diseases (63.8 ± 14.59 years), there were significant differences in the mean values of pO₂ and pCO₂ between patients with CPE and patients with UCPE: 65.0 ± 13.0 mm Hg versus 68.0 ± 16.0 mm Hg for pO₂, and 32.0 ± 4.0 mm Hg versus 34.0 ± 5.0 mm Hg for pCO₂ ⁽¹²⁾. In this study, 10% of the patients with PE had pO₂ >82 mm Hg and pCO₂ >37 mm Hg. In another study, it was found that the combination of pO₂ >80 mm Hg, pCO₂ >35 mm Hg, and D(A-a)O₂ <20 mm Hg could not exclude the diagnosis of PE in more than 30% of cases without cardiopulmonary diseases and in more than 14% of cases with previous cardiopulmonary diseases ⁽¹³⁾. More recently, it was demonstrated that the association between pO₂ >80 mm Hg and negative D-dimer has a predictive value of 1 for the exclusion of PE ⁽¹⁴⁾.

In studies in which elderly patients were considered, although there were lower mean values of pO₂ and higher mean values of D(A-a)O₂ with respect to younger patients with PE and with respect to the patients of the above-mentioned studies, no significant differences were found in patients with or without PE ^{(15) (16)}. In particular, in the study of Stein and colleagues ⁽¹⁵⁾, mean values of pO₂ of 61 ± 12 mm Hg and D(A-a)O₂ of 47 ± 14 mm Hg were found in patients over 70 years of age participating in the PIOPED study. The same authors ⁽¹⁷⁾ found that D(A-a)O₂ increases with age in patients with PE, both in subjects with or without pre-existing cardiopulmonary diseases and in patients with or without prior episodes of PE or deep vein thrombosis. It was also found that pO₂ and D(A-a)O₂ have an inverse correlation in patients with PE with or without prior cardiopulmonary diseases. Moreover, D(A-a)O₂ showed a direct correlation with the severity of PE.

In the study by Jones and colleagues ⁽¹⁶⁾ of patients 74 ± 6 years old without pre-existing cardiopulmonary diseases, the mean values of pO₂, pCO₂, and D(A-a)O₂ in subjects with confirmed PE were not significantly different from those of subjects without PE: 61.4 ± 12.2 mm Hg versus 59.9 ± 14.5 mm Hg, 33.8 ± 6.2 mm Hg versus 36.6 ± 9.6 mm Hg, and 46.6 ± 11.6 mm Hg versus 46.0 ± 14.9 mm Hg, respectively. In 5.6% of patients with PE, D(A-a)O₂ was normal. The specificity of an abnormal D(A-a)O₂ in PE was 8.7%, and the sensitivity was 94%.

In our study of older patients, we found lower mean values of pO₂ in patients with PE but similar mean values of D(A-a)O₂ with respect to previous studies. There were significant differences in pO₂, SO₂, and D(A-a)O₂ between the CPE and UCPE groups, although 79% of the UCPE group had pO₂ <75 mm Hg, and 74.5% of this group had abnormal values of D(A-a)O₂. We also found that more than 50% of the CPE group had pCO₂ values >38 mm Hg, in disagreement with previous studies in which typical hypocapnia was found. Moreover, in the CPE group, apparently normal values of D(A-a)O₂ were determined by high levels of pCO₂ (>60 mm Hg). A possible explanation of these results is that many elderly patients suffer from multiple pathology with basal levels of pCO₂ higher than 42 mm Hg. Another possible explanation is the poor response to acute stress in the elderly. The low percentage of respiratory alkalosis and the high percentage of acute respiratory failure with hypercapnia (26% CPE and 25.5% UCPE) found in our series reflect these observations.

We also found that BGA parameters and D(A-a)O₂ are not useful for short-term prognosis (except SO₂), although lower mean values of pO₂ and pH and higher values of pCO₂ were found in nonsurviving patients. However, metabolic acidosis and metabolic alkalosis were associated with higher mortality.

Conclusion
Hypoxemia and increased alveolar-arterial oxygen gradient have a high sensitivity and low specificity in the diagnosis of PE in old age. Respiratory alkalosis seems to be less frequent in elderly patients with PE than in younger patients, while many older patients have metabolic or respiratory acidosis associated with higher short-term mortality.

PE remains a serious medical problem in the elderly; in a previous study we demonstrated the difficulty to reach a quick diagnosis for nonspecificity of clinical, instrumental, and laboratory aspects in this group of patients ⁽¹⁸⁾. Although BGA is a quick and noninvasive examination for assessment of gas exchange and acid-base status, multiple pathology could influence its results for diagnosis of PE in older patients.

Received March 8, 2000

Accepted March 17, 2000

	References

Top Abstract Methods Results Discussion References

 

1. Palla A, Petruzzelli S, Donnamaria V, Giuntini C, 1995. The role of suspicion in the diagnosis of pulmonary embolism. Chest 107: (suppl) 21S-24S. [Abstract/Free Full Text]
2. Raffin TA, 1986. Indications for arterial blood gas analysis. Ann Intern Med 105:390-398.
3. ACCP Consensus Committee on Pulmonary Embolism1996. Opinion regarding the diagnosis and management of venous thromboembolic disease. Chest 109:233-237. [Free Full Text]
4. The PIOPED Investigators1990. Value of the ventilation/perfusion scan in acute pulmonary embolism. Results of the Prospective Investigation of Pulmonary Embolism Diagnosis (PIOPED). JAMA 263:2753-2759. [Abstract/Free Full Text]
5. Pesarin F. Permutation Testing of Multidimensional Hypotheses by Nonparametric Combination of Dependent Tests. Padova, Italy: CLEUP; 1999.
6. Tapson VF, Carroll BA, Davidson BL, et al. 1999. The diagnostic approach to acute venous thromboembolism. Clinical practice guideline. Am J Respir Crit Care Med 160:1043-1066. [Free Full Text]
7. Sorbini CA, Grassi V, Solinas E, et al. 1968. Arterial oxygen tension in relation to age in healthy subjects. Respiration 25:3-13. [Medline]
8. Cerveri MC, Zoia F, Fanfulla L, et al. 1995. Reference values of arterial oxygen tension in the middle-aged and elderly. Am J Respir Crit Care Med 152:934-941. [Abstract]
9. Padamadan WD, 1984. Respiratory diseases and A-a gradient measurement. JAMA 252:1344
10. Rossi A, Ganassini A, Tantucci C, Grassi V, 1996. Aging and the respiratory system. Aging Clin Exp Res 8:143-161.
11. Stein PD, Terrin ML, Hales CA, et al. 1991. Clinical, laboratory, roentgenographic, and electrocardiographic findings in patients with acute pulmonary embolism and no pre-existing cardiac or pulmonary disease. Chest 100:598-603. [Abstract/Free Full Text]
12. Miniati M, Prediletto R, Formichi B, et al. 1999. Accuracy of clinical assessment in the diagnosis of pulmonary embolism. Am J Respir Crit Care Med 159:864-871. [Abstract/Free Full Text]
13. Stein PD, Goldhaber SZ, Henry JW, Miller AC, 1996. Arterial blood gas analysis in the assessment of suspected acute pulmonary embolism. Chest 109:78-81. [Abstract/Free Full Text]
14. Egermayer P, Town GI, Turner JG, Heaton DC, Mee AL, Beard ME, 1998. Usefulness of D-dimer, blood gas, and respiratory rate measurements for excluding pulmonary embolism. Thorax 53:830-834. [Abstract/Free Full Text]
15. Stein PD, Gottschalk A, Saltzman HA, Terrin ML, 1991. Diagnosis of acute pulmonary embolism in the elderly. JACC 18:1452-1457. [Abstract]
16. Jones JS, VanDeleen N, White L, Dougherty J, 1993. Alveolar-arterial oxygen gradients in elderly patients with suspected pulmonary embolism. Ann Emerg Med 22:1177-1181. [Medline]
17. Stein PD, Goldhaber SZ, Henry JW, 1995. Aveolar-arterial oxygen gradient in the assessment of acute pulmonary embolism. Chest 107:139-143. [Abstract/Free Full Text]
18. Masotti L, Ceccarelli E, Cappelli R, Guerrini M, Forconi S, 2000. Pulmonary embolism in the elderly: clinical, instrumental and laboratory aspects. Gerontology 46:205-211. [Medline]

This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation