Pulse Oximetry Overestimates Oxygen Saturation in COPD

Discussion

The mean age (63 y) distribution of the subjects reflects the progressive respiratory insufficiency in COPD, which debilitates patients late in its natural history. The preponderance of male population could be attributed to the habit of smoking and also to significant exposure to pollutants outside of the house, both of these characteristics being more common to men. The studied subjects have largely been laborers in whom the exposure to air pollution might be contributory to their respiratory ailments.

Considering the agreement limits (−4.68 to 3.28), the handheld pulse oximeter performs poorly in acute respiratory failure in COPD as a whole. However, it can be used to aid in pointing out the trend of rise or fall in blood oxygen saturation.

The Bland-Altman plot (Fig. 1) shows some degree of difference between S_aO2 and S_pO2. With the mean difference (bias) being −3.98 (95% CI −4.68 to 3.28), the pulse oximeter consistently overestimates oxygen saturation. In COPD, this error may severely affect the treatment decisions in borderline cases.

In patients with COPD, the right-ventricular end-diastolic pressures are increased. The interventricular septum is pushed toward the left ventricle and compromises its function.¹¹ The arterial pulse dependent on left-ventricular contraction is thus altered, and faulty values may be seen in pulse oximetry. The amount of oxygen delivered to the patient with COPD has to be finely tuned, since both hypoxia and hyperoxgenation are known to increase the death rates in patients with COPD.¹² Hence, respiratory therapy based solely on pulse oximetry is ill-advised in patients with COPD.

Therefore, pulse oximetry should be interpreted in the light of other clinical features of respiratory failure pending ABG analysis. Different studies show a wide range (−0.02 to 3.39) of bias between the oxygen saturation assessed by pulse oximetry and arterial blood gas analysis (Table 4).^{2–4,13–16}

This may in part be due to the different clinical situations in which the studies were performed and the fact that the pulse oximeter interprets the ratio of oxyhemoglobin to total hemoglobin in the pulsatile arterial vessels.¹⁷ Therefore, factors altering the detection of pulsatility of the arteries limit the accuracy of the S_pO2 values. It is known to be unreliable in low perfusion states, dyshemoglobinemia, hypothermia, carbon monoxide, icterus, and anemia.¹⁸ It appears to be a reliable tool in conditions such as cyanotic congenital heart disease,¹⁵ where there is less direct dysregulation of the peripheral vasculature than in conditions like liver transplantation³ and systemic sclerosis.¹⁴

These observations should be kept in mind by the clinician when interpreting pulse oximetry in an emergency setting. In short, the disease itself has a bearing on the accuracy of the instrument.

When the data were analyzed under subclasses of chronic bronchitis and emphysema, the Bland-Altman analysis (Table 2) showed higher variation of agreement limits in subjects with chronic bronchitis than in emphysema. The sensitivity of the pulse oximeter to respiratory failure in subjects with chronic bronchitis (81%) (Table 3) was less than that noted in those with emphysema (85%). Adding to this the low specificity (75%) of pulse oximetry in chronic bronchitis makes it doubly inefficient. Its readings in chronic bronchitis may fail to show hypoxemia when it is present (low sensitivity), and they may show erroneously low values in the absence of hypoxia (low specificity). This may be due to the fact that hemodynamic instability is generally greater in patients with chronic bronchitis than in those with emphysema, which may cause fluctuations in recording S_pO2. The distinct pathophysiologies¹⁹ of these 2 disorders contribute to the differential findings.

In chronic bronchitis, cough and sputum production usually precede by many years the onset of clinically important airway obstruction. During the relentless progression of the disease, airway obstruction becomes irreversible due to recurrent respiratory tract infections, which lead to exacerbations of the obstruction and cause increased disability. When the disease is advanced, respiratory tract infections precipitate acute respiratory failure. As the patient progresses into chronic respiratory failure, right-sided heart failure (cor pulmonale) also ensues. In contrast, emphysema is characterized by the loss of elastic recoil of the lungs, which probably precedes the onset of symptoms by many years. Presenting initially with dyspnea on exertion, it increases inexorably without the fluctuations seen in chronic bronchitis. In the terminal stages, the patient is grossly dyspneic at rest and usually succumbs to intercurrent respiratory tract infections. Due to this difference in disease progression, central cyanosis and heart failure are features more common in chronic bronchitis than in emphysema. These features significantly affect the accuracy of pulse oximeter readings.

One of the shortcomings of this study is the lack of consideration of the effects of carboxyhemoglobin, which might affect the analysis because many smokers have higher levels of carboxyhemoglobin. The study is also limited by its small sample size, but it suggests that: (1) the handheld pulse oximeter on its own was suspect in assessing respiratory failure in subjects with COPD; (2) it may help in monitoring trends in patients with hypoxemia established by ABG analysis; (3) it is not conclusive in diagnosing respiratory failure but is an adjunct to the ABG analysis; (4) the variability of the readings was greater in subjects with chronic bronchitis than in those with emphysema due to hemodynamic instabilities, which are more frequent with the former subjects.