Research ArticleOriginal Research

Hypoxemia During Extreme Hyperleukocytosis: How Spurious?

Andry Van de Louw, Ruchi J Desai, Coursen W Schneider and David F Claxton
Respiratory Care January 2016, 61 (1) 8-14; DOI: https://doi.org/10.4187/respcare.04196

Abstract

BACKGROUND: Spurious hypoxemia has been described in case reports during extreme hyperleukocytosis and has led to recommendations for immediate cooling and analysis of arterial blood gases (ABGs). We sought to determine, in samples processed as recommended, the magnitude of spurious hypoxemia in acute leukemia subjects with hyperleukocytosis.

METHODS: A retrospective chart review was conducted of all subjects admitted between 2003 and July 2014 for acute leukemia, who presented with white blood cell (WBC) count > 50 × 109 cells/L and had ABGs performed. For each ABG, we collected PaO2, SaO2, simultaneous WBC count, and SpO2 when available. Bland and Altman analysis was used to assess the agreement between SpO2 and SaO2.

RESULTS: One-hundred forty-six samples (from 45 subjects) were included, of which 57 samples (from 18 subjects) had data available for Bland and Altman analysis. Mean (SpO2 − SaO2) was 2.5%, and 95% CI for limits of agreement between SpO2 and SaO2 was (−10.1,15.1)%. The mean (SpO2 – SaO2) was significantly higher for WBC count > 100 × 109/L as compared with WBC count < 100 × 109/L (3.8% vs 0.4%, P = .04), and the 95% CIs for limits of agreement were (−10.3,18)% versus (−7.9,8.6)%. SpO2 and SaO2 were poorly correlated (r2 = 0.19), whereas the difference (SpO2 – SaO2) was fairly correlated with WBC count (r2 = 0.44). Overall, 11 of 19 samples with WBC count > 150 × 109/L had PaO2 < 55 mm Hg whereas SpO2 was > 94%, the proportion being 5 of 62 samples for WBC count < 150 × 109/L (P < .001). Three subjects with WBC count > 150 × 109/L exhibited large SpO2 to SaO2 differences (10–20%) before leukapheresis, which decreased to below 5% afterward.

CONCLUSIONS: In subjects with acute leukemia and hyperleukocytosis, despite cooling and quickly analyzing the samples, we observed poor correlation and agreement between SpO2 and SaO2, unacceptably low for WBC count > 100 × 109/L. Our results suggest that current guidelines may not totally prevent the diagnosis of spurious hypoxemia.

Introduction

Between 10 and 30% of patients newly diagnosed with acute leukemia present with hyperleukocytosis, usually defined as a white blood cell (WBC) count > 100 × 109/L.1 Hyperleukocytosis not only negatively impacts overall survival but is also associated with increased early mortality, due to cerebral and pulmonary leukostasis.2 Patients requiring ICU admission and mechanical ventilation for acute respiratory failure have an overall poor survival3,4; thus, the respiratory status of acute leukemia patients with hyperleukocytosis is a major concern.

In acute leukemia, early detection of pulmonary compromise may prompt invasive intervention aimed at decreasing WBC count (leukapheresis) or improving gas exchange (mechanical ventilation). As clinical signs and chest x-ray findings may be subtle in immunocompromised hosts, the assessment of patient‘s oxygenation via arterial blood gas (ABG) measurement and/or pulse oximetry is crucial. The finding of artifactual decrease in PaO2 due to increased oxygen consumption by WBCs in collected samples, has been described and referred to as spurious hypoxemia.5 A clinically relevant decay in PaO2 may occur as soon as 2 min in samples with WBC count as low as 55 × 109/L and seems proportional to WBC count.5 Several case reports have been published,6–9 but other authors have emphasized that the hypoxemia observed during hyperleukocytosis might actually be real and reflect subtle pulmonary leukostasis.10 Published data come from small series, but no study has ever systematically investigated the reliability of PaO2 and SaO2 in this setting, probably because of the relative rarity of extreme hyperleukocytosis. Although the American Association for Respiratory Care (AARC) has recommended immediate cooling and analysis of ABG for hyperleukocytic patients,11 these guidelines rely on limited data, and their effect on the magnitude of spurious hypoxemia has never been evaluated.

Given the limited published data, we included a large series of acute leukemia subjects with hyperleukocytosis in this retrospective study and systematically assessed (1) the relationship between PaO2, SaO2, and WBC count and (2) the agreement between SaO2 and SpO2 in cooled and quickly analyzed ABG.

QUICK LOOK

Current knowledge

Patients newly diagnosed with acute leukemia may present with hyperleukocytosis, usually defined as a white blood cell (WBC) count > 100 × 109/L. The finding of artifactual decrease in PaO2, due to increased oxygen consumption by WBCs in collected samples, has been described in this population and referred to as spurious hypoxemia. A clinically relevant decay in PaO2 may occur as soon as 2 min in samples with WBCs as low as 55 × 109/L and seems proportional to WBC count.

What this paper contributes to our knowledge

In a retrospective review of acute leukemia subjects with hyperleukocytosis, despite cooling and quickly analyzing the samples, there was poor correlation and agreement between SpO2 and SaO2, unacceptably low for WBC count > 100 × 109/L. These results suggest that current guidelines may not always prevent the diagnosis of spurious hypoxemia.

Methods

This retrospective study was approved by the Pennsylvania State University College of Medicine Institutional Review Board (IRB number 00000374), and informed consent was waived.

Subjects

The charts of all subjects admitted to our institution with a newly diagnosed acute leukemia between January 2003 and July 2014 were screened, and all subjects with at least one ABG performed with a concomitant hyperleukocytosis (WBC count > 50 × 109/L) were included for further analysis.

Data Collected

Baseline demographic data, significant comorbidities, type of acute leukemia, the use of invasive or noninvasive mechanical ventilation during admission, and 28-d and 6-month survival were collected. Whenever an ABG and a complete blood count were performed simultaneously, we collected PaO2, SaO2, and WBC count, along with SpO2 recorded at the same time, when available. Mean arterial pressure, heart rate, and vasopressor requirements (yes/no) at the time of ABG were also recorded. As per hospital policy, ABGs were collected in plastic syringes, stored in ice, and immediately sent (via pneumatic tube system) and processed in the laboratory, in agreement with published recommendations.12 The type of oxygen or ventilatory support was recorded as follows: no oxygen requirement, oxygen on nasal cannula, high-flow oxygen device (high-flow nasal cannula, Venturi mask), noninvasive mechanical ventilation, invasive mechanical ventilation.

Statistical Analysis

All results were reported as mean ± SD (SPSS 20, SPSS, Chicago, Illinois). To assess the agreement between SaO2 and SpO2, we used the method described by Bland and Altman,13 calculating the mean difference (bias, d) and the SD of the differences (precision, s) between SpO2 and SaO2 and the 95% CIs for limits of agreement between SpO2 and SaO2 (d ± 2s). We used the 2-tailed Student t test to compare the bias, mean arterial pressure, and heart rate between samples with a WBC count below versus above 100 × 109/L. A 2-tailed Fisher exact test was used to compare the proportion of cases with hypoxemia (PaO2 < 55 mm Hg) and vasopressor requirements between samples with WBC below or above 100 × 109/L, respectively. Analysis of variance for repeated measures was used to assess the effect of leukapheresis on (SpO2 – SaO2) differences. P < .05 was considered statistically significant.

Results

Subjects

Of 874 patients admitted during the 11½ year period with a newly diagnosed acute leukemia, 45 subjects were included in this analysis. In total, these subjects had 146 simultaneous complete blood counts and ABGs performed. All subjects had acute myeloid leukemia, and 6 had the acute promyelocytic leukemia subtype. None of the 63 patients with hyperleukocytic acute lymphoblastic leukemia had an ABG performed. Baseline demographic, clinical, and biological data as well as survival are detailed in Table 1. ABGs were performed while subjects were breathing room air (n = 3), receiving oxygen on nasal cannula (n = 23) or high-flow device (n = 23), or supported by noninvasive (n = 4) or invasive (n = 93) mechanical ventilation.

View this table:
Table 1.

Demographic, Clinical, Biological, and Survival Data for the 45 Subjects

Agreement and Correlation Between SaO2 and SpO2

Among the 146 ABGs performed in 45 subjects, we were able to collect simultaneous SaO2 and SpO2 and performed Bland and Altman analyses for 57 samples in 18 subjects. The mean (SpO2 – SaO2) was 2.5%, and the 95% CI for limits of agreement between SpO2 and SaO2 were (−10.1,15.1)% (Fig. 1). Plotting SpO2 versus SaO2 (Fig. 2), we observed a poor correlation with a coefficient of determination (r2) of 0.19.

Fig. 1.
Fig. 1.

Bland and Altman representation of the agreement between SpO2 and SaO2. For the 57 samples (18 subjects), (SpO2 – SaO2) was plotted on the Y axis versus (SpO2 + SaO2)/2 on the X axis. The mean (SpO2 – SaO2) (bias) was 2.5% (center line), and the 95% CI for limits of agreement was (−10.1,15.1)% (dotted lines).

Fig. 2.
Fig. 2.

Correlation between SpO2 and SaO2 for 57 samples (18 subjects). The correlation equation is SpO2 = 0.243SaO2 + 73.3 (r2 = 0.19).

Effect of WBC Count on PaO2 and on the Agreement Between SpO2 and SaO2

Figures 3 and 4 represent the distributions of PaO2 and (SpO2 – SaO2) versus WBC count, respectively. The coefficient of determination for the correlation between (SpO2 – SaO2) and WBC count was 0.44; thus, 44% of the (SpO2 – SaO2) variance was accounted for by variations in WBC count. Figure 5 displays the effect of WBC count on the agreement between SpO2 and SaO2. The mean (SpO2 – SaO2) was significantly higher for WBC count > 100 × 109/L as compared with WBC count < 100 × 109/L (3.8% vs 0.4%, P = .04). Likewise, the 95% CIs for limits of agreement were (−10.3,18.0)% for WBC count > 100 × 109/L versus (−7.9,8.6)% for WBC count < 100 × 109/L. There was no difference between the 2 groups in terms of vasopressor requirements (11 of 35 samples with WBC count > 100 × 109/L vs 11 of 22 samples with WBC count < 100 × 109/L, P = .28) or heart rate (96 ± 22 beats/min for samples with WBC count > 100 × 109/L vs 93 ± 20 beats/min for samples with WBC count < 100 × 109/L, P = .60). Mean arterial pressure at the time of ABG was higher for WBC count > 100 × 109/L as compared with WBC count < 100 × 109/L (90 ± 22 mm Hg vs 77 ± 14 mm Hg, P = .002). Including only the 10 samples with WBC count > 200 × 109/L, the bias and 95% CI for limits of agreement were 11.0% and (−2.4,24.5)%, respectively.

Fig. 3.
Fig. 3.

Distribution of PaO2 and white blood cell count for 146 samples (45 subjects).

Fig. 4.
Fig. 4.

Distribution of (SpO2 – SaO2) and white blood cell count for 57 samples (18 subjects). The correlation equation is SpO2 – SaO2 = 0.05 × white blood cell count – 4.0 (r2 = 0.44).

Fig. 5.
Fig. 5.

Bland and Altman representation of the agreement between SpO2 and SaO2 according to white blood cell (WBC) count. (SpO2 – SaO2) was plotted against (SpO2 + SaO2)/2 for the 22 samples with a WBC count < 100 × 109/L (A) and for the 35 samples with a WBC count > 100 × 109/L (B). The bias was 0.4% for white blood cell count < 100 × 109/L versus 3.8% for WBC count > 100 × 109/L (P = .04) (center lines). The 95% CIs for limits of agreement were (−7.9,8.6)% and (−10.3,18.0)%, respectively (dotted lines).

Among 44 samples with WBC count > 100 × 109/L, 18 (10 subjects) exhibited a PaO2 < 55 mm Hg, whereas SpO2 was > 88%, the proportion being only 4 of 36 samples for WBC count < 100 × 109/L (P = .003). Similarly, 11 of 19 samples with WBC count > 150 × 109/L had PaO2 < 55 mm Hg, whereas SpO2 was > 94%, the proportion being only 5 of 62 samples for WBC count < 150 × 109/L (P < .001). The lowest PaO2 observed was 22 mm Hg in a subject with a WBC count of 365 × 109/L, breathing O2 4 L/min via nasal cannula with a SpO2 of 96% and no clinical sign of respiratory distress.

Comparison of (SpO2 – SaO2) Differences Before/After Leukapheresis

Six subjects who underwent leukapheresis to quickly decrease their WBC count had ABGs performed before and after the procedure. (SpO2 – SaO2) was 9.4 ± 9.2% before leukapheresis versus 0.5 ± 3.8% afterward (P = .055). Three subjects with very high WBC count (> 150 × 109/L) all exhibited large SpO2 to SaO2 differences (10–20%) before leukapheresis, which decreased to < 5% after the procedure (Fig. 6).

Fig. 6.
Fig. 6.

Effect of leukapheresis on white blood cell count and (SpO2 – SaO2) in 6 subjects who had arterial blood gas analysis performed before (shapes on the right) and after (shapes on the left) leukapheresis. For the 3 subjects with large SpO2 to SaO2 differences (10–20%) before leukapheresis, (SpO2 – SaO2) decreased to below 5% after the procedure. The 6 different shapes represent the 6 individual subjects.

Discussion

The main findings of this study are the large bias and limits of agreement and the poor correlation between SaO2 and SpO2 in hyperleukocytic acute myeloid leukemia. This poor agreement seems directly related to the severity of hyperleukocytosis, as it worsened with increasing WBC count and disappeared after leukapheresis in some subjects. It is clinically relevant especially for very high WBC count (> 150 × 109/L), as 50% of ABGs suggested severe hypoxemia (PO2 < 55 mm Hg), whereas subjects were probably not hypoxic (SpO2 > 94%).

Spurious hypoxemia (or leukocyte larceny) was first described in 1979,5,14 and a few small series have been published since, mostly in subjects with chronic myeloid or lymphocytic leukemias6,7,9,14,15 and severe hyperleukocytosis, but no study thus far had investigated the magnitude and clinical relevance of this phenomenon. The pathophysiology of spurious hypoxemia has been thought to involve oxygen consumption by increased numbers of leukocytes inside the syringes (leukocyte larceny) before analysis.5 This hypothesis was supported by an in vitro study using continuous PO2 measurement in blood samples from 5 hyperleukocytic leukemia subjects and also from granulocyte donors, where authors observed a progressive decrease in PO2 in hyperleukocytic samples, reaching values as low as 20 mm Hg within 10 min.16 This progressive decline was grossly proportional to WBC count and was not subsequently observed in the same subjects after cytoreductive therapy or in whole blood from the granulocyte donors.16 These results are also consistent with the decrease in PO2 expected when applying Henry's law to hyperleukocytic samples, using known data on in vitro oxygen consumption by leukocytes17; assuming, for instance, that the oxygen consumption of leukocytes is about 0.1 μl O2/106 cells/h,18 one would expect a drop in PO2 of about 50 mm Hg within 10 min for a WBC count of 100 × 109/L at ambient temperature. This PaO2 decay could be related not only to the WBC count but also to the type of cells, as immature leukocytes seem to have higher metabolic rates than normal cells.19

Although scattered case reports or small series of spurious hypoxemia have been published,5–8 our study is the first to systematically investigate the magnitude of spurious hypoxemia in a large population of acute leukemia subjects with hyperleukocytosis. Other authors have challenged the magnitude of spurious hypoxemia and suggested that part of the observed hypoxemia might actually be real and related to either unrecognized leukostasis or increased methemoglobinemia.10 Indeed, increased methemoglobinemia described during acute leukemia20 is associated with false SpO2 readings and was also proposed to account for low SaO2 in hyperleukocytic subjects.10 However, methemoglobinemia is unlikely to explain spurious hypoxemia, as it is associated with normal PaO221 and could not account for the extremely low PaO2 observed in the present study and others.5,9,22,23

Overall, we described a poor correlation between SaO2 and SpO2 in the presence of a WBC count > 50 × 109/L, as attested by the low coefficient of determination (r2 = 0.19) observed when plotting SpO2 versus SaO2. Although correlation analysis may not be the best approach to compare SaO2 and SpO2, a large meta-analysis of 74 studies reported an r2 of 0.8 between SaO2 and SpO2,24 much higher than the r2 of 0.19 that we observed for subjects with WBC count > 50 × 109/L. The bias and 95% CI for limits of agreement in our whole population (2.5% and (−10.1,15.1)%) were also larger than those reported by most studies in non-hyperleukocytic subjects24–26 but were probably negatively affected by samples with extreme WBC count. Indeed, the bias and 95% CI for limits of agreement that we observed in subjects with WBC count < 100 × 109/L (0.4% and (−7.9,8.6)%) are similar to those reported in studies in critically ill27 and COPD26 populations. Conversely, the bias and 95% CI for limits of agreement were 3.8% and (−10.3,18.0)% for WBC count > 100 × 109/L, whereas one of the largest studies on 102 general ICU subjects reported a bias of only 0.02% with 95% CI for limits of agreement of (−4.2,4.2)%.25 Although peripheral vasoconstriction can cause false SpO2 readings,28 the large (SaO2 − SpO2) differences for WBC count > 100 × 109/L are unlikely to be explained by a worse hemodynamic status associated with hyperleukocytosis, as we found no difference in heart rate or vasopressor requirements between the 2 groups, and if anything, mean arterial pressure was higher for those with WBC count > 100 × 109/L. As expected, bias and 95% CI for limits of agreement were even poorer for WBC count > 200 × 109/L (11.0% and (−2.4,24.5)%, respectively). The relationship between (SpO2 − SaO2) and WBC count was reinforced by the correlation observed when plotting the difference (SpO2 − SaO2) versus WBC count (r2 = 0.44).

Following the publication of several case reports6–8 and of in vitro studies,5 recent clinical practice guidelines from the AARC recommend immediate cooling and analysis of ABGs for patients with very high leukocyte counts.11 Although our ABGs were processed according to these recommendations (immediately cooled and analyzed without delay), we still observed clinically important spurious hypoxemia. Some authors have described a blunted PaO2 decay in samples placed on ice,5,10 but others have argued that immediate cooling is not sufficient to eliminate spurious hypoxemia, because the oxygen consumption by leukocytes continues during the time (as short as a few minutes) required by samples to gradually reach ice water temperature.23 As in the present study, Loke et al22 observed a significant PaO2 decay even in samples stored on ice and analyzed immediately. The addition of potassium cyanide appears to be the best way to stop the PaO2 decay,5 but it is more difficult to implement in daily practice. Rapid processing of the samples alone is certainly not sufficient to prevent spurious hypoxemia, as elegantly shown by Fox et al5; performing in vitro studies of blood samples using a tonometer, they observed a drop in PO2 from 130 to 55 mm Hg within 2 min only for a WBC count of 276 × 109/L at ambient temperature. Our results reinforce the AARC recommendations11 but suggest that they may not be sufficient to prevent spurious hypoxemia.

Spurious hypoxemia may have deleterious consequences if not adequately recognized. First, PaO2 may be used as a criterion to initiate mechanical ventilation in patients with hematological malignancies,29 along with clinical signs of respiratory distress. Coexisting conditions (pain, anxiety, cerebral leukostasis) may also cause tachypnea, reinforcing the pivotal role of PaO2 in the clinical decision making. Second, during mechanical ventilation, unrecognized spurious hypoxemia may trigger inappropriate increase in FIO2 and expose patients who are already prone to respiratory complications to additional pulmonary oxygen toxicity.30 Third, PaO2 is part of severity scores widely used in critically ill patients (APACHE [Acute Physiology and Chronic Health Evaluation] and SOFA [Sequential Organ Failure Assessment]),31 and low recorded PaO2 may falsely increase those scores. Finally, initial PaO2 has been shown to be an independent predictor of mortality in immunocompromised populations with pulmonary infiltrates.32 In summary, assessment of oxygenation is crucial in acute leukemia patients but is particularly challenging in the presence of extreme hyperleukocytosis.

Interestingly, most if not all cases of spurious hypoxemia published involved myeloid leukemias, acute or chronic.5–8,10 In agreement with the literature, all of our subjects had acute myeloid leukemia. The most probable explanations for the lineage specificity of this observation are: (1) acute myeloid leukemia is the most frequent acute leukemia in adults33 and (2) symptomatic pulmonary leukostasis is more frequent in acute myeloid leukemia subjects,34 who are therefore more likely to have ABGs performed. Indeed, among 874 patients admitted for acute leukemia in the present study, only 63 had acute lymphoblastic leukemia, and none of them had ABGs performed during their admission.

The main limitation of this study is its retrospective design and the inherent concern about the reliability of the data collected. Pulse oximetry is subject to plethysmography artifacts,28 and we were not able to account for some factors affecting its reliability (methemoglobinemia); thus, SpO2 recorded in the charts could be false. We tried to minimize this bias by ensuring that consistent SpO2 values were recorded around the time of ABG, but we cannot rule out, despite all of our precautions, the possibility of false SpO2 readings. However, extreme hyperleukocytosis in acute leukemia is so infrequent that a study prospectively recording SpO2 and SaO2 would be difficult to perform. Although our hospital policy requires all ABGs to be stored on ice and immediately processed, we could not control for the exact time elapsed between ABG collection and analysis. Because of the irreducible time required for transport to the laboratory, receipt, labeling, and accessioning of the samples, we estimate this delay at about 10–15 min. As discussed above, however, even a few minutes may be enough to cause PaO2 decay, and the only way to prevent that delay would be to use portable gas analyzers at bedside. Another limitation is the lack of a control group with normal WBC count; however, the bias and 95% CI for limits of agreement that we observed for WBC count between 50 and 100 × 109/L were similar to those reported in general ICU subjects,25,26 so that the poor agreement observed for WBC count > 100 × 109/L can probably be ascribed to hyperleukocytosis. Our choice to include subjects with WBC count > 50 × 109/L is also arguable, but it was based on a study reporting a rapid decay in PaO2 in blood samples with a WBC count of 55 × 109/L.5 Although these results were obtained by in vitro studies using tonometry in a few samples, our data actually suggest that spurious hypoxemia is clinically unacceptable for WBC count > 100 × 109/L. Finally, oxygen supplementation may have altered the relationship between PaO2 and WBC count, as presented in Figure 3. We were not able to present data with uniform FIO2 (due to the small number of subjects for a given FIO2) or PaO2/FIO2 ratios, as accurate computation of FIO2 is challenging for patients receiving oxygen via nasal cannula or face mask. Notwithstanding this limitation, the relationship between PaO2 and WBC count as depicted in Figure 3 remains convincing. Indeed, one would expect the subjects with higher WBC count to have presented more severe respiratory symptoms and to have received more oxygen, which could have artificially increased their PaO2 as compared with subjects with lower WBC count. The fact that we observed, on the contrary, extremely low PaO2 in subjects with higher WBC count suggests that supplemental oxygen did not distort the overall relationship between PaO2 and WBC count and, if anything, strengthens our results.

Conclusions

In acute leukemia subjects with hyperleukocytosis, even cooled and quickly analyzed ABGs exhibited unacceptably poor agreement and correlation between SpO2 and SaO2 for WBC count > 100 × 109/L. This spurious hypoxemia seems grossly proportional to WBC count and may lead to significant misinterpretation of the subject's oxygenation status. The clinical practice guidelines for immediate cooling and analysis of ABG for very high WBC count are reinforced but may not be sufficient to prevent spurious hypoxemia.

Footnotes

  • Correspondence: Andry Van de Louw MD PhD, Division of Pulmonary and Critical Care Medicine, Milton S Hershey Medical Center, 500 University Drive, Hershey, PA 17033. E-mail: avandelouw{at}hmc.psu.edu.

  • The authors have disclosed no conflicts of interest.

  • See the Related Editorial on Page 119

REFERENCES

  1. 1.↵
    1. Porcu P,
    2. Cripe LD,
    3. Ng EW,
    4. Bhatia S,
    5. Danielson CM,
    6. Orazi A,
    7. McCarthy LJ
    . Hyperleukocytic leukemias and leukostasis: a review of pathophysiology, clinical presentation and management. Leuk Lymphoma 2000;39(1):1–18.
    OpenUrlCrossRefPubMed
  2. 2.↵
    1. Dutcher JP,
    2. Schiffer CA,
    3. Wiernik PH
    . Hyperleukocytosis in adult acute nonlymphocytic leukemia: impact on remission rate and duration, and survival. J Clin Oncol 1987;5(9):1364–1372.
    OpenUrlAbstract/FREE Full Text
  3. 3.↵
    1. Chaoui D,
    2. Legrand O,
    3. Roche N,
    4. Cornet M,
    5. Lefebvre A,
    6. Peffault de Latour R,
    7. et al
    . Incidence and prognostic value of respiratory events in acute leukemia. Leukemia 2004;18(4):670–675.
    OpenUrlCrossRefPubMed
  4. 4.↵
    1. Roze des Ordons AL,
    2. Chan K,
    3. Mirza I,
    4. Townsend DR,
    5. Bagshaw SM
    . Clinical characteristics and outcomes of patients with acute myelogenous leukemia admitted to intensive care: a case-control study. BMC Cancer 2010;10:516.
    OpenUrlPubMed
  5. 5.↵
    1. Fox MJ,
    2. Brody JS,
    3. Weintraub LR
    . Leukocyte larceny: a cause of spurious hypoxemia. Am J Med 1979;67(5):742–746.
    OpenUrlCrossRefPubMed
  6. 6.↵
    1. Lele AV,
    2. Mirski MA,
    3. Stevens RD
    . Spurious hypoxemia. Crit Care Med 2005;33(8):1854–1856.
    OpenUrlCrossRefPubMed
  7. 7.↵
    1. Mizock BA,
    2. Franklin C,
    3. Lindesmith P,
    4. Shah PC
    . Confirmation of spurious hypoxemia using continuous blood gas analysis in a patient with chronic myelogenous leukemia. Leuk Res 1995;19(12):1001–1004.
    OpenUrlCrossRefPubMed
  8. 8.↵
    1. Sacchetti A,
    2. Grynn J,
    3. Pope A,
    4. Vasso S
    . Leukocyte larceny: spurious hypoxemia confirmed with pulse oximetry. J Emerg Med 1990;8(5):567–569.
    OpenUrlCrossRefPubMed
  9. 9.↵
    1. Haynes JM
    . A case of disparity between pulse oximetry measurements and blood gas analysis values. Respir Care 2004;49(9):1059–1060.
    OpenUrlFREE Full Text
  10. 10.↵
    1. Gartrell K,
    2. Rosenstrauch W
    . Hypoxaemia in patients with hyperleukocytosis: true or spurious, and clinical implications. Leuk Res 1993;17(11):915–919.
    OpenUrlPubMed
  11. 11.↵
    1. Davis MD,
    2. Walsh BK,
    3. Sittig SE,
    4. Restrepo RD
    . AARC clinical practice guideline: blood gas analysis and hemoximetry: 2013. Respir Care 2013;58(10):1694–1703.
    OpenUrlAbstract/FREE Full Text
  12. 12.↵
    1. Knowles TP,
    2. Mullin RA,
    3. Hunter JA,
    4. Douce FH
    . Effects of syringe material, sample storage time, and temperature on blood gases and oxygen saturation in arterialized human blood samples. Respir Care 2006;51(7):732–736.
    OpenUrlAbstract/FREE Full Text
  13. 13.↵
    1. Bland JM,
    2. Altman DG
    . Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1(8476):307–310.
    OpenUrlCrossRefPubMed
  14. 14.↵
    1. Hess CE,
    2. Nichols AB,
    3. Hunt WB,
    4. Suratt PM
    . Pseudohypoxemia secondary to leukemia and thrombocytosis. N Engl J Med 1979;301(7):361–363.
    OpenUrlCrossRefPubMed
  15. 15.↵
    1. Wen YF,
    2. Kuo PH
    . Unusual discrepancy between SpO2 and SaO2 in a 31-year-old man with chronic myelogenous leukemia. Respir Care 2012;57(11):1977–1979.
    OpenUrlFREE Full Text
  16. 16.↵
    1. Chillar RK,
    2. Belman MJ,
    3. Farbstein M
    . Explanation for apparent hypoxemia associated with extreme leukocytosis: leukocytic oxygen consumption. Blood 1980;55(6):922–924.
    OpenUrlAbstract/FREE Full Text
  17. 17.↵
    1. Malawista SE,
    2. Bodel PT
    . The dissociation by colchicine of phagocytosis from increased oxygen consumption in human leukocytes. J Clin Invest 1967;46(5):786–796.
    OpenUrlPubMed
  18. 18.↵
    1. Holmes B,
    2. Page AR,
    3. Good RA
    . Studies of the metabolic activity of leukocytes from patients with a genetic abnormality of phagocytic function. J Clin Invest 1967;46(9):1422–1432.
    OpenUrlCrossRefPubMed
  19. 19.↵
    1. Kempner W
    . The nature of leukemic blood cells as determined by their metabolism. J Clin Invest 1939;18(3):291–300.
    OpenUrlCrossRefPubMed
  20. 20.↵
    1. Vaidya MS,
    2. Das Gupta A,
    3. Pavri RS,
    4. Baxi AJ,
    5. Advani SH
    . Acquired methemoglobinemia in leukemia: its etiopathogenesis and possible effects on red cell structure and function. Leuk Res 1981;5(3):265–269.
    OpenUrlPubMed
  21. 21.↵
    1. Rausch-Madison S,
    2. Mohsenifar Z
    . Methodologic problems encountered with cooximetry in methemoglobinemia. Am J Med Sci 1997;314(3):203–206.
    OpenUrlCrossRefPubMed
  22. 22.↵
    1. Loke J,
    2. Duffy TP
    . Normal arterial oxygen saturation with the ear oximeter in patients with leukemia and leukocytosis. Cancer 1984;53(8):1767–1769.
    OpenUrlCrossRefPubMed
  23. 23.↵
    1. Chillar RK,
    2. Belman MJ,
    3. Farbstein M
    . Pseudohypoxemia due to leukemia and thrombocytosis. N Engl J Med 1980;302(10):584.
    OpenUrlCrossRefPubMed
  24. 24.↵
    1. Jensen LA,
    2. Onyskiw JE,
    3. Prasad NG
    . Meta-analysis of arterial oxygen saturation monitoring by pulse oximetry in adults. Heart Lung 1998;27(6):387–408.
    OpenUrlCrossRefPubMed
  25. 25.↵
    1. Van de Louw A,
    2. Cracco C,
    3. Cerf C,
    4. Harf A,
    5. Duvaldestin P,
    6. Lemaire F,
    7. Brochard L
    . Accuracy of pulse oximetry in the intensive care unit. Intensive Care Med 2001;27(10):1606–1613.
    OpenUrlCrossRefPubMed
  26. 26.↵
    1. Hannhart B,
    2. Michalski H,
    3. Delorme N,
    4. Chapparo G,
    5. Polu JM
    . Reliability of six pulse oximeters in chronic obstructive pulmonary disease. Chest 1991;99(4):842–846.
    OpenUrlCrossRefPubMed
  27. 27.↵
    1. Wilson BJ,
    2. Cowan HJ,
    3. Lord JA,
    4. Zuege DJ,
    5. Zygun DA
    . The accuracy of pulse oximetry in emergency department patients with severe sepsis and septic shock: a retrospective cohort study. BMC Emerg Med 2010;10:9.
    OpenUrlCrossRefPubMed
  28. 28.↵
    1. Jubran A
    . Pulse oximetry. Crit Care 1999;3(2):R11–R17.
    OpenUrlCrossRefPubMed
  29. 29.↵
    1. Conti G,
    2. Marino P,
    3. Cogliati A,
    4. Dell'Utri D,
    5. Lappa A,
    6. Rosa G,
    7. Gasparetto A
    . Noninvasive ventilation for the treatment of acute respiratory failure in patients with hematologic malignancies: a pilot study. Intensive Care Med 1998;24(12):1283–1288.
    OpenUrlCrossRefPubMed
  30. 30.↵
    1. Jackson RM
    . Pulmonary oxygen toxicity. Chest 1985;88(6):900–905.
    OpenUrlCrossRefPubMed
  31. 31.↵
    1. Vincent JL,
    2. Moreno R
    . Clinical review: scoring systems in the critically ill. Crit Care 2010;14(2):207.
    OpenUrlCrossRefPubMed
  32. 32.↵
    1. Poe RH,
    2. Wahl GW,
    3. Qazi R,
    4. Kallay MC,
    5. Utell MJ,
    6. Morrow GR
    . Predictors of mortality in the immunocompromised patient with pulmonary infiltrates. Arch Intern Med 1986;146(7):1304–1308.
    OpenUrlCrossRefPubMed
  33. 33.↵
    1. Deschler B,
    2. Lübbert M
    . Acute myeloid leukemia: epidemiology and etiology. Cancer 2006;107(9):2099–2107.
    OpenUrlCrossRefPubMed
  34. 34.↵
    1. Bunin NJ,
    2. Pui CH
    . Differing complications of hyperleukocytosis in children with acute lymphoblastic or acute nonlymphoblastic leukemia. J Clin Oncol 1985;3(12):1590–1595.
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

 

Print
Download PDF
Article Alerts
Email Article
Citation Tools

Jump to section

Related Articles

Cited By...

Keywords