GET THE APP

Albumin, pH, PaCO2 and Alveolar-Arterial Gradient Difference Can Predict 30Day Mortality in COPD Exacerbation

Journal of Research in Medical and Dental Science
eISSN No. 2347-2367 pISSN No. 2347-2545

All submissions of the EM system will be redirected to Online Manuscript Submission System. Authors are requested to submit articles directly to Online Manuscript Submission System of respective journal.

Research - (2021) Volume 9, Issue 4

Albumin, pH, PaCO2 and Alveolar-Arterial Gradient Difference Can Predict 30Day Mortality in COPD Exacerbation

Sema Avci1 and Gokhan Perincek2

Author info »

Abstract

Introduction: Early treatment, prognosis and short-term mortality prediction of COPD exacerbation are substantial for physicians in the ED. To compare the A-a O2 gradient (together with expected A-a O2 gradient, A-a O2 gradient difference), inflammatory markers (CRP, CAR, procalcitonin, NLR, PLR), and arterial blood gas (pH, pCO2, lactate clearance) in predicting COPD exacerbation patients’ 30-day mortality.

Methods: A questionnaire was designed for each patient, including detailed demographic profile, smoking status, comorbidities, vital signs, laboratory results and outcomes (discharge, hospitalization, ICU, 30-day mortality).

Results: 135 (60.3%) of the cases were male and 89 (39.7%) were female. Their ages ranged from 48 to 95 years, with an average of 72.22 ± 10.09. The LOS of cases varied between 1 and 28 days, with an average of 7.00 ± 3.83, ICU admission rate was 8% and the overall mortality rate was 5.4%. Albumin (AUC 0.81) and pH (AUC 0.68) showed the highest 30-day mortality prediction. While A-a O2 gradient difference (AUC 0.68) showed the highest 30-day mortality prediction, expected A-a O2 gradient (AUC 0.53) indicated a statistically lower 30-day mortality prediction. While PaCO2 (AUC 0.71) showed the highest 30-day mortality prediction, lactate clearance (AUC 0.54) indicated a statistically lower mortality estimation.

Conclusion: Albumin is a strong predictor of 30-day mortality in COPD exacerbation patients in the ED. In addition, an arterial blood gas sampling measurement including pH, PaCO2 and A-a O2 gradient difference are simple, precise and practical measurements for estimating 30-day mortality in these patients.

Keywords

Albumin, pH, PaCO2, Alveolar-arterial gradient, Chronic obstructive pulmonary disease

Introduction

Chronic obstructive pulmonary disease (COPD) is an inflammatory progressive condition which causes serious morbidity and mortality, constitutes approximately 1.5 million emergency department (ED) visits and is the fourth cause of all deaths [1,2]. COPD exacerbation is defined as the acute worsening of respiratory symptoms including dyspnea, cough and purulent sputum, and periods of exacerbation are the most common causes of ED admissions, hospitalization and mortality in COPD patients.3 The increasing number of COPD exacerbations leads to decreased quality of life, increased COPDrelated mortality rates and high healthcare costs [2,3]. Obviously, early treatment, prognosis and short-term mortality prediction are essential for physicians in the EDs.

Several laboratory results including well-known inflammatory markers such as neutrophil-tolymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), c-reactive protein (CRP)/albumin ratio (CAR), procalcitonin and arterial blood gas analysis including (pH, PaCO2, lactate clearance) have been used to predict severity, outcomes and short-long term mortality of patients with COPD exacerbation [4-8]. In addition to these laboratory parameters, the alveolar-arterial oxygen (A-a O2) gradient which is a measure of difference between the alveolar and arterial concentration of oxygen, has also been used to evaluate the outcomes (mortality, length of stay, severity of the disease) of the patients with pneumonia and pulmonary embolism [9,10].

The aims of the study were to compare the A-a O2 gradient (together with expected A-a O2 gradient, A-a O2 gradient difference), inflammatory markers (CRP, CAR, procalcitonin, NLR, PLR) and arterial blood gas (pH, pCO2, lactate clearance) in predicting COPD exacerbation patients’ 30-day mortality in the ED.

Materials and Methods

Study design and population

This prospective cross-sectional study was conducted with the approval of Kafkas University Medical Faculty Ethics Commitee between January and April 2020. The study included 224 (89 female, 135 male) patients with COPD exacerbation who admitted to ED, stages I-IV, for all patients. The diagnosis of COPD exacerbation was in accordance with the criteria established by Global Initiative for Chronic Obstructive Lung Disease 2019 [11]. Patients with a diagnosis of COPD, who were admitted to ED due to exacerbation and hospitalized were included in the study. A questionnaire was designed for each patient, including detailed demographic profile, smoking status, comorbidities, vital signs, laboratory results and outcomes (discharge, hospitalization, intensive care unit, 30-day mortality).

The laboratory findings were analyzed within 3 hours after admission to ED including arterial blood gas analysis, serum electrolytes, liver and kidney function tests, complete blood count, NLR, PLR, CRP, CAR, and procalcitonin.

Arterial blood gas samples were drawn from the radial artery in all patients while they were breathing room air to prevent any intervention caused by the maintenance of supplementary oxygen. Atmospheric pressure (mmHg), partial oxygen pressure (PaO2, mmHg), fraction of inspired oxygen (FiO2, 21% for room air), partial carbon dioxide pressure (PaCO2, mmHg) and age (for expected A-a O2 gradient) were recorded for all patients and calculated with https://www.mdcalc.com/a-a-O2-gradient [12]. After calculation process, the A-a O2 gradient and expected A-a O2 gradient for age were obtained. A-a O2 difference was calculated as A-a O2 gradient-expected A-a O2 gradient. Daily atmospheric pressure (mmHg) changes of the region where the study was carried out were obtained from the Turkish State Meteorological Service. Lactate clearance of the patients was calculated as ([initial lactate – second lactate (6 hr later)] /initial lactate) × 100 [13]. After one-month follow up, 30-day mortality after ED admission was evaluated.

Statistical reviews

All statistical calculations were performed with IBM SPSS Statistics for Windows, Version 23.0. Armonk, NY. The continuous variables were expressed as mean ± standard deviation; categoric variables were defined as percentages (%). The normal distribution was determined by histogram and Kolmogorov-Smirnov test. Mean values of continuous variables were compared between the groups using Mann- Whitney U test. Prediction accuracy was assessed using the area under the receiver operating characteristic (ROC) curve. The results were evaluated as 95% confidence interval and p value <0.05, which was considered statistically significant.

Results

Table 1 lists sociodemographic and clinical characteristic of the patients. Accordingly, 135 (60.3%) of the cases were male and 89 (39.7%) were female. Their ages ranged from 48 to 95 years, with an average of 72.22 ± 10.09. 104 (49.3%) of the cases were ex-smoker, 25 (11.8%) were current smoker and 81 (38.4%) were nonsmoker. The respiratory rate of the patients ranged between 13 and 29 per minute with an average of 19.99 ± 1.90. The systolic blood pressure of cases varied between 50 and 160 mmHg, with an average of 120.01 ± 11.64. The fever of the cases ranged from 36 to 37.2oC, with an average of 36.41 ± 0.23. The LOS of cases varied between 1 and 28 days, with an average of 7.00 ± 3.83, (intensive care unit) ICU admission rate was 8% and the overall mortality rate was 5.4%. Coexisting diseases were not detected 43 (19.2%) of the patients and anyone was living in nursing home resident. The frequency of comorbidities seen among the patients were: Hypertension (n=53), diabetes mellitus (n=31), congestive heart failure (n=19), asthma (n=8), hyperlipidemia (n=3), chronic liver disease (n=1), coronary artery disease (n=4), chronic renal disease (n=4).

n %
Gender Female 89 39.7
Male 135 60.3
Monthly income 0-150 USD 99 45.8
151-450 USD 102 47.2
451-750 USD 13 6
751-1100 USD 2 0.9
Oxygen concentrator (+) 80 36.7
(-) 138 63.3
Home nebulizer (+) 122 56
(-) 96 44
BiPAP machine (+) 20 9.2
(-) 198 90.8
Education status Illiterate 113 52.3
Primary School 82 38
Middle School 9 4.2
High School 6 2.8
University 2 0.9
Literate 4 1.9
Marital Status Single 6 2.8
Married 148 68.5
Divorced 3 1.4
Death of partner 59 27.3
Social condition Living with family 197 93.4
Alone 12 5.7
Support of distant relative or neighbour 1 0.5
Support of civil society organization 1 0.5
Place of residence Village 137 64.3
County 18 8.5
City 58 27.2
Smoking status Ex-smoker 104 49.3
Current smoker 25 11.8
Never smoker 81 38.4
Passive smoker 1 0.5
COPD drugs Short-acting beta agonists 1 0.6
Inhaled corticosteroids 49 29.9
Combination inhalers 114 69.5
Result of hospitalization Discharge 215 96
Death 9 4
Hospitalization unit Respiratory medicine unit 206 92
ICU 18 8

Table 1: Sociodemographic properties and clinical findings of the cases.

Table 2 lists comparisons of vital signs and age with Mann Whitney U test between survivors and non-survivors. The mean age of the non-survivors was significantly higher than the survivors (z=-2.153; p=0.031). Oxygen saturation, respiratory rate and systolic blood pressure did not show significantly difference between two groups) (p>0.05).

  Groups Xsıra Σ sira U z p
Oxygen saturation Deaths 68 680 625 -1.412 0.158
Survivors 91.82 15610
Respiratory rate Deaths 117.14 1288.5 801.5 -1.192 0.233
Survivors 96.86 17821.5
Temperature Deaths 100.59 1106.5 1016.5 -0.066 0.948
Survivors 99.44 18594.5
Systolic blood pressure Deaths 109.09 1200 923 -0.599 0.549
Survivors 98.94 18501
Age Deaths 145.54 1746.5 767.5 -2.153 0.031*
Survivors 105.78 21473.5

Table 2: Comparison of oxygen saturation, respiration, fever, systolic blood pressure and age values by groups.

Table 3 lists analysis of blood parameters with Mann Whitney U test between survivors and nonsurvivors. Blood urea nitrogen (BUN) (p=0.009) and creatinine (p=0.028) were significantly lower and hemoglobine (p=0.048) was higher in non-survivors.

  Groups Xsıra Σ sira U z p
White blood cell Non-survivors 107.58 21516.5 983.5 -1.049 0.294
Survivors 88.46 1061.5
Monocyte Non-survivors 105.99 20986 1091 -0.475 0.635
Survivors 97.42 1169
Neutrophil Non-survivors 106.24 21036.5 1040.5 -0.722 0.471
Survivors 93.21 1118.5
Lymphocyte Non-survivors 108.5 21699.5 800.5 -1.936 0.053
Survivors 73.21 878.5
Eosinophil Non-survivors 104.68 20727.5 1026.5 -0.792 0.428
Survivors 118.96 1427.5
Basophil Non-survivors 106.59 21104.5 972.5 -1.072 0.284
Survivors 87.54 1050.5
Hemoglobine Non-survivors 108.54 21707.5 792.5 -1.975 0.048
Survivors 72.54 870.5
Hematocrit Non-survivors 108.04 21608 892 -1.492 0.136
Survivors 80.83 970
Platelet Non-survivors 107.51 21502.5 997.5 -0.981 0.327
Survivors 89.63 1075.5
Red cell distribution width Non-survivors 103.98 20691.5 791.5 -1.96 0.051
Survivors 139.54 1674.5
Glucose Non-survivors 106.54 21308.5 1191.5 -0.041 0.967
Survivors 105.79 1269.5
BUN Non-survivors 104.29 20963 662 -2.623 0.009
Survivors 152.33 1828
Creatinine Non-survivors 105.74 21464.5 758.5 -2.195 0.028
Survivors 146.29 1755.5
Uric acid Non-survivors 86.83 14413.5 552.5 -1.774 0.076
Survivors 116.25 1162.5
Phosphor Non-survivors 83.84 13331 611 -0.187 0.851
Survivors 87.13 697
Aspartate amino transferase Non-survivors 109.89 22307.5 834.5 -1.832 0.067
Survivors 76.04 912.5
Alanine aminotransaminase Non-survivors 108.51 21919 806 -1.533 0.125
Survivors 79.27 872
Protein Non-survivors 106.26 20826 832 -1.7 -0.089
Survivors 75.83 910
Calcium Non-survivors 106.42 21178.5 910.5 -0.94 0.347
Survivors 88.77 976.5
Sodium Non-survivors 105.38 21076 976 -1.09 0.276
Survivors 125.17 1502
Magnesium Non-survivors 85.49 13763.5 565.5 -0.581 0.561
Survivors 75.19 601.5
Bicarbonate Non-survivors 106.72 21451.5 1150.5 -0.268 0.789
Survivors 111.63 1339.5
Lactate (initial) Non-survivors 107.27 21453 1047 -0.743 0.458
Survivors 93.75 1125
Lactate (second) Non-survivors 106.64 21115 962 -1.107 0.268
Survivors 86.67 1040

Table 3: Comparison of blood parameters between survivors and non-survivors.

Table 4 and Figure 1 demonstrate the accuracy of pH, pCO2, NLR, PLR, CRP, albumin, CAR, procalcitonin, A-a O2 gradient, A-a O2 gradientexpected, A-a O2 gradient difference and lactate clearance in predicting 30-day mortality. Albumin (AUC 0.81, 95% CI: 0.69-0.92) and pH (AUC 0.68, 95% CI: 0.49-0.88) showed highest 30-day mortality prediction. pCO2 (AUC 0.25, 95% CI: 0.09-0.41) and A-a O2 gradient-expected (AUC 0.29, 95% CI: 0.12-0.46) indicated statistically lower 30-day mortality prediction.

  AUC SE 95% CI
pH 0.684 0.1 0.487 to 0.881
PaCO2 0.25 0.08 0.094 to 0.407
NLR 0.415 0.087 0.244 to 0.586
PLR 0.36 0.089 0.186 to 0.535
CRP 0.506 0.084 0.342 to 0.670
Albumin 0.811 0.061 0.692 to 0.929
CAR 0.484 0.085 0.317 to 0.652
Procalcitonin 0.37 0.087 0.199 to 0.541
 A-a O2 gradient 0.364 0.109 0.149 to 0.578
A-a O2 gradient-expected 0.288 0.086 0.120 to 0.456
A-a O2 gradient difference 0.456 0.105 0.250 to 0.662
Lactate clearance 0.425 0.099 0.232 to 0.619

Table 4: Investigation of measurements effective in estimating 30-day mortality.

advances

Figure 1. The ROC curves for prediction of 30-day mortality for pH, PaCO2, NLR, PLR, CRP, albumin, CAR, procalcitonin, A-a O2 gradient, A-a O2 gradient-expected, A-a O2 gradient difference and lactate clearance.

Table 5 and Figure 2 demonstrate the accuracy of A-a O2 gradient, expected A-a O2 and A-a O2 gradient difference in predicting 30-day mortality. While A-a O2 gradient difference (AUC 0.68, 95% CI: 0.62-0.75) showed highest 30-day mortality prediction, expected A-a O2 gradient (AUC 0.53, 95% CI: 0.46-0.60) indicated statistically lower 30-day mortality prediction.

  AUC SE 95% CI
A-a O2 gradient 0.598 0.102 0.528 to 0.664
A-a O2 gradient-expected 0.53 0.0956 0.460 to 0.598
A-a O2 gradient difference 0.684 0.0775 0.617 to 0.746

Table 5: Investigation of measurements effective in estimating 30-day mortality.

advances

Figure 2. The ROC curves for prediction of 30-day mortality for A-a O2 gradient, A-a O2 gradient-expected and A-a O2 gradient difference.

Table 6 and Figure 3 demonstrate the accuracy of pH, PaCO2 and lactate clearance in predicting 30- day mortality. While PaCO2 (AUC 0.71, 95% CI: 0.64-0.77) showed the highest 30-day mortality prediction, lactate clearance (AUC 0.54, 95% CI: 0.47-0.61) indicated statistically lower mortality estimation.

  AUC SE 95 % CI
pH 0.658 0.102 0.590 to 0.721
PaCO2 0.71 0.0874 0.644 to 0.770
Lactate clearance 0.54 0.0802 0.470 to 0.608

Table 6: Investigation of measurements effective in estimating 30-day mortality.

advances

Figure 3. The ROC curves for prediction of 30-day mortality for pH, PaCO2 and lactate clearance.

Table 7 and Figure 4 demonstrate the accuracy of procalcitonine, CRP, CAR, albumin and NLR in predicting 30-day mortality. While albumin (AUC 0.81, 95% CI: 0.73-0.87) showed highest 30-day mortality prediction, CRP (AUC 0.51, 95% CI: 0.42-0.59) and CAR (AUC 0.52, 95% CI: 0.43-0.60) indicated statistically lower mortality estimation.

  AUC SE 95% CI
Procalcitonin 0.63 0.0924 0.543 to 0.712
CRP 0.506 0.0876 0.418 to 0.593
CAR 0.516 0.0895 0.428 to 0.603
Albumin 0.811 0.0632 0.734 to 0.873
NLR 0.585 0.0916 0.497 to 0.670

Table 7: Investigation of measurements effective in estimating 30-day mortality.

advances

Figure 4. The ROC curves for prediction of 30-day mortality for procalcitonin, CRP, CAR, albumin and NLR.

Discussion

Mortality prediction in patients with COPD exacerbation in the ED is important for patient management. The study aimed to compare arterial blood gas (pH, PaCO2, lactate clearance), A-a O2 gradient and various inflammatory markers (NLR, PLR, CRP, CAR, procalcitonin). In the study, the 30-day mortality rate and ICU admission rate were 5.4% and 8%, respectively. The mean age of the non-survivors was higher than the survivors. The long and short term mortality rates of COPD exacerbation are quite different and the portion of this study is lower compared to these studies [14-17]. Although non-survivors in our study were elderly and the number of comorbidities was high, the lower mortality rate may be due to early admission to the ED and the high rate of hospitalization.

In the study, lower BUN, creatinine and higher hemoglobine were related to 30-day mortality among COPD exacerbation patients. Dehydration due to high fever, acute inflammatory reaction, decreased oral intake in the elderly population, smoking, increased number of comorbidities, hypoperfusion caused by hypoxia and hypercapnia may lead to impaired renal functions in COPD exacerbation patients [18,19]. Moreover, the impairment of renal functions is a factor which increases the severity of COPD and acute renal failure may increase mortality and hospital admissions.20 Serum creatinine level provides indirect information about the total muscle mass in the body [20]. Furthermore, acute serum creatine elevation is an indicator of muscle breakdown, and this increase is not seen patients with body muscle atrophy or wasting such as severe COPD [21,22]. In addition, BUN increases with the catabolism of body mass, low BUN level indicates cessation of catabolism [21]. Low creatinine and BUN levels in non-survivors may be result from severe COPD stage and muscle atrophy. Hemoglobin abnormalities such as anemia and polycythemia are common in patients with COPD and hypoxiainduced erythropoiesis leads to secondary polycythemia in these patients [23]. So, cor pulmonale and pulmonary hypertension are based on polycythemia in COPD [24]. In addition, high hemoglobin level concentration increases mortality rate and leads to poor outcomes as it predisposes to hypertension, stroke, thrombosis and cardiovascular events [23,24].

In this study, albumin (0.81) and pH (0.68) were the strongest predictor of 30-day mortality in COPD exacerbation patients among pH, pCO2, NLR, PLR, CRP, albumin, CAR, procalcitonin, A-a O2 gradient, A-a O2 gradient-expected, A-a O2 gradient difference and lactate clearance. Albumin (0.81) was also superior in predicting mortality compared to the other inflammatory markers. The mortality predictive power of PaCO2 (0.71) was higher than pH and lactate clearance. A-a O2 gradient difference (AUC 0.68) was better than expected A-a O2 gradient and A-a O2 gradient in predicting mortality. Besides albumin, pH, PaCO2 and A-a O2 gradient difference can also be used mortality prediction for COPD exacerbation in the ED. Trauma, critical conditions such as sepsis, organ failure, chronic inflammatory diseases cause increased vascular permeability and lower serum albumin level [25,26]. Moreover, albumin is an acute plasma protein responsible for microvascular permeability, acid-bas equilibrium and prevention of platelet aggregation. So, decreased serum level of albumin is associated with poor outcomes including morbidity, mortality and ICU admission in particularly critically ill patients [26-29]. COPD is a chronic and inflammatory disease and the worsening of the disease severity during exacerbation may explain the role of albumin as a predictor of mortality. Additionally, albumin is an inflammatory marker which can be obtained more easily and practically than arterial blood gas in the ED. Although the blood gas sampling procedure is difficult, pH, PaCO2 and A-a O2 gradient difference are substantial in terms of showing respiratory acidosis, compensation status and deep hypoxia [30]. The effects of pH and PaCO2 in predicting mortality, hospital admission rate, long-term changes, and non-invasive ventilation duration in COPD patients have been shown in several studies [31- 34]. In addition, we could not find any study in the literature on the relationship between A-a O2 gradient difference and COPD exacerbation.

This single-center study had some limitations. To begin with, the patient population was comparatively small and we did not evaluate a coexisting pneumonia for arterial blood gas analysis including A-a O2 gradient, expected A-a O2 gradient and A-a O2 gradient difference. A detailed effect of comordid diseases to mortality was unknown. 30-day follow-up period after ED admission, merely mortality was evaluated and the treatment protocol was not recorded. Furthermore, local hematoma, aneurysm, air or thrombus embolism, infection, laceration, hemorrhage, needle stick injuries and pain may ocur after arterial blood gas sampling [35,36].

Conclusion

Albumin is strong predictor of 30-day mortality in COPD exacerbation patients in the ED. In addition, an arterial blood gas sampling measurement including pH, PaCO2 and A-a O2 gradient difference are simple, precise and practical measurements for estimating 30-day mortality in these patients.

Acknowledgements

We thank to Murat Yasar Capan for extraction of data.

Author Contributions

S.A. and G.P. helped in the concept analysis, data collection, design, literature search, interpretation of data, preparation of initial and final draft. All the authors read and approved the final draft.

Funding

None declared.

Availability of Data and Materials

The authors agree to the conditions of publication including the availability of data and materials in our manuscript.

Conflict of Interest

None declared.

Informed Consent

Informed consent was obtained from the participants or their legally authorized representatives.

Ethical Approval

This study was approved by the local ethics committee of Kafkas University Medical Faculty.

Human Rights

The principles outlined in the Declaration of Helsinki have been followed.

References

Author Info

Sema Avci1 and Gokhan Perincek2

1Department of Emergency Medicine, Usak University Medical Faculty, Usak, Turkey
2Department of Pulmonology, Kars Harakani State Hospital, Kars, Turkey
 

Citation: Didem Atabek, Nagehan Akta?, Didem Sakaryal? Uyar, Two Years Clinical Evaluation of Sonic-Resin Placement System with Self-Etch and Total-Etch Adhesive Modes, J Res Med Dent Sci, 2021, 9 (4):84-92.

Received: 15-Feb-2021 Accepted: 26-Mar-2021

cappadocia tours
cappadocia hotels
cappadocia balloon flights

paper.io

agar io

wormax io