Annals of Thoracic Medicine
: 2015  |  Volume : 10  |  Issue : 2  |  Page : 94--99

The association between glucose levels and hospital outcomes in patients with acute exacerbations of chronic obstructive pulmonary disease

Ebtesam A Islam, Chok Limsuwat, Teerapat Nantsupawat, Gilbert G Berdine, Kenneth M Nugent 
 Department of Internal Medicine, Texas Tech University Health Sciences Center, Lubbock, Texas, USA

Correspondence Address:
Kenneth M Nugent
3601 4th Street, Lubbock, Texas - 79430


BACKGROUND: Corticosteroids used for chronic obstructive pulmonary disease (COPD) exacerbations can cause hyperglycemia in hospitalized patients, and hyperglycemia may be associated with increased mortality, length of stay (LOS), and re-admissions in these patients. MATERIALS AND METHODS: We did three retrospective studies using charts from July 2008 through June 2009, January 2006 through December 2010, and October 2010 through March 2011. We collected demographic and clinical information, laboratory results, radiographic results, and information on LOS, mortality, and re-admission. RESULTS: Glucose levels did not predict outcomes in any of the studied cohorts, after adjustment for covariates in multivariable analysis. The first database included 30 patients admitted to non-intensive care unit (ICU) hospital beds. Six of 20 non-diabetic patients had peak glucoses above 200 mg/dl. Nine of the ten diabetic patients had peak glucoses above 200 mg/dl. The maximum daily corticosteroid dose had no apparent effect on the glucose levels. The second database included 217 patients admitted to ICUs. The initial blood glucose was higher in patients who died than those who survived using bivariate analysis (P = 0.015; odds ratio, OR, 1.01) but not in multivariable analysis. Multivariable logistic regression analysis also demonstrated that glucose levels did not affect LOS. The third database analyzing COPD re-admission rates included 81 patients; the peak glucose levels were not associated with re-admission. CONCLUSIONS: Our data demonstrate that COPD patients treated with corticosteroids developed significant hyperglycemia, but the increase in blood glucose levels did not correlate with the maximum dose of corticosteroids. Blood glucose levels were not associated with mortality, LOS, or re-admission rates.

How to cite this article:
Islam EA, Limsuwat C, Nantsupawat T, Berdine GG, Nugent KM. The association between glucose levels and hospital outcomes in patients with acute exacerbations of chronic obstructive pulmonary disease.Ann Thorac Med 2015;10:94-99

How to cite this URL:
Islam EA, Limsuwat C, Nantsupawat T, Berdine GG, Nugent KM. The association between glucose levels and hospital outcomes in patients with acute exacerbations of chronic obstructive pulmonary disease. Ann Thorac Med [serial online] 2015 [cited 2021 Dec 6 ];10:94-99
Available from:

Full Text

Chronic obstructive pulmonary disease (COPD) has significant adverse effects on quality of life and life expectancy. According to the World Health Organization (WHO) website, COPD was the fifth leading cause of death in 2002, total deaths will increase by more than 30% over the next 10 years, and by 2030, COPD will be the third leading cause of death worldwide. [1] Patients with COPD often have acute exacerbations (AECOPD) which require hospitalization, and these hospitalized patients have higher mortality rates and costs than patients managed as outpatients. Hospital mortality has been associated with some clinical parameters, such as a low Glasgow Coma Scale score, cardiorespiratory arrest prior to admission, and higher acute physiology scores. [2] A cochrane database systematic review found that patients given corticosteroids had significantly fewer treatment failures within 30 days. Hospital stays were shorter, and lung function and dyspnea improved. However, adverse events, including hyperglycemia, were more frequent. [3] Hyperglycemia has been associated with adverse clinical outcomes in patients with AECOPD in several studies. [4] Baker reported that hyperglycemia was associated with an increased relative risk of death or longer inpatient stay and that a 1 mmol/L (18 mg/dl) increase in blood glucose increased the risk of adverse outcomes by 15%. [4] The authors also reported more frequent isolation of multiple pathogens and Staphylococcus aureus from sputum of patients with higher blood glucose levels. Hyperglycemia has been associated with poor outcomes in several medical disorders, including stroke, myocardial infarction, and pneumonia. [5],[6],[7] We have reviewed the associations between glucose levels and corticosteroid dosage, between initial glucose levels and length of stay (LOS) and mortality, and between peak glucose levels and re-admission in three separate cohorts of patients hospitalized for acute exacerbations of COPD. We thought this approach could help us understand the relationships between glucose levels and management strategies and key outcomes.


The information reported in this manuscript was collected during three separate projects in the same hospital. The first cohort included COPD patients admitted for an acute exacerbation to a non-ICU bed to determine if there was a relationship between corticosteroid dosage and hyperglycemia. The second cohort included COPD patients with acute exacerbations who were admitted to the MICU to determine factors associated with mortality and LOS. The third cohort included COPD patients who were monitored for re-admission within 30 days after discharge for an acute exacerbation. The Texas Tech University Health Sciences Center Lubbock/Odessa Institutional Review Board approved these studies.

Cohort # 1: Glucose levels and corticosteroid dosage

We did a retrospective study at University Medical Center in Lubbock, TX, using charts from July 1, 2008 through June 30, 2009. These charts were identified by the billing office in the Department of Internal Medicine. We included all adult patients with a discharge diagnosis of COPD exacerbation identified by the coders in the Department of Internal Medicine; we excluded patients younger than 18 years and those admitted to an ICU. Data collected included forced expiratory volume in 1 second (FEV1), history of diabetes, blood glucose at admission, peak glucose, and maximum daily dose of corticosteroid.

Cohort #2: Glucose levels and mortality and LOS

We retrospectively reviewed charts of patients at University Medical Center in Lubbock, TX, diagnosed with COPD exacerbations who needed ICU admission. The medical records department used ICD 9 codes 491.21 to identify patients admitted between January 1, 2006 and December 31, 2010. The inclusion criteria were age 45 years or older, a diagnosis with COPD exacerbation defined by at least two criteria (increased dyspnea, cough increased in frequency and severity, and sputum production increased in volume and/or changed character), and admission to the ICU. The exclusion criteria included a history of another respiratory disease, such as asthma, or a cardiac disorder resulting in congestive heart failure which caused the acute respiratory distress.

Data collected included the patients' ages, sex, body mass indices (BMI), baseline pulmonary function tests, comorbidities, complete blood counts (CBCs), complete metabolic profiles, albumin levels, initial arterial blood gases, sputum cultures, blood cultures, chest X-rays, Acute Physiology and Chronic Health Evaluation II (APACHE II) scores, confusion, urea, respiratory rate, systolic blood pressure (CURB65) scores, final diagnoses, total duration of hospital stay, and in-hospital mortality. We separated patients into two groups: Group One died during hospitalization and Group Two survived to hospital discharge. The primary outcome in this study was the identification factors associated with increased in-hospital mortality; the secondary outcome was the identification of the factors associated with increased LOS. We separated patients by LOS into four quartiles to identify parameters associated with increased LOS. [8],[9]

Cohort # 3: Peak glucose levels and re-admission

We retrospectively reviewed COPD patients who were admitted to University Medical Center in Lubbock, TX, between October 1, 2010 and March 31, 2011 and then re-admitted within 30 days. The medical record numbers were available from the University Medical Center (UMC) case management office which was collecting information on these Centers for Medicare and Medicaid Services performance measure. From electronic medical records and case management resources, we collected the highest blood glucose level during the hospital stay for each patient and other clinical parameters which might be associated with re-admissions. [10]

Data analysis

We used descriptive statistics to summarize base line characteristics of patients in all three cohorts. T-tests and Chi square tests were used to analyze differences between patients who died and patients who survived the acute flare. Multivariable logistic regression models were used to determine factors that predicted increased mortality and LOS. T-tests were used to compare highest blood glucose levels in the early re-hospitalization group with the non-early re-hospitalization group. We then used logistic regression analysis to analyze which factors affected the early re-hospitalization rate. Statistical analysis was performed using Statistical Package for the Social Sciences (SPSS) version 16.0; P - values <0.05 were consider statistically significant.


The basic clinical characteristics for the three cohorts are summarized in [Table 1].{Table 1}

Cohort #1: Glucose levels and corticosteroid dosage in non-ICU patients

Twenty non-diabetic patients had a mean initial glucose of 122 ± 33 mg/dl with a range of 83 to 238 mg/dl [Table 2]. Six patients had maximum glucose levels above 200 mg/dl. Ten patients with diabetes (DM) had mean initial glucose levels of 196 ± 110 mg/dl with a range of 77 to 446 mg/dl, and nine had maximum glucose levels above 200 mg/dl. We stratified the patients into tertiles based on the maximum daily dose of corticosteroids received during hospitalization [Table 3]. There were no differences in the initial or maximum glucose levels in these three groups (P > 0.05 by analysis of variance, ANOVA).{Table 2}{Table 3}

Cohort #2: Glucose levels and mortality in AECOPD patients requiring ICU admission

We identified 325 patients admitted between January 1, 2005 and January 1, 2011 to intensive care units with the ICD 9 code for AECOPD. However, after chart reviews 108 cases were excluded from the study (5 deaths, 103 survivors) because they did not meet inclusion criteria or met exclusion criteria. The majority of excluded patients had congestive heart failure with exacerbation or acute asthma and did not have an AECOPD as the primary diagnosis. Therefore, this study enrolled 217 cases with AECOPD who needed ICU admission; 26 died during hospitalization and 191 survived to hospital discharge. Eighty-nine percent of the patients who died in the hospital had a FEV1 less than 50% of predicted; 68% of the survivors had a FEV1 less than 50 % of predicted. The mean initial blood glucose on admission in this cohort was 156 ± 74.3 mg/dl (range 60 to 485 mg/dL). Univariate logistic regression analysis identified a statistically significant association between initial glucose and mortality (OR = 1.01; 95%CI: 1.001-1.01, P = 0.015). Multivariable logistic regression including statistically significant mortality risk factors from univariate logistic regression showed that patients who lived in a nursing home (OR 50.02; 95% CI: 2.7-923.19), had a low hemoglobin (OR 0.40; 95% CI: 0.17-0.94), were intubated (OR 27.54; 95%1.78-425.39), and had a high APACHE II score (OR 1.55; 95% CI: 1.07-2.23) were at increased risk for mortality, but the initial glucose levels were not associated with increased risk [Table 4]. [8]{Table 4}

Cohort #2: Initial glucose levels and LOS in AECOPD patients requiring ICU admission

The overall LOS was 9.01 ± 6.00 days. It was 8.37 ± 4.99 days in survivors and 13.69 ± 9.78 days in non-survivors. Based on LOS, we separated patients into quartiles: The first quartile LOS: 1-4 days (n = 44), the second quartile LOS: 5-6 days (n = 53), the third quartile LOS: 6-10 days (n = 58), and the fourth quartile LOS: 11-33 days (n = 62). Initial mean blood glucose was 156 ± 74.3 mg/dl and did not predict a statistically significant increased LOS in univariate logistic regression analysis (OR 1.00; 95% CI: 0.99-1.01, P = 0.59). Univariate analysis did demonstrate that nursing home status, low albumin, the presence of a pleural effusion, intubation, and high APACHE II scores were associated with increased LOS (P < 0.05 for each factor). A multivariable logistic regression model demonstrated that only intubation (OR 5.93; 95%CI 2.78-12.65) predicted a statistically significant increase in hospital LOS [Table 5]. [9]{Table 5}

Cohort # 3: Peak glucose levels and re-admission in patients with AECOPD

These patients had 103 admissions, including 13 early re-hospitalizations (12.6 %), defined as a re-hospitalization within 30 days after discharge. The mean peak blood glucose levels were 212 ± 105mg/dl with a range from 94mg/dl to 600 mg/dl. There were no differences in mean peak glucose levels between early re-hospitalization and non-early re-hospitalization patients (P = 0.318). The characteristics of these two groups are reported in [Table 6]. Multivariable logistic regression analysis showed that only the presence of unilateral pulmonary infiltration on admission (OR 5.14 [0.86-30.86], P = 0.074) and an ejection fraction < 55% (OR 4.57 (0.86-24.31), P = 0.075) possibly had an effect on early re-hospitalizations. [10]{Table 6}


Our study includes information from three separate cohorts with a total of 318 COPD patients with acute exacerbations requiring hospitalization. The initial glucose levels had a wide range in the first cohort, and the mean glucose level was nearly 200 mg/dl in the patients with diabetes. Ninety percent of the patients with diabetes had a maximum glucose level during hospitalization above 200 mg/dL. There was no apparent correlation between the maximum corticosteroid dose and the maximum glucose level in this small group of patients. The other two cohorts included patients admitted to the ICU and patients tracked for re-admission to the hospital. In the ICU cohort the initial glucose level did not predict length of stay or mortality using a multivariable model. In the cohort tracked for re-admission, the peak glucose level did not predict re-admission when used in a multivariable model. Consequently, our results demonstrate that patients with acute COPD exacerbations frequently have hyperglycemia, especially if they have diabetes, but the elevated glucose levels do not predict short-term outcomes. This does not mean that hyperglycemia is not consequential in some patients or when present over longer periods of time or that it does not require attention during patient management.

Glucose levels in patients with AECOPD

Baker et al., measured glucose levels in 284 patients with acute exacerbations of COPD requiring hospitalization and reported the highest value measured either at admission or during the admission. [4] The median value in the entire cohort was 126 mg/dl (interquartile range (IQR): 108-162 mg/dl). Two hundred and four patients (72%) had values greater than 110 mg/dl, and 32(11%) had values greater than 200 mg/dl. The highest quartile had glucose levels greater than 162 mg/dl. Fifteen patients (5.3%) in this cohort had diabetes (DM). Parappil reviewed 246 episodes of AECOPD. [11] Forty-two percent of the admissions had a random glucose greater than 180 mg/dl, but mean values were not provided. Fifty-three patients (22%) had DM. Burt measured glucose values in 47 patients using a continuous monitoring system; the mean level was 137 ± 34 mg/dl. [12] The peak levels in their cohort were 220 ± 61 mg/dl. Seven patients (15%) had DM. Chakrabarti studied 88 patients with decompensated COPD (mean pH = 7.25) who required non-invasive ventilation. [13] Fifty percent of the patients had an admission glucose >126 mg/dl; 16 (18%) had DM. The initial glucose levels in our Cohort 2 patients were 151 mg/dl in survivors and 192 mg/dl in non-survivors; 62 patients (29%) had DM. The mean peak glucose levels ranged from 184 mg/dl in non-diabetic patients in Cohort 1 to 344 mg/dl in diabetic patients in Cohort 1. Consequently, these studies in patients with AECOPD requiring hospitalization demonstrate that peak glucose levels are variable but often high, and that 5-29% of patients with AECOPD has DM. Patients with COPD frequently have comorbid diseases, including DM, and COPD is a risk factor for the development of DM. [14],[15] This association may reflect the adverse effects of medication, physical inactivity with weight gain, and possibly intermittent stress associated with acute exacerbations.

Outcomes in AECOPD patients with hyperglycemia

Several studies have evaluated the relationship between in-hospital glucose levels and AECOPD. Baker found significantly higher mortality rates and longer LOS in341 patients with acute exacerbations of COPD and increasing blood glucose levels measured during the hospitalization, independent of age, sex, and previous diagnosis of DM. [4] The absolute risk of mortality and length of hospital stay increased by 15% for each 1 mmol/L (18mg/dl) increase in blood glucose. The authors also reported more frequent isolation of multiple pathogens and Staphylococcus aureus from sputum of patients with increases in blood glucose. Burt reported that the mean glucose level based on continuous measurement independently predicted an increased LOS. [12] Parappil et al., reported a non-significant increase in LOS and mortality in 39 patients with DM admitted with acute exacerbations of COPD. [11] However, Kasirye reported a longer LOS in patients with low glucose levels (1 daily mean value < 90 mg/dl). [16] Chakrabarti reported that hyperglycemia was associated with an increased likelihood of failure of noninvasive ventilation in 88 patients; the explanation for this association is unclear. [13] Dave reviewed outcomes in 498 patients requiring 731 admissions for acute hypercapneic respiratory failure, including 109 patients who required noninvasive ventilation. [17] Admission glucose levels were similar in patients managed on the ward and patients managed in a high dependency unit or ICU. Our studies did not identify a relationship between LOS, mortality, or re-admission. Overall, these studies report an inconsistent effect of hyperglycemia on the outcomes in patients hospitalized with COPD exacerbations. This could be explained, in part, by the glucose measurement protocols which included admission values, mean daily values, peak values, and mean continuous values. Glucose levels have a circadian pattern with higher values in the late afternoon in corticosteroid treated patients with AECOPD, and routine morning lab tests may not detect the range of glucose elevations. [18],[19] In addition, glycemic variability is associated with increased length of stay and mortality in hospitalized patients and needs to be considered in these studies. [20]

Possible mechanisms for hyperglycemic effects

Patients with AECOPD frequently have hyperglycemia which could have acute adverse effects in critically ill patients through several mechanisms. High intracellular glucose concentrations increase the production of reactive oxygen species, such as superoxide, and of peroxynitrite which can impair mitochondrial function and damage proteins, cellular membranes, and nucleic acids. [21],[22] Hyperglycemia causes glycosylation of proteins, impaired leukocyte function, and activation of pro-inflammatory genes through transcription factors. [21],[22],[23] Glycosuria potentially causes an osmotic diuresis with a loss of electrolytes and volume contraction which could adversely affect the clinical course. Occasionally, patients on corticosteroids develop hyperglycemic hyperosmolar syndrome or diabetic ketoacidosis. Other studies have reported that hyperglycemia increases the mortality in patients with acute myocardial infarction, with ischemic or hemorrhagic stroke, and with community-acquired pneumonia. [5],[6],[7],[24]

Strengths and limitations

Our study includes a relatively large number of patients with AECOPD requiring hospitalization and considers three important outcomes. However, it does have some limitations. Medical records were identified by discharge codes, and some patients with COPD exacerbations could have been misdiagnosed with acute respiratory distress secondary to pulmonary embolus, CHF, or asthma. Some information was not consistently available in the records. Different physicians managed the patients, and different management approaches would influence the choice of medication, the dose of medication, and the frequency and type of laboratory testing. The small number of patients in our first cohort might have limited our ability to identify a statistically significant association between corticosteroid dose and hyperglycemia, but there was a very wide range of corticosteroid doses in these patients. Finally, we don't know if the development of hyperglycemia introduces practical problems in the management of the patients both in the hospital and after discharge.

In summary, our study does not identify any important associations between glucose levels and outcomes of patients with AECOPD requiring hospitalization. This conclusion seems somewhat counterintuitive but could reflect the fact that in-hospital management of hyperglycemia is relatively easy and prevents complications. [25],[26] We need more information on the frequency and severity of hyperglycemia in these patients, the factors associated with hyperglycemia, the effects of hyperglycemia on short term outcomes, and the effect of intermittent hyperglycemia on long term outcomes.


1World health organization. Burden of COPD. Available from: [Last accessed on 2014 Jan 4].
2Messer B, Griffiths J, Baudouin SV. The prognostic variables predictive of mortality in patients with an exacerbation of COPD admitted to the ICU: An integrative review. QJM 2012;105:115-26.
3Walters JA, Gibson PG, Wood-Baker R, Hannay M, Walters EH. Systemic corticosteroids for acute exacerbations of chronic obstructive pulmonary disease. Cochrane Airways Group, Published Online: 21 JAN 2009. [Last accessed on 2014 Jan 4].
4Baker EH, Janaway CH, Philips BJ, Brennan AL, Baines DL, Wood DM, et al. Hyperglycaemia is associated with poor outcomes in patients admitted to hospital with acute exacerbations of chronic obstructive pulmonary disease. Thorax 2006;61:284-9.
5Capes SE, Hunt D, Malmberg K, Pathak P, Gerstein HC. Stress hyperglycemia and prognosis of stroke in nondiabetic and diabetic patients: A systematic overview. Stroke 2001;32:2426-32.
6Hadjadj S, Cosinet D, Mauco G, Ragot S, Duengler F, Sosner P, et al. Prognostic value of admission plasma glucose and HbA 1C in acute myocardial infarction. Diabet Med 2004;21:305-10.
7McAlister FA, Majumdar SR, Blitz S, Rowe BH, Romney J, Marrie TJ. The relation between hyperglycemia and outcomes in 2,471 patients admitted to the hospital with community-acquired pneumonia. Diabetes Care 2005;28:810-5.
8Limsuwat C, Nantsupawat N, Umyarova E, Ussavarungsi K, Nugent K. Factors affecting mortality in patients with COPD exacerbations requiring ICU admission. Southwest Respir Crit Care Chron 2013;1.
9Limsuwat C, Mankongpaisarnrung C, Dumrongmongcolgul N, Nugent K. Factors influencing the length of hospital stay in patients with acute exacerbations of chronic obstructive pulmonary disease admitted to intensive care units. Qual Manag Health Care 2014;23:86-93.
10Nantsupawat T, Limsuwat C, Nugent K. Factors affecting chronic obstructive pulmonary disease early rehospitalization. Chron Respir Dis 2012;9:93-8.
11Parappil A, Depxynski B, Collett P, Marks GB. Effect of comorbid diabetes on length of stay and risk of death in patients admitted with acute exacerbations of COPD. Respirology 2010;15:918-22.
12Burt MG, Roberts GW, Aguilar-Loza NR, Quinn SJ, Frith PA, Stranks SN. Relationship between glycaemia and length of hospital stay during acute exacerbation of chronic obstructive pulmonary disease. Intern Med J 2013;43:721-4.
13Chakrabarti B, Angus RM, Agarwal S, Lane S, Calverley PM. Hyperglycaemia as a predictor of outcome during non-invasive ventilation in decompensated COPD. Thorax 2009;64:857-62.
14Feary JR, Rodrigues LC, Smith CJ, Hubbard RB, Gibson JE. Prevalence of major comorbidities in subjects with COPD and incidence of myocardial infarction and stroke: A comprehensive analysis using data from primary care. Thorax 2010;65:956-62.
15Lee CT, Mao IC, Lin CH, Lin SH, Hsieh MC. Chronic obstructive pulmonary disease: A risk factor for type 2 diabetes: A nationwide population-based study. Eur J Clin Invest 2013;43:1113-9.
16Kasirye Y, Simpson M, Mamillapalli CK, Epperla N, Liang H, Yale SH. Association between blood glucose level and outcomes in patients hospitalized for acute exacerbation of chronic obstructive pulmonary disease. WMJ 2013;112:244-9.
17Dave C, Turner A, Thomas A, Beauchamp B, Chakraborty B, Ali A, et al. Utility of respiratory ward-based NIV in acidotic hypercapnic respiratory failure. Respirology 2014;19:1241-7.
18Burt MG, Roberts GW, Aguilar-Loza NR, Frith P, Stranks SN. Continuous monitoring of circadian glycemic patterns in patients receiving prednisone for COPD. J Clin Endocrinol Metab 2011;96:1789-96.
19Roberts GW, Monteiro VE. Pattern of high-dose prednisolone-induced hyperglycaemia in COPD exacerbations. J Pharm Pract Res 2009;39:50-4.
20Mendez CE, Mok KT, Ata A, Tanenberg RJ, Calles-Escandon J, Umpierrez GE. Increased glycemic variability is independently associated with length of stay and mortality in non-critically ill hospitalized patients. Diabetes Care 2013;36:4091-7.
21Clement S, Braithwaite SS, Magee MF, Ahmann A, Smith EP, Schafer RG, et al., American Diabetes Association Diabetes in Hospitals Writing Committee. Management of diabetes and hyperglycemia in hospitals. Diabetes Care 2004;27:553-91.
22Chaudhuri A, Umpierrez GE. Oxidative stress and inflammation in hyperglycemic crises and resolution with insulin: Implications for the acute and chronic complications of hyperglycemia. J Diabetes Complications 2012;26:257-8.
23Nielson CP, Hindson DA. Inhibition of polymorphonuclear leukocyte respiratory burst by elevated glucose concentrations in vitro. Diabetes 1989;38:1031-5.
24Umpierrez GE, Isaacs SD, Bazargan N, You X, Thaler LM, Kitabchi AE. Hyperglycemia: An independent marker of in-hospital mortality in patients with undiagnosed diagnosis. J Clin Endocrinol Metab 2002;87:978-82.
25Kavanagh BP, McCowen MB, McCowen KC. Clinical practice. Glycemic control in the ICU. N Eng J Med 2010;363:2540-6.
26Moghissi ES. Addressing hyperglycemia from hospital admission to discharge. Curr Med Res Opin 2010;26:589-98.