DOI: 10.14744/ejmo.2017.

30974
EJMO 2017;1(1):14–18

Research Article
Pneumonia in Patients with Diabetes Mellitus:
A Single-Center Experience
Mahmut Polat,1 Mehmet Nuri Ozdemir,2 Erdal Akyer,3 Mehmet Uzunlulu,3 Aytekin Oguz3
1
Department of Internal Medicine, Bitlis Adilcevaz Oncology Hospital, Bitlis, Turkey
2
Department of Internal Medicine, Batman Sason State Hospital, Batman, Turkey
3
Department of Internal Medicine, Istanbul Medeniyet University Goztepe Training and Research Hospital, Istanbul, Turkey

Abstract
Objectives: The rate of infection and hospitalization due to infection as well as the risk of mortality are greater in pa-
tients with diabetes mellitus (DM). The aim of the present study was to determine the frequency of pneumonia, length
of hospital stay, rate of patients requiring hospitalization in the intensive care unit (ICU), and rates of discharge or mor-
tality among patients hospitalized due to DM in internal medicine inpatient clinics.
Methods: A total of 399 patients with DM (170 males and 229 females) were enrolled in the study. The frequency of
pneumonia in the group was determined, and the length of hospital stay, rate of patients requiring hospitalization in
ICU, discharge rate, mortality, and factors related to mortality were evaluated, comparing those with pneumonia and
DM and those with DM alone.
Results: The frequency of pneumonia among the DM patients was 12% (48 cases). Hypertension was the most com-
mon comorbidity in the 2 groups (31.8% and 32.8%, respectively, among those with pneumonia and DM and those
with DM alone). The rate of discharge, mortality, and the rate of patients transferred to the ICU was 92% (367 cases), 4%
(16 cases), and 4% (16 cases). The length of hospital stay, discharge rate, mortality, and the rate of patients transferred
to the ICU was similar between the patients with and without pneumonia (p>0.05 for each), but the length of hospital
stay and the rate of patients transferred to the ICU was higher among patients with pneumonia (p<0.05 for each).
Conclusion: About 10% of diabetic patients were treated for pneumonia. Infection in patients with DM leads to a
longer hospital stay and fewer hospital discharges. The Infection-related disease burden leads to serious morbidity in
patients with DM, and therefore, longer hospital stay and more patients transferred to ICU.
Keywords: Diabetes mellitus, morbidity, mortality, pneumonia

Cite This Article: Polat M, Ozdemir M, Akyer E, Uzunlulu M, Oguz A. Pneumonia in Patients with Diabetes Mellitus: A Single-
Center Experience. EJMO. 2017; 1(1): 14-18

P atients with diabetes mellitus (DM) have an increased

propensity to develop infections. It is a known fact
that the risk of hospitalization and mortality is greater in
mucosal colonization, and surgical wound infections ren-
der diabetic patients more sensitive to infection related
to hyperglycemia.[1] Hospitalization risk due to pneumo-
diabetic patients. Factors such as an impaired immune re- nia has been reported to be 26% higher in diabetic cases
sponse, vascular insufficiency, sensory peripheral neurop- than in control cases, and the risk increases even more in
athy, autonomic neuropathy, serious infections, commu- patients with longer duration of diabetes and in patients
nity-acquired pneumonia (CAP), urinary tract infection, with poor glycemic control.[2] There is a dramatic increase
inclination to necrotizing bacterial or fungal skin and in the frequency of diabetes in parallel with the increase

Address for correspondence: Mahmut Polat, MD. Bitlis Adilcevaz Onkoloji Hastanesi, Dahiliye Klinigi, Bitlis, Turkey
Phone: +90 434 311 21 00 E-mail: mpolat333@hotmail.com
Submitted Date: May 20, 2017 Accepted Date: July 21, 2017 Available Online Date: August 03, 2017
©
Copyright 2017 by Eurasian Journal of Medicine and Oncology - Available online at www.ejmo.org
EJMO 15

in obesity and metabolic syndrome, which are considered Statistical Analysis

to be a global epidemic. This leads to an increase in the Statistical analyses were conducted using SPSS Statistics
frequency of other infections, particularly pneumonia, in for Windows, Version 21.0 (IBM Corp., Armonk, NY, USA).
the diabetic population, which is already susceptible to The Kolmogorov-Smirnov test was employed to deter-
infection. The primary outcome of this study was a de- mine if the distribution of variables was appropriate pri-
termination of the frequency of pneumonia in diabetic or to analysis. Continuous variables were expressed as
patients admitted to internal medicine clinics and a com- mean±SD. Classifiable data were analyzed by using the
parison with literature information. A secondary outcome chi-square test, and numerical data were analyzed using
was an analysis of the relationship between diabetes and Student's t-test and the Mann-Whitney U test. Independent
variables such as mortality, duration of hospitalization, factors affecting mortality were demonstrated in logistic
and intensive care need. regression analysis and significance levels, in addition to
estimated relative risk (odds ratio [OR] and 95% confidence
Methods interval [CI]) values. Results obtained with p<0.05 in 95%
Study Patients CI were accepted as meaningful in terms of statistical sig-
Patients aged 18 and over who were admitted to the nificance.
Internal Medicine Department of the Ministry of Health
İstanbul Medeniyet University Göztepe Training and Re- Results
search Hospital with a diagnosis of diabetes between A total of 399 patients (170 males constituting 42.6%
June 2009 and June 2013 were included in the study and 229 females constituting 57.4%) were included and
consecutively and retrospectively. Approval of institu- evaluated in the study. Pneumonia was detected in 12%
tional ethics committee was received for the study (date: of these patients. The mean age of all of the patients
June 25, 2013; approval no.: 2013/0010). The principles was 68 years (range: 56–78 years), while the mean age
of the Declaration of Helsinki were observed during the of those with pneumonia was 73 years (range: 68–81
study. Inclusion criteria were a diagnosis of diabetes and years), and the mean age of non-pneumonia patients
pneumonia, respectively, according to the International was 67 years (range: 54–76 years). The mean age of the
Classification of Diseases code in the patient file. Exclu- diabetic patients with pneumonia was significantly high-
sion criteria from the study were as follows: aspiration er (p=0.001). The mean length of stay in hospital was 6
pneumonia, possibility of ventilator-associated pneu- days (range: 4–9 days); there was no significant differ-
monia, administration of immunosuppressive therapy, ence between the patients with and without pneumonia
presence of immunosuppressive disease (asplenia, AIDS, (p>0.05). About 4% of diabetic patients admitted to the
organ transplantation, hematological malignancies), ICU for any reason died, while 4% were referred to ICU
and pregnancy. and 92% were discharged. The demographic character-
istics, outcomes, and infections of diabetic patients are
Study Design provided in Table 1. Glycated hemoglobin (HbA1c) level
This study was a retrospective descriptive case series study. in those who died was 6.9±1.5%, while HbA1c in cases
Patient data regarding age, sex, smoking and alcohol us- referred to ICU was 6.6±1.3% and level was 9.6±3.2% in
age, duration of hospital stay, presence of diabetes, treat- the discharged group. HbA1c level in cases of death and
ment received for diabetes (oral antidiabetics and/or in- referral to ICU was statistically significantly lower com-
sulin), and diagnosis of comorbid diseases (hypertension, pared to discharged patients (p=0.009). The most fre-
chronic obstructive pulmonary disease, chronic renal dis- quent infection observed in diabetic patients in the inter-
ease, cardiac insufficiency, dementia, chronic stroke, neo- nal medicine clinics was pneumonia, with a rate of 12%.
plasia, asthma, peripheral artery disease), as well as clinical, This was followed by urinary tract infection, with a rate
radiological, and laboratory data and outcomes (discharge, of 6.8%. Concomitant comorbidities in diabetic patients
referral to intensive care, death) of all participants were re- are provided in Table 2. The most frequent comorbidity
corded from their files. The study’s primary outcome was a observed in diabetic patients was essential hypertension,
determination of the frequency of pneumonia in diabetic with 131 patients (32.8%). The intensive care referral rate
patients. Furthermore, the mean duration of hospitaliza- was significantly higher in diabetic pneumonia patients
tion; rates of discharge, mortality, and referral to the inten- (p<0.05); however, the mortality rate was similar between
sive care unit (ICU); and the frequency of comorbidities and patients with diabetic pneumonia and those without di-
infections in patients hospitalized with diagnosis of diabe- abetic pneumonia (p>0.05). The discharge rate was lower
tes were determined.
16 Polat et al., Pneumonia in Diabetes Mellitus / doi: 10.14744/ejmo.2017.30974

Table 1. Demographic and clinical characteristics of study subjects

Discharged Death ICU p
Age, year 63±16 67±8 70±12 0.07
Gender (%)
Male 157 (40) 5 (1.3) 8 (2) 0.54
Female 210 (53) 11 (2.8) 8 (2)
Length of hospital stay, days 7.5±6.5 4.4±3.7 4.1±2 0.001
HbA1c (%) 9.6±3.2 6.9±1.5 6.6±1.3 0.009
Infectious disease
Pneumonia 36 (34) 2 (1.9) 6 (5.7) 0.31
Urinary tract infection 24 (22.6) 1 (0.9) 2 (1.9) 0.56
Diabetic ulcer 9 (10.2) 0 0 0.36
Celulitis 5 (4.7) 0 0 0.38
COPD exacerbation 4 (4.5) 0 0 0.65
Diarrhea 5 (4.7) 0 1 (0.9) 0.65
Sepsis 2 (1.9) 5 (4.7) 1 (0.9) 0.001
Acute cholangitis 1 (0.9) 0 0 0.98
Acute peritonitis 2 (1.9) 0 0 0.81
COPD: Chronic obstructive pulmonary disease; HbA1c: glycated hemoglobin; ICU: intensive care unit.

Table 2. Comorbidites of diabetes mellitus patients

for pneumonia. It was revealed, however, that pneumonia
did not have an effect on mortality, although it does lead
n %
Hypertension 131 32.8
to longer hospitalization and increased referral to ICU. It
COPD 32 8 is known that the incidence of infection is greater in dia-
Coronary artery disease 55 13.8 betic patients. Diabetes is a significant risk-raising factor
Congestive heart failure 25 6.3 for lower respiratory tract, urinary tract, and bacterial skin
Chronic renal failure 47 11.8 and mucous membrane infections.[3] Shah et al. observed
Cerebrovscular disease 6 1.5 that diabetes increased the risk of pneumonia in a cohort
Chronic liver disease 4 1 study performed in Canada.[4] Pneumonia was detected in
COPD: Chronic obstructive pulmonary disease
12% of 399 diabetic patients in our study. It was followed
by urinary tract infection at a rate of 6.8%. Type 2 diabetes
increased the risk of hospitalization for pneumonia by 1.23
in patients with diabetic pneumonia (p<0.05) (Table 3). times, while type 1 diabetes increased the risk of hospital-
Factors that were not revealed to be significant in multi- ization for pneumonia by 4.43 times, compared with a con-
variate analysis were age, gender, presence chronic renal trol group.[2]
failure (CRF), level of urea, aspartate aminotransferase Data regarding the effect of diabetes on the duration of
(AST), alanine aminotransferase (ALT), HbA1C, and C-re- pneumonia are conflicting.[5–7] Patients among the diabetic
active protein (CRP) (p>0.05). group with pneumonia in our study had a significantly lon-
ger hospitalization time and greater rate of referral to the
Discussion ICU compared with patients without pneumonia (p<0.05).
Approximately 1 of 10 of patients admitted to the inter- However, there was no difference with respect to mortal-
nal medicine clinic with diabetes in this study was treated ity rate (p>0.05). These results may have been due to the

Table 3. Clinical outcome of patients

All (n=399) Pneumonia (+) (n=48) Pneumonia (-) (n=351) p
n % n % n %
Discharged 367 92 40 83.3 327 93.2 0.02
Death 16 4 2 4.2 14 4 0.95
ICU 16 4 6 12.5 10 2.8 0.001
ICU: Intensive care unit
EJMO 17

fact that the average age of the patients with pneumonia of hospitalization was longer for diabetic pneumonia pa-
was higher. We found that factors affecting mortality in di- tients. Similarly, the rate of diabetic pneumonia patients re-
abetic patients such as age, gender, presence of CRF, and ferred to intensive care was significantly higher compared
level of urea, AST, ALT, CRP and HbA1c were not statistically to non-diabetic pneumonia patients. However, mortality
significant. On the other hand, in numerous studies exam- rates were similar. More extensive and prospective studies
ining etiology, clinical features, outcomes of CAP in diabet- and more consistent information are required with regard
ic patients, age, presence of gram-negative pneumonia, to the presence of diabetes in the clinical course of infec-
septic shock table result upon admission, and bacteremia tious diseases.
mortality have been found to be associated in multivariate
Disclosures
analysis.[7] Chronic pulmonary disease, chronic heart dis-
ease, hypertension, male gender, CRP and urea level were Ethics Committee Approval: The study was approved by the
found to be factors affecting mortality in multivariate anal- Local Ethics Committee (date: June 25, 2013; approval no.:
ysis in our study. However, the presence of diabetes was 2013/0010).
not found to be associated with mortality. It has been re- Peer-review: Externally peer-reviewed.
ported that diabetes history does not predict mortality in Conflict of Interest: None declared.
hospital, though hyperglycemia prolongs stay in hospital,
and increases likelihood of morbidity, mortality.[8] Umpier- References
rez et al. found that that glucose level is a more important 1. Kovalaske MA, Gandhi GY. Glycemic control in the medical in-
prognostic factor than DM.[9] Similarly, a comprehensive tensive care unit. J Diabetes Sci Technol 2009;3:1330–41. [CrossRef]
study of the prognosis of CAP found that plasma glucose 2. Kornum JB, Thomsen RW, Riis A, Lervang HH, Schønheyder
was an independent risk factor for mortality, while diabe- HC, Sørensen HT. Diabetes, glycemic control, and risk of hospi-
tes was not.[10] Moreover, a correlation between the severi- talization with pneumonia: a population-based case-control
ty of pneumonia and non-diabetic hyperglycemia was de- study. Diabetes Care 2008;31:1541–5. [CrossRef ]
tected.[11] No statistically significant difference was found 3. Muller LM, Gorter KJ, Hak E, Goudzwaard WL, Schellevis FG,
in death rate when non-diabetic patients were compared Hoepelman AI, et al. Increased risk of common infections in
with diabetic patients in our study. This may indicate that patients with type 1 and type 2 diabetes mellitus. Clin Infect
hyperglycemia level is more important than the presence Dis 2005;41:281–8. [CrossRef ]
of diabetes. HbA1c percentage in patients who died, were 4. Shah BR, Hux JE. Quantifying the risk of infectious diseases for
referred to the ICU, and those discharged, was found to people with diabetes. Diabetes Care 2003;26:510–3. [CrossRef ]
be 6.9±1.5, 6.6±1.3, and 9.6±3.2, respectively, in this study 5. Kornum JB, Thomsen RW, Riis A, Lervang HH, Schønheyder
population. HbA1c level in patients who died and those re- HC, Sørensen HT. Type 2 diabetes and pneumonia outcomes: a
ferred to the ICU were statistically significantly lower com- population-based cohort study. Diabetes Care 2007;30:2251–
pared with discharged patients. This may be related to the 7. [CrossRef ]
fact that intense control of blood glucose level increases 6. Falguera M, Pifarre R, Martin A, Sheikh A, Moreno A. Etiology
the frequency of hypoglycemia, as well as increased mor- and outcome of community-acquired pneumonia in patients
tality.[12–14] with diabetes mellitus. Chest 2005;128:3233–9. [CrossRef ]
There are several limitations to our study. First and fore- 7. Di Yacovo S, Garcia-Vidal C, Viasus D, Adamuz J, Oriol I, Gili F, et
most, it is a retrospective study and does not include data al. Clinical features, etiology, and outcomes of community-ac-
such as duration of the presence of diabetes or organ dam- quired pneumonia in patients with diabetes mellitus. Medi-
age of the patients, which may have had an effect on mor- cine (Baltimore) 2013;92:42–50. [CrossRef ]
tality. Furthermore, diabetes subtypes, namely type 1 DM 8. McAlister FA, Majumdar SR, Blitz S, Rowe BH, Romney J, Mar-
and type 2 DM, were not differentiated and no association rie TJ. The relation between hyperglycemia and outcomes in
was made between mechanisms such as insulin resistance 2,471 patients admitted to the hospital with community-ac-
to infection or metabolic syndrome and pneumonia. In ad- quired pneumonia. Diabetes Care 2005;28:810–5. [CrossRef ]
dition, there was no differentiation made between patients 9. Umpierrez GE, Isaacs SD, Bazargan N, You X, Thaler LM, Kitab-
with diabetes who received a pneumococcal vaccine and chi AE. Hyperglycemia: an independent marker of in-hospital
those who did not. mortality in patients with undiagnosed diabetes. J Clin Endo-
In conclusion, the incidence of pneumonia in patients with crinol Metab 2002;87:978–82. [CrossRef ]
diabetes (399 cases), was 12% (48 cases). Sixteen (4%) of 10. Fine MJ, Auble TE, Yealy DM, Hanusa BH, Weissfeld LA, Singer
399 diabetic patients died, and 16 patients (4%) were re- DE, et al. A prediction rule to identify low-risk patients with
ferred to intensive care. It was observed that the duration community-acquired pneumonia. N Engl J Med 1997;336:243–
18 Polat et al., Pneumonia in Diabetes Mellitus / doi: 10.14744/ejmo.2017.30974

50. [CrossRef ] sensus statement on inpatient diabetes and glycemic control.

11. Rueda AM, Ormond M, Gore M, Matloobi M, Giordano TP, Diabetes Care 2006;29:1955–62. [CrossRef ]
Musher DM. Hyperglycemia in diabetics and non-diabetics: 13. Inzucchi SE. Clinical practice. Management of hyperglycemia in
effect on the risk for and severity of pneumococcal pneumo- the hospital setting. N Engl J Med 2006;355:1903–11. [CrossRef]
nia. J Infect 2010;60:99–105. [CrossRef ] 14. Castellanos MR, Fadel DA, Seminara DP. Brief report: glucose
12. ACE/ADA Task Force on Inpatient Diabetes. American College control in nondiabetic elderly patients with pneumonia may
of Endocrinology and American Diabetes Association Con- be an overlooked therapeutic target. Clin Geriatr 2008;16:25–6.