Results:
Eighty-nine patients were included in the study, accounting for 91
affected ears. The study group included 53 males (59.5%) and 36 females
(40.5%), with a mean age of 69.8 years. Eighty-three (94.3%) patients
had DM at time of admission. Table-1 depicts patient’s characteristics.
The most common complaints at admission were otalgia (n=79, 86.8%) and
aural discharge (n= 43, 47.2%). Edema of the external ear canal (n=67,
73.6%) was the most common finding on physical examination, followed by
Granulation tissue (n=66, 72.5%) and aural discharge (n=44, 48.3%).
Six patients (6.5%) had facial palsy.
Swab cultures were taken from all patients at admission. PA-NOE was the
most common pathogen (n=41, 45.0%), followed by sterile-NOE (n=19,
20.8%), and candida species (n=11, 12.0%). For statistical analysis we
excluded those with missing data, and grouped the results into 4 main
categories:1) PA-NOE (n=41) 2)sterile-NOE (n=19) 3)Fungal-NOE (n=14) 4)
other bacteria’s (n=4).
Mean duration of DM at time of hospitalization was 136.6 months, ranging
from 6–396 months. The most common DM associated complications were
retinopathy (n=26, 30.9%), nephropathy (n=24, 28.5%) and peripheral
neuropathy (n=16, 19.0%). Mean HbA1c, which wasretrieved from 28
patients, was 8.13% (5.8%-12.6%). Mean urine microalbumin, collected
from 40 patients was 195.5 mcg. Blood glucose measurement during
hospitalization was collected from 75 effected ears. Forty-nine patients
(53.8%) had mean blood glucose measurements of ≥140mg\dL
and 26 patients (28.5%) had ≤ 140mg/dL. Table-2 depicts DM related
patient characteristics.
Average duration of hospitalization was 22.3 days. Twenty five patients
(27.4%) required surgical intervention. At time of data collection 55
patients died, among them 3 died during hospitalization.
Tables 3 and 4 evaluate the relationship between DM characteristics and
NOE outcomes. Mean duration of diabetes was 157.88 months among NOE
patients who required surgery in comparison to 127.6 months among
patients who were treated conservatively (p-value 0.25). Thirty six
patients (54.5%) had mean glucose levels ≥140 mg\dLin
the non-surgical group, as opposed to 13 patients (52.0%) in the
surgical group (p-value 0.69). No statistical difference was seen
between the two groups in regards to HbA1c levels (p-value 0.59) and
microalbuminuria (p-value 1.0).Fifty one patients were hospitalized less
than 20 days (56.0%) and 40 patients (44%) were hospitalized longer
than 20 days. Average age was 68.3 years among patients hospitalized
less than 20 days in comparison to 71.7 years among people hospitalized
more than 20 days (p-value 0.15). Mean HbA1camong NOE patients
hospitalized less than 20 days was 7.6% in comparison to 8.7% among
patients hospitalized ≥20 days (p-value 0.027). Duration of diabetes
(p-value 0.65), long term complications of diabetes (p-value 0.45) and
mean glucose measurements during hospitalization (p-value 0.16) were not
statistically significant between the two groups.
In order to further evaluate the effects of glycemic control during
hospitalization on NOE, we compared patients with mean blood glucose
measurements above and below 140 mg\dL (Table-5),
regardless of disease outcome. Among patients with blood glucose
measurements of ≤140mg\dL, 11 patients had long term
complications of DM (42.3%) in comparison to 30 patients (61.2%) with
mean blood glucose measurements of≥140mg\dL (p=0.11).
Duration of DM was associated with higher mean glucose measurements
during hospitalization (p=0.005). Similarly,higher levels of HbA1C were
associated with higher mean glucose measurements during hospitalization
(p=0.001). Sevenpatients with glucose measurements
≤140mg\dL had PA-NOE (26.7%), in comparison to 25
patients (51.0%) among those with glucose
measurements≥140mg\dL (p=0.045). Figure -1 presents the
incidence of isolated pathogen based on mean glucose measurements during
hospitalization.