practice

Predicting diabetic foot ulcer infection

using the neutrophil-to-lymphocyte
ratio: a prospective study

Objective: To investigate whether the neutrophil-to-lymphocyte ratio
(NLR) may be used in the early stage risk assessment and follow-up
in diabetic foot infection
Methods: Over a five-year study, NLR values on admission and
day 14 of treatment were matched with their laboratory and clinical
data in a cohort study. Patients were followed-up or consulted in
several clinics or polyclinics (infectious diseases).
Results: Admission time NLR was higher, in severe cases as
indicated by both Wagner and PEDIS infection scores (severe
versus mild Wagner score NLR 6.7 versus 4.2; p=0.04; for PEDIS
score NLR 6.3 versus 3.6; p=0.03, respectively). In patients who
underwent vascular intervention (12.6 versus 4.6; p=0.02);
amputation indicated (9.2 versus 4.1; p=0.005) and healed
afterwards (6.9 versus 4.3; p<0,001), when matched with others.
NLR was also found to be correlated with duration of both IV
antibiotic treatment (r=0.374; p=0.005) and hospitalisation (r=0.337;
p=0.02). Day 14 NLR was higher in patients who underwent
vascular intervention (5.1 versus 2.9; p=0.007) when matched
to others.
Conclusion: Patients with higher NLR values at admission had more
severe diabetic foot infection, higher risk for amputation, need for
long-term hospitalisation and aggressive treatment. However, they
also have more chance of benefit from treatment.
Declaration of interest: The authors have no conflicts of interest.

amputation ● diabetic foot infection ● neutrophil-to-lymphocyte ratio ● primary care

I
nfected diabetic foot ulcers (DFU) are an important
community health problem because they threaten
limb and life by damaging tissues, and presents a
great financial burden on the community.1,2
Overall costs were found to be increased fourfold
by DFUs and sixfold by amputations in patients with
diabetes for CODE-2 data from Germany.3 Some 7–20%
of the resources laid out for diabetes in North America
and Europe are attributed to diabetic foot.4 Strategies to
reduce the risk of lower-extremity amputation may
generate substantial economic benefits and should be a
standard component of routine diabetes care.5
The neutrophil-to-lymphocyte ratio (NLR) has been
found to be useful for predicting degree of severity and
clinical outcomes in a wide variety of clinical conditions,
from malignancy to acute appendicitis.6–9 NLR was
previously studied in people with diabetes in the context
of chronic kidney disease,10,11 impaired glucose regulation,
microvascular complications,12,13 retinopathy14 and other
complications.15 But there are only a few studies evaluating
NLR in DFU infection16,17 and, to our knowledge, no
study has investigated how it is affected by treatment.
For this reason, we aimed to investigate whether NLR
may be used in the early stage risk assessment and
prediction of prognosis in diabetic foot infection, in
addition to gaining knowledge about its correlation
© 2019 MA Healthcare ltd
is followed up by several clinical branches. Over five
years, we recruited patients consulted in clinics
(infectious diseases, internal medicine, plastic surgery)
or polyclinics (infectious diseases). All patients were
prospectively followed, and clinical and laboratory
changes were recorded normally in follow-up forms.
We had categorised DFU infection using both the
Wagner and PEDIS infection scores. The Wagner scoring
system gives priority to the status of the infected ulcer
and extremity. The PEDIS infection scoring system
considers not only the ulcer’s status but systemic
manifestations of infection as well. To secure convenience
for matching, we placed the patients into two groups, as
mild and severe patients indicated by both scoring
systems. Wagner 0 (no ulcer but acute cellulitis), 1 and 2
and PEDIS 2 were defined as mild cases, while Wagner
3–5 and PEDIS 3, 4 cases were defined as severe.
We followed up the patients until the treatment was
finished by medical or surgical treatment or the patient
ended the process by his/her own decision. We used the
*Fatma Aybala Altay,1 MD, Associate Professor; Semanur Kuzi,2 MD, Specialist;
Mustafa Altay,3 MD, Professor; İhsan Ateş,4 MD, Associate Professor; Yunus Gürbüz,1
MD, Associate Professor; Emin Ediz Tütüncü,1 MD, Professor; Gönül Çiçek Şentürk,1
MD, Specialist; Nilgün Altın,1 MD, Specialist; İrfan Şencan,1 MD, Professor
*Corresponding author email: aybalaaltay@hotmail.com

with the treatment results.
Methods
Our hospital is a 700-bed tertiary care hospital, which
is also a teaching and research facility. DFU infection
1 Department of Infectious Diseases and Clinical Microbiology, University
of Health Sciences, Dışkapı Yıldırım Beyazıt Research and Training Hospital,
Ankara, Turkey. 2 Infectious Diseases Department, Artvin State Hospital, Artvin,
Turkey. 3 Department of Endocrinology and Metabolism, University of Health Sciences,
Keçiören SUAM, Ankara, Turkey. 4 Department of Internal Medicine, University of
Health Sciences, Ankara Numune SUAM, Ankara, Turkey.

JOURNAL OF WOUND CARE VOL 28, NO 9, SEPTEMBER 2019 601

Downloaded from magonlinelibrary.com by 128.243.044.244 on September 13, 2019.
 practice

Table 1. Laboratory results at admission and at day 14 admission time value of WBC and day 14 value
of WBC).
Variables At admission At day 14 p-value
n=101 n=101
Statistical analysis
WBC, x10 6/l 11,100 8500 <0.001* Statistical analysis was made by using the Statistical
(min–max) (3000–38,200) (2800–16,900)
Package for Social Sciences (SPSS) for Windows 20 (IBM
PMNL, x10 6/l 8300 (910–36,500) 5375 (500–13,600) <0.001* SPSS Inc., Chicago, IL) programme. Normal distribution
(min–max) of the data was assessed by the Kolmogorov-Smirnov
Lymphocyte, x10 6/l 1600 (400–3440) 1800 (117–4600) 0.09 test. Normally distributed numeric variables were
(min–max) expressed as mean±standard deviation (SD), while
abnormally distributed ones were shown as median
NLR (min–max) 4.7 (0.5–53.0) 3.0 (0.29–40.5) <0.001*
(range: min-max). Categorical variables were
Hgb, g/dl, ±SD 11.6±1.9 10.9±2.0 <0.001* denominated in numbers and percentiles. For
matching, two groups of numeric variables, student’s
MPV, fl, ±SD 8.4±1.2 8.1±1.2 0.003*
T-test and Mann-Whitney U test were used, while Chi-
CRP, mg/l, (min–max) 76.8 (3–449) 17.6 (1.3–306) <0.001* square and Fischer’s exact test were used for matching
categorical variables. Comparison of initial and day 14
ESR, mm/h, (min–max) 80 (6–140) 74 (5–157) 0.01*
data was made by paired sample T-test and Wilcoxon
Numerical variables are shown as mean±standard deviation (SD) or median (min–max—minimum/ signed rank test. Differences between initial and day
maximum); *p<0.05 is accepted as statistical significance cut-off; WBC—white blood cell count;
PMNL—neutrophil; NLR—neutrophil-to-lymphocyte ratio; Hgb—haemoglobin; MPV—mean 14 values were shown with ‘Δ’. Spearman’s correlation
platelet volume; CRP—C-reactive protein; ESR—erythrocyte sedimentation rate analysis was used to analyse the relation between
numeric variables. Probable risk factors for dependent
variables were chosen, as the variables which had a
data of intravenous (IV) antibiotic treatment duration, p-value of <0.25. Independent predictors were assessed
total treatment duration and duration of hospitalisation by multivariable robust regression and multivariable
separately for matching. To define the term ‘amputation’, logistic regression analysis backward methods. A value
we considered the indication for amputation as stated of p<0.05 was accepted as statistically significant.
by a specialist. Vascular interventions included not only Unstandardised coefficients were symbolised as β±(SE)
vascular surgery but placement of a stent(s) or (SE, standard error) and odds ratio was symbolised
application of balloon dilatation catheters. as OR.
Patients’ demographic characteristics and concurrent
laboratory results were taken from their follow-up forms Results
and records. Age, Wagner and PEDIS scores, The study population consisted of a total of 101 patients
hospitalisation time, treatment duration, duration of with DFU infection, 45 woman (44.6%) and 56 men
diabetes, peripheral vascular status, osteomyelitis or (55.4%). Mean age was 60.7±11.9 years. Of these, 5%
abscess presence, treatment modalities are all normally (n=5) has diabetes type 1, while 95% (n=96) had
recorded for each patient and were used in this study. diabetes type 2. The duration of diabetes was between
White blood cell (WBC), neutrophil, lymphocyte, 1–40 years, median 13 years. Infected DFU time was
haemoglobin, HbA1c, C-reactive protein (CRP), median 15 days (range: 2–365 days). Bacterial growth
erythrocyte sedimentation rate (ESR) values on was detected in 57.4% (n=58) of the sample cultures
admission and all but HbA1c values on day 14 of taken from patients. Mild Wagner patients were 33.7%
treatment were taken. Hospitalisation statistics, clinical (n=34) and severe Wagner patients were 66.3% (n=67)
results and exposure to administrations during of total. Mild PEDIS and severe PEDIS were 25.7% (n:26)
follow-up were also reviewed. and 74.3% (n=75), respectively. The number of healing
NLR was calculated by dividing the neutrophil count patients on day 14 of treatment was 47 (46.5%).
by the lymphocyte count. Clinical and laboratory data Laboratory values of the patients on admission and on
of these patients were reviewed via their records. Both day 14 of treatment are seen in Table 1. There was a
clinical and laboratory findings recorded at admission significant decrease in WBC, NLR, CRP and ESR on day
and on day 14 of treatment were chosen, to determine 14 of treatment.
the effect of treatment on these results and to be able to
understand whether these findings could cast a light on Admission value of NLR and correlations
patients’ follow-up. Admission value of NLR was higher for the patients who
DFU healing was described as at least 50% reduction had: positive culture results; higher Wagner or PEDIS
of wound area or improvement in Wagner grade18 and scores; undergone vascular invention; been given
© 2019 MA Healthcare ltd

absence or regression of drainage, cellulitis or pain,
which were defined previously.19
The difference between admission and day 14 values
of laboratory test results were showed with the label ‘Δ’
in front of it (e.g. ΔWBC means: the difference between
amputation indication; healed by treatment when
matched with others (Table 2). Admission NLR was also
positively correlated with admission MPV, CRP, ESR and
day 14 WBC, PMNL, CRP and ESR. It was found to be
positively correlated with duration of IV antibiotic and

602 JOURNAL OF WOUND CARE VOL 28, NO 9, SEPTEMBER 2019

Downloaded from magonlinelibrary.com by 128.243.044.244 on September 13, 2019.

 practice

Table 2. Match of admission and day 14 neutrophil-lymphocyte ratio (NLR), according to patient characteristics

Variables Admission time Patients

∆p
NLR p-value NLR p-value
Gender
Female 4.8 (1.2–53.0) 2.6 (1.4–13.6)
0.494 0.303 0.092
Male 4.7 (0.5–26.7) 3.2 (0.3–40.5)

Type of diabetes
Type 1 3.3 (1.2–20.4) 3.4 (1.6–7.3)
0.253 0.938 0.398
Type 2 4.9 (0.5–53.0) 3.0 (0.3–40.5)

Bacterial growth
No 4.3 (0.5–20.7) 3.2 (0.3–40.5)
0.043* 0.908 0.043*
Yes 6.6 (1.2–53.0) 3.0 (1.3–16.0)

Wagner score
Mild 4.2 (1.2–33.2) 2.9 (1.4–40.5)
0.040* 0.867 0.037*
Severe 6.7 (0.5–53.0) 3.1 (0.3–13.6)

PEDIS score
Mild 3.6 (1.2–20.7) 3.1 (1.4–40.5)
0.027* 0.692 0.003*
Severe 6.3 (0.5–53.0) 3.0 (0.3–13.6)

Circulation deficiency
No 4.4 (0.5–33.2) 2.9 (0.3–40.5)
0.249 0.25 0.598
Yes 5.5 (1.2–53.0) 3.2 (1.3–16.0)

Osteomyelitis
No 5.2 (1.2–53.0) 3.3 (1.4–40.5)
0.454 0.148 0.654
Yes 4.5 (0.5–21.7) 2.8 (0.3–9.3)

Hyperbaric oxygen
No 4.6 (0.5–53.0) 2.9 (0.3–40.5)
0.887 0.731 0.933
Yes 6.9 (1.2–14.3) 3.4 (1.6–13.6)

Abscess
No 4.4 (0.5–53.0) 2.9 (0.3–40.5)
0.32 0.530 0.027*
Yes 6.8 (1.6–21.7) 3.1 (1.7–7.8)

Debridement
No 4.8 (0.5–33.2) 3.2 (0.3–40.5)
0.854 0.064 0.406
Yes 4.7 (1.2–53.0) 2.4 (1.3–9.8)

Vascular intervention
No 4.6 (0.5–17.5) 2.9 (0.3–9.5)

Yes 0.018* 0.007* 0.022*

12.6 5.1 (4.6–40.5)
(3.1–53.0)

Amputation indication
No 4.1 (0.5–33.2) 2.9 (0.3–40.5)
0.005* 0.06 0.012*
Yes
© 2019 MA Healthcare ltd

9.2 (1.8–53.0) 3.3 (1.3–13.6)

Healing
No 4.3 (0.5–33.2) 3.3 (0.3–16)
<0.001* 0.382 0.009*
Yes 6.9 (1.2–53.0) 2.9 (1.3–40.5)

JOURNAL OF WOUND CARE VOL 28, NO 9, SEPTEMBER 2019 603

Downloaded from magonlinelibrary.com by 128.243.044.244 on September 13, 2019.
 practice

Table 3. Correlations of neutrophil-to-lymphocyte ratio (NLR)

NLR1 NLR2 ΔNLR

Variables
r p r p r p

Age 0.031 0.755 0.077 0.453 0.006 0.952

DM time 0.097 0.336 0.112 0.274 –0.068 0.505

WBC1 0.733 <0.001* 0.329 0.001* –0.664 <0.001*

PMNL1 0.835 <0.001* 0.388 <0.001* –0.739 <0.001*

Lymphocyte1 –0.651 <0.001* –0.463 <0.001* 0.464 <0.001*

Hgb1 –0.399 0.002* –0.194 0.055 0.306 0.042*

MPV1 0.388 0.003* 0.114 0.266 –0.342 0.001*

CRP1 0.740 <0.001* 0.385 <0.001* –0.637 <0.001*

ESR1 0.444 <0.001* 0.349 0.014* –0.350 <0.001*

WBC2 0.317 0.032* 0.382 <0.001* –0.041 0.689

PMNL2 0.403 <0.001* 0.652 <0.001* –0.117 0.251

Lymphocyte2 –0.345 <0.001* –0.668 <0.001* 0.031 0.764

Hgb2 –0.511 <0.001* –0.345 0.001* 0.418 <0.001*

MPV2 0.051 0.617 –0.008 0.935 –0.144 0.156

CRP2 0.517 <0.001* 0.564 <0.001* –0.336 0.001*

ESR2 0.549 <0.001* 0.389 0.006* –0.484 <0.001*

DFI duration (day) –0.140 0.162 –0.123 0.227 0.104 0.307

IVs antibiotics time 0.374 0.005* 0.040 0.695 –0.364 <0.001*

HbA1c –0.037 0.728 –0.394 0.005* –0.178 0.095

Hospitalisation days 0.337 0.022* 0.072 0.483 –0.338 0.002*

Total treatment days 0.060 0.556 –0.021 0.843 –0.189 0.067

ΔWBC –0.685 <0.001* –0.140 0.170 0.741 <0.001*

ΔPMNL –0.742 <0.001* –0.114 0.266 0.819 <0.001*

Δlymphocyte 0.338 0.001* –0.319 0.030* –0.496 <0.001*

ΔHgb –0.355 0.01** –0.336 0.019* 0.327 0.025*

ΔMPV –0.357 <0.001* –0.135 0.187 0.390 0.003*

ΔCRP –0.638 <0.001* –0.188 0.075 0.598 <0.001*

ΔESR 0.116 0.282 0.100 0.355 –0.108 0.321

DM—diabetes mellitus; WBC—white blood cell count; 1—variable 1: the admission time value of parameter; 2—variable 2: day 14 value of parameter; PMNL—neutrophil;
HgB—haemoglobin; MPV—mean platelet volume; CRP—C-reactive protein; ESR—erythrocyte sedimentation rate; NLR1—admission time neutrophil/lymphocyte ratio; NLR2—day 14
neutrophil/lymphocyte ratio; ΔNLR—the difference between NLR2 and NLR1; ΔWBC—the difference between WBC2 and WBC1; ΔPMNL—the difference between PMNL2 and PMNL1;
Δlymphocyte—the difference between lymphocyte2 and lymphocyte1; ΔHgb—the difference between Hgb2 and Hgb1; ΔMPV—the difference between MPV2 and MPV1; ΔCRP—the
difference between CRP2 and CRP1; ΔESR—the difference between ESR1 and ESR2

hospitalisation as well (Table 3). Admission NLR was
negatively correlated with ΔWBC, ΔPMNL,
Δhaemoglobin, ΔMPV and ΔCRP, while it was positively
correlated with Δlymphocyte and not correlated with
ΔESR. (Table 3).
intervention were an exception, having a higher NLR
than those who had not undergone vascular
intervention (Table 2). Day 14 NLR showed positive
correlation with admission WBC, PMNL, CRP, ESR,
while it was negatively correlated with admission
© 2019 MA Healthcare ltd

lymphocyte level. Furthermore, it was also positively

NLR and correlations day 14 correlated with day 14 CRP and ESR and negatively
The day 14 median NLR was similar between patient correlated with day 14 haemoglobin and HbA1c. Day 14
subgroups categorised by different features; only the NLR was negatively correlated with Δlymphocyte and
subgroup of patients which had undergone vascular Δhaemoglobin level (Table 3).

604 JOURNAL OF WOUND CARE VOL 28, NO 9, SEPTEMBER 2019

Downloaded from magonlinelibrary.com by 128.243.044.244 on September 13, 2019.


Alteration of NLR between admission and day 14 and
its correlations
Statistical analysis of the subgroups about admission
and day 14 treatment values of NLR showed that NLR
had decreased more at day 14 of treatment in: culture
positive patients when compared with negative ones;
serious Wagner patients when compared with mild
ones; serious PEDIS patients when compared with mild
ones; patients with abscesses when compared with
those not having an abscess; patients with vascular
intervention compared with those without it;
amputation indicated patients when compared with
not indicated ones; and lastly, healing wounds when
compared with non-healing wounds (Table 2).
ΔNLR was found to be negatively correlated with
admission time values of WBC, PMNL, MPV, CRP and
ESR and positively correlated with lymphocyte and
haemoglobin. It was positively correlated with day 14
haemoglobin, while it was negatively correlated with
day 14 CRP and ESR. ΔNLR showed negative correlation
with duration of IV antibiotic treatment and duration
of hospitalisation. ΔNLR was positively correlated with
each of ΔWBC, ΔPMNL, ΔHGB, ΔMPV, and ΔCRP, while
negatively correlated with Δlymphocyte (Table 3).
Associations between NLR values
and clinical/laboratory outcomes
Admission value of NLR was determined as an
independent predictor of amputation indication
(OR:1.081; p=0.021). Severe PEDIS score was an
independent predictor of unchanged NLR on day 14 of
treatment (β±SE=−3.423±1.525; p=0.027). Also, the
decrease in CRP level at day 14 of treatment was
detected as a predictor of decrease in NLR at day 14
(β±SE= 2.471±0.689; p=0.001).
Admission time NLR was higher in severe cases when
matched with mild ones by considering both Wagner
and PEDIS infection scoring systems (for Wagner
scoring 6.7 versus 4.2; p=0.04; for PEDIS scoring 6.3
versus 3.6; p=0.03, respectively). Admission time NLR
was also found to be higher in patients who had
undergone vascular intervention during follow-up
when matched to those who had not (12.6 versus 4.6;
p=0.02). It was also higher in patients who were
indicated to have amputation when matched to those
who were not so indicated (9.2 versus 4.1; p=0.005).
When admission time NLR was analysed in respect to
laboratory results, it was higher in cases where bacterial
growth was detected in cultures of samples taken from
diabetic ulcers (6.6 versus 4.3; p=0.04). Admission time
NLR was positively correlated with admission time
values of WBC (r=0.733; p<0.001), PMNL (r=0.835;
p<0.001), MPV (r=0.388; p=0.003), CRP (r=0.740;
p<0.001) and ESR (r=0.444; p<0.001).
© 2019 MA Healthcare ltd
Another interesting result was that admission time
NLR was found to be correlated with duration of IV
antibiotic treatment (r=0.374; p=0.005) and duration of
hospitalisation (r=0.337; p=0.02). Also, interestingly,
admission time NLR was higher in patients who healed
JOURNAL OF WOUND CARE VOL 28, NO 9, SEPTEMBER 2019
practice
afterwards when compared with those who did not
(6.9 versus 4.3; p<0.001).
The day 14 value of NLR was significantly higher in
patients who had undergone vascular intervention at
any time during the treatment (5.1 versus 2.9; p=0.007)
when matched to ones who had not; while no
significant relation with other clinical parameters was
observed. The day 14 NLR was positively correlated
with day 14 values of WBC (r=0.382; p<0.001), PMNL
(r=0.652; p<0.001), CRP (r=0.564; p<0.001) and ESR
(r=0.389; p=0.006).
Discussion
Our results suggested that the patients who have higher
NLR values at admission are those whose DFUs are more
severe, who would need vascular intervention, who
have higher risk for amputation, who would need
hospitalisation and long-term IV treatment.
Furthermore, these patients seem to have a better
chance of gaining benefit from treatment. Such
conclusions are valuable because they give important
clues for prediction of the prognosis and consequently
management of DFU infections.
Our hospital is a large tertiary care hospital that gives
health-care to patients coming from both its residential
area and nearby provinces. We usually meet with severe
or complicated cases, but our mild cases are not rare
because of easily attainable traits of our health-care
system. So we have a large, wide-ranging population of
DFU infection patients.
An important problem we met is that the patients lost
time by waiting for the results of a treatment that was
actually ineffective. This was because a DFU is usually not
as simple as it seems and abscess, osteomyelitis or
vascularisation troubles cannot be detected in primary
care facilities. So we have many patients coming with poor
results of one or two weeks’ treatments. Infected DFU time
was median 15 days with a range between 2–365 days.
Clinicians need an easily attainable and inexpensive
method to predict the need for an aggressive and
extensive approach for the treatment of infected
DFUs, especially in primary care facilities, where
imaging techniques or specialist consultations are not
normally available.
Although NLR is shown to be associated with severity
and predictive for prognosis in many inflammatory
conditions20,21 and some severe infections,22–24 the
association between DFU infection and NLR is less
known. There are some studies investigating the
association between NLR and DFUs,16,17 but our study
focuses on infected DFUs only.
We chose infected DFUs, all of which were untreated
or not cured by previous antibiotic treatment. We also
chose those patients having only one lesion at the time
of admission and no coinfection in any part of the
body, in order to avoid misdirecting high levels of
inflammatory markers.
We used both admission time and day 14 of treatment
data to be able to determine the relation between
605
Downloaded from magonlinelibrary.com by 128.243.044.244 on September 13, 2019.
 practice
healing and NLR alteration. This data gave us very
beneficial results that the decrease of NLR by appropriate
treatment is correlated with healing in infected DFU
patients. The positive correlation between ΔNLR and
ΔCRP supports this conclusion, because we know CRP
decrease correlates with clinical improvement.25
Admission time median NLR was higher in severe
cases when matched with mild ones, as measured by
both the Wagner and PEDIS infection scoring systems.
This result is consistent with the data from past studies
that NLR is associated with the severity of infection or
inflammatory event.20–24
Past studies have not included a categorisation of the
infected DFU. For example, Ong et al. had investigated
infected and non-infected DFU inpatients
retrospectively.16 They had found that initial ESR, WBC
and NLR were higher in osteomyelitis patients and only
ESR was higher in soft tissue infected patients when
compared with non-infected DFU patients. Their match
was between infected and non-infected DFU patients.
Control results taken at six months showed a difference
at only WBC in osteomyelitis patients compared with
soft tissue infected ones. Their results showed that NLR
had been weakly correlated with WBC, ESR and CRP
initially and not correlated with any of them at the last
visit that occurred at six months. Most of the data was
unavailable because the number of patients was halved
at the last visit. They showed that coexistence of high
WBC and high NLR at initial visit was superior to other
inflammatory markers in discrimination of infected and
non-infected DFU.
Yapıcı et al.17 had investigated 75 infected DFUs
retrospectively to learn whether NLR is related with
osteomyelitis and the need for amputation. The patients
who had undergone amputation or debridement had
higher NLR values compared with the ones who had
not had any surgical intervention (15.7±10.3, 9.9±5.6
and 6.0±2.8, respectively). But it should be considered
that 71% of all patients had osteomyelitis and
osteomyelitis was solely found to be related with higher
NLR values.
Luo et al. had found that post-treatment NLR was an
independent predictor of amputation in inoperable
critical limb ischaemia patients. This result is consistent
with ours, but this study’s population had a diabetes
rate of only 17% and furthermore none of them had an
infected DFU,26 these patients also showed increased
risk of amputation. We found that admission time NLR
was also higher in patients who had undergone a
vascular intervention during the study, when matched
to those who had not (12.6 versus 4.6; p=0.018). Day 14
NLR was also higher in those patients who underwent
vascular intervention (5.1 versus 2.9; p=0.007) when
matched to others. This result can be explained by
ongoing vascular problems at day 14 of treatment,
because we had found it difficult to make our patients
receive an early vascular intervention at the times when
the participants were enrolled to the study. This result
could be used to comment on the condition of the
606
circulation in patients who had not had a proper
response to the treatment. High values of admission
NLR may be a sign of severe ischaemia, indicating a need
to examine the arterial blood supply potentially leading
to an intervention.
In our study, admission time median NLR was found
to be higher in patients who were indicated to have
amputation when matched to ones who were not
indicated. This is an important result, because it can
directly affect the management of the patient. It can be
an early warning for referring the patient to a centre
where he/she could be managed in an multidisciplinary
approach. Because one of the main goals for DFU
infection management is to prevent amputation, this
simple information may be critically important for
saving the limb.
These findings reveal that admission time NLR is
correlated with the severity of the DFU infection, by
means of arterial insufficiency, presence of osteomyelitis
or deep abscess, and indication of amputation. Our
results are consistent with the results of Demirdal and
Sen.27 Other noteworthy results we found were that
admission time NLR was correlated with duration of IV
antibiotic treatment and duration of hospitalisation.
This finding makes admission time NLR a good predictor
of severity of DFU infection.
It is good news that admission NLR was higher in
patients who healed afterwards by the treatment,
compared with those who did not. This result suggests
that patients with high NLRs would be difficult and
high-risk cases, but also it would be a worthwhile to
manage them in a multidisciplinary and careful
approach, because they would then have a greater
chance of healing.
When admission time NLR was analysed in respect
to laboratory results, it was positively correlated with
admission time values of WBC, PMNL, MPV, CRP and
ESR. MPV has been shown to be increased in diabetic
complications before and our findings are consistent
with this.28 Correlation of NLR with WBC and PMNL is
a logical necessity because of the arithmetical relation
between WBC and PMNL. Positive correlations between
NLR, CRP and ESR are understandable because we know
CRP and ESR are both good markers of infection such
as neutrophils.
Of course, a cut-off value making the differentiation
between the cases and the probabilities would make
NLR a highly efficient marker. But, sadly, our patients’
data did not permit the setting up of a net value with
good levels of sensitivity and specificity to identify
severe cases or to guess about various projections. Still,
NLR values of six and more would suggest that the
patient had a severe infection and should be referred to
high-level care.
© 2019 MA Healthcare ltd
Limitations
A limitation of the study was that our patient
population, although 101 in size, was not sufficiently
large enough to provide sufficient sensitivity and
JOURNAL OF WOUND CARE VOL 28, NO 9, SEPTEMBER 2019
Downloaded from magonlinelibrary.com by 128.243.044.244 on September 13, 2019.

specitify values, and thus it was difficult to give a useful
cut off value to differentiate mild cases from more
severe ones.
Conclusions
The authors believe use of NLR may be beneficial to
assess the clinical severity and probabilities in DFU
infection. It can act as a guide to predict the presence
of ischaemia, osteomyelitis and/or deep abscess of the
foot, and so the need for aggressive treatment, even
surgical intervention. The first evaluation is usually
carried out in primary care, where imaging techniques
or specialist consultations could not be held and the
clinician usually feels that he/she needs a crystal ball
References
1 Laing P. The development and complications of diabetic foot ulcers.
Am J Surg 1998; 176(2 Suppl):11S–19S. https://doi.org/10.1016/
S0002-9610(98)00182-2
2 Gordois A, Scuffham P, Shearer A et al. The health care costs of
diabetic peripheral neuropathy in the US. Diabetes Care 2003;
26(6):1790–1795. https://doi.org/10.2337/diacare.26.6.1790
3 Liebl A, Spannheimer A, Reitberger U, Görtz A. [Costs of long-term
complications in type 2 diabetes patients in Germany. Results of the
CODE-2 Study] [Article in German]. Med Klin (Munich) 2002; 97(12):713–
719. https://doi.org/10.1007/s00063-002-1215-z
4 Boulton AJ, Vileikyte L, Ragnarson-Tennvall G, Apelqvist J. The global
burden of diabetic foot disease. Lancet 2005; 366(9498):1719–1724.
https://doi.org/10.1016/S0140-6736(05)67698-2
5 Ollendorf DA, Kotsanos JG, Wishner WJ et al. Potential economic
benefits of lower-extremity amputation prevention strategies in diabetes.
Diabetes Care 1998; 21(8):1240–1245. https://doi.org/10.2337/
diacare.21.8.1240
6 Tang L, Li X, Wang B et al. Prognostic value of neutrophil-to-
lymphocyte ratio in localized and advanced prostate cancer: a
systematic review and meta-analysis. PLoS One 2016; 11(4):e0153981.
https://doi.org/10.1371/journal.pone.0153981
7 Wu Y, Chen Y, Yang X et al. Neutrophil-to-lymphocyte ratio (NLR) and
platelet-to-lymphocyte ratio (PLR) were associated with disease activity
in patients with systemic lupus erythematosus. Int Immunopharmacol
2016; 36:94–99. https://doi.org/10.1016/j.intimp.2016.04.006
8 Aykan AÇ, Hatem E, Kalaycıoğlu E et al. Neutrophil-to-lymphocyte
ratio may be a marker of peripheral artery disease complexity. Anatolian
Journal of Cardiology. 2016; 16(7):497-503. https://doi.org/ 10.5152/
Anatol J Cardiol.2015.6240.
9 Markar SR, Karthikesalingam A, Falzon A, Kan Y. The diagnostic value
of neutrophil:lymphocyte ratio in adults with suspected acute
appendicitis. Acta Chir Belg 2010; 110(5):543–547. https://doi.org/10.10
80/00015458.2010.11680673
10 Nakhjavani M, Aghajani Nargesi A et al. Changes in leukocyte
subpopulations with decline in glomerular filtration rate in patients with
type 2 diabetes. Acta Med Iran 2015; 53(7):425–431 .
11 Huang W, Huang J, Liu Q et al. Neutrophil-lymphocyte ratio is a
reliable predictive marker for early-stage diabetic nephropathy.
Horumon To Rinsho 2015; 82(2):229–233. https://doi.org/10.1111/
cen.12576
12 Demirtas L, Degirmenci H, Akbas EM et al. Association of
hematological indicies with diabetes, impaired glucose regulation and
microvascular complications of diabetes. Int J Clin Exp Med.
2015;8(7):11420-11427. eCollection 2015.
13 Öztürk ZA, Kuyumcu ME, Yesil Y et al. Is there a link between
neutrophil-lymphocyte ratio and microvascular complications in geriatric
diabetic patients? J Endocrinol Invest 2013; 36(8):593–599. https://doi.
org/10.3275/8894
14 Yue S, Zhang J, Wu J et al. Use of the monocyte-to-lymphocyte ratio
to predict diabetic retinopathy. Int J Environ Res Public Health 2015;
© 2019 MA Healthcare ltd
12(8):10009–10019. https://doi.org/10.3390/ijerph120810009
15 Azab B, Chainani V, Shah N, McGinn JT. Neutrophil-lymphocyte ratio
as a predictor of major adverse cardiac events among diabetic
population: a 4-year follow-up study. Angiology 2013; 64(6):456–465.
https://doi.org/10.1177/0003319712455216
16 Ong E, Farran S, Salloum M et al. Does everything that’s counted
count? Value of inflammatory markers for following therapy and
JOURNAL OF WOUND CARE VOL 28, NO 9, SEPTEMBER 2019
practice
Reflective questions
●● How could neutrophil-to-lymphocyte ratio (NLR) be used in
primary care facilities?
●● How can NLR be used in predicting need for early
vascularisation in patients?
●● Can the use of NLR help reduce the risk of amputation? If
so, how?
to evaluate the patient. So such a prediction is critically
important in primary care. We think more studies on
this topic would be helpful to build up a new and
simple tool for the better management of
DFU infection. JWC
predicting outcome in diabetic foot infection. Int J Low Extrem Wounds
2017; 16(2):104–107. https://doi.org/10.1177/1534734617700539
17 Yapıcı O, Berk H, Öztoprak Net al. Can ratio of neutrophil-to-
lymphocyte count and erythrocyte sedimentation rate in diabetic foot
infection predict osteomyelitis and/or amputation? Hematology Reports.
2017; 9(1):6981. https://doi.org/ 10.4081/hr.2017.6981. eCollection 2017
Feb 23.
18 Apelqvist J, Ragnarson Tennvall G. Cavity foot ulcers in diabetic
patients: a comparative study of cadexomer iodine ointment and
standard treatment. An economic analysis alongside a clinical trial. Acta
Derm Venereol 1996; 76(3):231–235 .
19 Davis JC, Heckman JD, DeLee JC et al. Chronic non-hematogenous
osteomyelitis treated with adjuvant hyperbaric oxygen. J Bone Joint
Surg Am. 1986; 68(8):1210–1217. https://doi.
org/10.2106/00004623-198668080-00010
20 Sevinc C, Yilmaz G, Ustundag S et al. The relationship between
mean platelet volume and neutrophil/lymphocyte ratio with inflammation
and proteinuria in chronic kidney disease. Saudi J Kidney Dis Transpl
2017; 28(1):90–94. https://doi.org/10.4103/1319-2442.198152
21 Liu X, Shen Y, Wang H et al. Prognostic significance of neutrophil-to-
lymphocyte ratio in patients with sepsis: a prospective observational
study. Mediators Inflamm 2016;2016:1–8. https://doi.
org/10.1155/2016/8191254
22 Naess A, Nilssen SS, Mo R et al. Role of neutrophil to lymphocyte
and monocyte to lymphocyte ratios in the diagnosis of bacterial
infection in patients with fever. Infection 2017; 45(3):299–307. https://
doi.org/10.1007/s15010-016-0972-1
23 Baglam T, Binnetoglu A, Yumusakhuylu AC et al. Predictive value of
the neutrophil-to-lymphocyte ratio in patients with deep neck space
infection secondary to acute bacterial tonsillitis. Int J Pediatr
Otorhinolaryngol 2015; 79(9):1421–1424. https://doi.org/10.1016/j.
ijporl.2015.06.016
24 Lowsby R, Gomes C, Jarman I et al. Neutrophil to lymphocyte count
ratio as an early indicator of blood stream infection in the emergency
department. Emerg Med J 2015; 32(7):531–534. https://doi.org/10.1136/
emermed-2014-204071
25 Altay FA, Şencan İ, Şentürk GÇ et al. Does treatment affect the
levels of serum Interleukin-6, Interleukin-8 and procalcitonin in diabetic
foot infection? A pilot study. J Diabetes Complications 2012;
26(3):214–218. https://doi.org/10.1016/j.jdiacomp.2012.03.018
26 Luo H, Yuan D, Yang H et al. Post-treatment neutrophil-lymphocyte
ratio independently predicts amputation in critical limb ischemia without
operation. Clinics (São Paulo) 2015; 70(4):273–277. https://doi.
org/10.6061/clinics/2015(04)09
27 Demirdal T, Sen P. The significance of neutrophil-lymphocyte ratio,
platelet-lymphocyte ratio and lymphocyte-monocyte ratio in predicting
peripheral arterial disease, peripheral neuropathy, osteomyelitis and
amputation in diabetic foot infection. Diabetes Res Clin Pract 2018;
144:118–125. https://doi.org/10.1016/j.diabres.2018.08.009
28 Buch A, Kaur S, Nair R, Jain A. Platelet volume indices as predictive
biomarkers for diabetic complications in Type 2 diabetic patients. J Lab
Physicians 2017; 9(2):84–88. https://doi.org/10.4103/0974-2727.199625
607
Downloaded from magonlinelibrary.com by 128.243.044.244 on September 13, 2019.