779550

research-article2018
JVA0010.1177/1129729818779550The Journal of Vascular AccessSpatola et al.

Original research article

JVA The Journal of
Vascular Access

The Journal of Vascular Access

Subjective Global Assessment–Dialysis

2019, Vol. 20(1) 70­–78
© The Author(s) 2018
Article reuse guidelines:
Malnutrition Score and arteriovenous sagepub.com/journals-permissions
https://doi.org/10.1177/1129729818779550
DOI: 10.1177/1129729818779550

fistula outcome: A comparison with journals.sagepub.com/home/jva

Charlson Comorbidity Index

Leonardo Spatola1, Silvia Finazzi, Albania Calvetta,

Claudio Angelini and Salvatore Badalamenti

Abstract
Introduction: Malnutrition is a well-recognized risk factor for all-cause mortality in hemodialysis patients. However, its
role for arteriovenous fistulas outcome has not been exhaustively investigated. Our aim was to point out the impact of
Subjective Global Assessment–Dialysis Malnutrition Score as independent predictor of arteriovenous fistulas thrombosis
(vascular access thrombosis) and/or significant stenosis (vascular access stenosis). In addition, we compared it with the
widespread Charlson Comorbidity Index.
Methods: We assessed 57 hemodialysis patients for a 2-year interval and evaluated the incidence of vascular access
thrombosis and/or stenosis. Linear regression analysis was used to test the relation of variables with Subjective Global
Assessment–Dialysis Malnutrition Score at baseline. Logistic and Cox regression analysis evaluated markers as predictors
of both vascular access thrombosis and stenosis. Receiver operating characteristic curve analysis was used to compare
area under the curve values of Subjective Global Assessment–Dialysis Malnutrition Score, Charlson Comorbidity Index,
and modified Charlson Comorbidity Index.
Results: Age and Charlson Comorbidity Index were positively related to Subjective Global Assessment–Dialysis
Malnutrition Score: B = 0.06 (95% CI = 0.01; 0.11) and B = 0.31 (95% CI = 0.01; 0.63). Higher albumin and normalized
protein catabolic rate levels had a protective role against vascular access failure: OR = 0.67 (95% CI = 0.56; 0.81) and
OR = 0.46 (95% CI = 0.32; 0.67), respectively. Higher Subjective Global Assessment–Dialysis Malnutrition Score and
Charlson Comorbidity Index values were significant risk factors: HR = 1.42 (95% CI = 1.04; 1.92) and HR = 1.48 (95%
CI = 1.01; 2.17), respectively. Area under the curve of Subjective Global Assessment–Dialysis Malnutrition Score was
significantly higher than those of both Charlson Comorbidity Index and modified Charlson Comorbidity Index: 0.70 (95%
CI = 0.50; 0.88) versus 0.61 (95% CI = 0.41; 0.80) and 0.55 (95CI% = 0.41; 0.70).
Conclusion: Subjective Global Assessment–Dialysis Malnutrition Score, as well as Charlson Comorbidity Index, are
useful tools to predict vascular access failure and should be carefully and periodically evaluated in order to check
significant variations that may compromise vascular access survival.

Keywords
Malnutrition, Subjective Global Assessment, Subjective Global Assessment–Dialysis Malnutrition Score, Charlson
Comorbidity Index, arteriovenous fistula failure, vascular access stenosis, vascular access thrombosis, maintenance HD

Date received: 5 March 2018; accepted: 7 May 2018

1Humanitas,
Clinical and Research Center, Renal and Hemodialysis
Introduction Unit, Rozzano (MI), Italy

Corresponding author:
Vascular access thrombosis (VAT) and/or significant steno- Leonardo Spatola, Humanitas, Clinical and Research Center, Renal and
sis (VAS) are the main causes of morbidity, reduced patient Hemodialysis Unit, via Manzoni 56, 20089 Rozzano (MI), Italy.
survival, hospitalization, and decreased quality of life in Email: leonardospatola@yahoo.it
Spatola et al. 71

Table 1. SGA-DMS: all the medical history and physical examination clinical data,22 range interval 7–35.

Patients-related medical history

Clinical data 1 2 3 4 5
Weight Change or WL No weight change WL <5% WL 5%–10% WL 10%–15% WL >15%
or gain <5%
Dietary intake No change Sub-optimal solid Full Liquid or any Hypocaloric liquid Starvation
diet moderate overall
disease
Gastrointestinal symptoms No symptoms Nausea Vomiting Diarrhea Severe
anorexia
Functional capacity None, normal or Difficulty with no/ Difficulty with Light activity Bed/chair
(nutritionally related improved activity little activity ambulation ridden
impairment)
Comorbidity HD <12 months and HD 1–2 years or HD 2–4 years or HD >4 years or Very severe
healthy otherwise mild comorbidity >75 years of age or severe comorbidity multiple
moderate comorbidity comorbiditya
Physical examination
Decreased fat stores or loss None (no change) Mild Moderate Moderately severe Severe
of subcutaneous fat (below
eyes, triceps, biceps, chest)
Signs of muscle wasting None (no change) Mild Moderate Moderately severe Severe
(temple, clavicle, scapula,
ribs, quadriceps, knee,
interosseous muscles)

WL: weight loss; HD: hemodialysis; SGA-DMS: Subjective Global Assessment–Dialysis Malnutrition Score; AIDS: acquired immunodeficiency syndrome.
aVery severe multiple comorbidity includes chronic heart failure Class III or IV, full-blown AIDS, severe coronary artery disease, moderate to severe

chronic obstructive pulmonary disease, and metastatic malignancies and/or recent chemotherapy.

chronic hemodialysis (HD) patients.1–3 Native arterio-
venous fistula (AVF) is the gold standard of vascular access
for HD patients when vessel conditions allow its creation
since it is characterized by long-term superior survival and
a reduced burden of complications compared with syn-
thetic graft fistulas and central venous catheter.3–5 In most
cases, the leading cause of dysfunction of AVF is VAT, usu-
ally due to a pre-existing VAS caused by neointimal hyper-
plasia in the venous outflow tract6,7 and/or to hemodynamic
stasis.8 Different inherited and/or acquired risk factors for
VAT have been recognized, but risk stratification is not
always identifiable.9 Among acquired prothrombotic states,
malnutrition has been investigated in several studies,8,10,11
and both hypoalbuminemia and decreased normalized pro-
tein catabolic rate (nPCR) have been associated with native
AVF failure. In this context, low albumin levels have been
related to inflammation status, showing an inverse relation-
ship with acute phase proteins such as high sensitivity
C-reactive protein.12,13 According to this fact, local hemo-
dynamic factors such as vascular inflammation and/or
endothelial damage associated to malnutrition status could
play a role in the formation of VAS.14,15
However, none of the current biochemical markers
seems to accurately reflect nutritional status in HD
patients,16 considering that 20%–60% of them are mal-
nourished,17,18 and these biomarkers may reflect rather
acute diseases such as inflammations and/or infections.
Due to this fact, various clinical scores such as the
Subjective Global Assessment (SGA) and the Malnutrition
Inflammation Score (MIS) have been progressively intro-
duced into the current clinical practice19 in order to fill the
gap left by the biohumoral markers. According to the evi-
dence provided by the Dialysis Outcomes and Practice
Patterns Study (DOPPS), malnourished HD patients esti-
mated by both scores SGA and MIS had greater odds to
have tunneled catheter rather than well-functioning AVF as
vascular access, underlying how malnutrition can affect vas-
cular access type.20,21 In this setting, our aim was to assess
the Subjective Global Assessment–Dialysis Malnutrition
Score (SGA-DMS), derived by prior SGA, as an independ-
ent predictor of AVF failure (VAS and/or VAT) across a
2-year interval in chronic HD patients. SGA-DMS is a semi-
quantitative index that consists of seven items in order to
evaluate malnutrition in HD patients,22 see also Table 1.
SGA-DMS is a simple and cost-free tool that ranges between
7 (normal) and 35 (severely malnourished) that can be eas-
ily performed by both medical and paramedical staff.
We also compared its own impact for vascular access sur-
vival with a well-known and commonly used comorbidity
score, the Charlson Comorbidity Index (CCI), see Table 2,
and with the modified CCI (mCCI), without considering age
factor. The CCI is a commonly used scale for assessing mor-
bidity, where different diseases are considered, and for each
decade, >40 years of age, a score of 1 is added.23
72 The Journal of Vascular Access 20(1)

Table 2. CCI score.

Score Condition
1 •• Myocardial infarction (history, not ECG only)
•• Congestive heart failure
•• Peripheral disease (including aortic aneurysm ≥6)
•• Cerebrovascular disease: CVA with mild or no residual or TIA
•• Dementia
•• Chronic pulmonary disease
•• Connective tissue disease
•• Peptic ulcer disease
•• Mild liver disease (without portal hypertension, includes chronic hepatitis)
•• Diabetes without end-organ damage (excludes diet-controlled alone)
2 •• Hemiplegia
•• Moderate or severe renal disease
•• Diabetes with end-organ damage (retinopathy, neuropathy, nephropathy, or brittle diabetes)
•• Tumor without metastasis (exclude if >5 years from diagnosis)
•• Leukemia(acute or chronic)
•• Lymphoma
3 •• Moderate or severe liver disease
6 •• Metastatic solid tumor
•• AIDS (not just being HIV positive)

ECG: electrocardiogram; CVA: cerebrovascular accident; TIA: transient ischemic attack; AIDS: acquired immunodeficiency syndrome; HIV: human im-
munodeficiency virus.
For each decade, >40 years, a score of 1 is added to the total score.23

Table 3. Clinical, biohumoral, and demographic variables of the 57 enrolled HD patients.

Clinical data HD patients VAS- and/or VAT-positive VAS- and/or VAT-negative P value
(n = 57) patients (n = 12) patients (n = 45)
Age (years) 68.5 ± 14.5 75.5 ± 10.24 67.24 ± 14.5 0.03
Gender % (M/F) 73/27 63/27 75/25 0.46
DV (years) 4.5 ± 4.11 3.82 ± 3.25 4.78 ± 4.3 0.42
Ht (%) 32.3 ± 3.65 32.7 ± 4 32.24 ± 3.6 0.73
Albumin (gr/dL) 3.67 ± 0.25 3.69 ± 0.26 3.67 ± 0.24 0.78
nPCR (gr/Kg/day) 1.86 ± 0.5 1.79 ± 0.4 1.90 ± 0.53 0.44
Kt/V 1.35 ± 0.41 1.34 ± 0.31 1.35 ± 0.33 0.96
P (mEq/L) 1.59 ± 0.52 1.77 ± 0.77 1.55 ± 0.44 0.38
CRP (mg/dL) 1 ± 2.04 1.77 ± 3.21 0.81 ± 1.64 0.36
Diabetes % (Y/N) 35/65 27/73 35/65 0.23
Hypertension % (Y/N) 71/29 72/28 73/27 0.38
Previous CV events % (Y/N) 37/63 50/50 36/64 0.39
SGA-DMS 10.6 ± 2.7 11.82 ± 2.63 10.31 ± 2.74 0.11
CCI 6.5 ± 2.2 7.45 ± 2.5 6.33 ± 2.15 0.19
mCCI 3.95 ± 1.40 4.18 ± 1.66 3.87 ± 1.39 0.57

Y/N: yes/no; DV: dialysis vintage; Ht: hematocrit; nPCR: normalized protein catabolic rate; P: phosphoremia; CRP: C-reactive protein; HD: hemodi-
alysis; SGA-DMS: Subjective Global Assessment–Dialysis Malnutrition Score; VAT: vascular access thrombosis; VAS: vascular access stenosis; CCI:
Charlson Comorbidity Index; mCCI: modified Charlson Comorbidity Index; CV: cerebrovascular.
Data are expressed as mean ± standard deviation or as percentage.

Materials and methods
Patients and study design
This is a 2-year observational cohort study involving 57
maintenance HD patients followed at the Hemodialysis
Unit of Humanitas Hospital, Rozzano (MI), Italy. All the
clinical and biohumoral data are reported in Table 3.
Inclusion criteria were adults aged ≥18 years with
autologous AVF, HD treatment for almost 3 months, and
AVF in good status without complications for at least
3 months since the first use.
Exclusion criteria were prosthetic graft fistulas, patients
with inherited and/or acquired prothrombotic diseases;
treatment with oral anticoagulation therapy; acute
Spatola et al. 73

Table 4. Linear regression model with SGA-DMS as thrill, and also unexplained persistent decreased Kt/V
dependent variable. (>0.2 units).
Data B (95% CI) p value VAT was diagnosed upon the physical examination report
of the lack of thrill at the palpation and/or auscultation.
Albumin –1.04 (–4.07; 1.98) 0.42
nPCR –0.87 (–1.15; 1.34) 0.94
Ht –0.05 (–0.25; 1.14) 0.52 Statistical analysis
Kt/V 1.56 (–0.77; 3.89) 0.18 All variables are expressed as mean (standard deviation
P –0.06 (–1.78; 1.08) 0.62 (SD)) unless otherwise indicated. Statistical significance
CRP 0.06 (–0.28; 0.45) 0.63
was set at the level of p < 0.05.
Age 0.06 (0.01;0.11) 0.02
Normal distribution of the variables was assessed using
Diabetes –0.63 (–2.18; 0.92) 0.41
Kolmogorov–Smirnov test. Student’s t test and Fisher test
Hypertension –0.17 (–2.70; 0.55) 0.19
were used to assess clinical and biohumoral data.
CCI 0.31 (0.01; 0.63) 0.05
mCCI 0.34 (–1.16; 0.85) 0.17
Linear regression analysis was assessed between the
continuous variables tested.
nPCR: normalized protein catabolic rate. Ht: hematocrit; P: phosphore- Logistic regression models were used to evaluate con-
mia; CRP: C-reactive protein; SGA-DMS: Subjective Global Assessment– tinuous and categorical variables as risk predictors of AVF
Dialysis Malnutrition Score; CCI: Charlson Comorbidity Index; mCCI:
modified Charlson Comorbidity Index
failure.
Both age and CCI were significantly related to SGA-DMS, B = 0.06 Cox regression models were used to evaluate the hazard
(95% CI = 0.01;0.11), p 0.02 and B = 0.31 (95% CI = 0.01; 0.63), p = ratios of the clinical and biohumoral data. Receiver operat-
0.05, respectively. ing characteristic curve (ROC) analysis was performed in
order to compare the area under the curve (AUC) of SGA-
i­nflammation and/or infections at the beginning of the DMS, CCI, and mCCI.
observation in June 2015. Kaplan–Meier curves for AVF failure (VAS or VAT)
Informed consent was obtained from each patient were performed for three different groups of SGA-DMS
before inclusion, and the study was approved by the local according to score value—Group 1: n = 21 patients (SGA-
Ethical Committee. DMS: 7–9); group 2: n = 26 patients (SGA-DMS: 10–12);
Blood samples were collected before starting the first Group 3: n = 10 patients (SGA-DMS: ≥13).
HD session of the third week of June 2015. The nPCR was All statistical analyses were performed using Statistical
calculated from pre-dialysis and post-dialysis blood urea Package for the Social Sciences (SPSS, Inc., Chicago, IL,
nitrogen (BUN) measurements in patients receiving inter- USA) version 20.0.
mittent dialysis according to the two BUN, single pool,
and variable-volume model.5 SGA-DMS, CCI, and mCCI
were all calculated according to the clinical data at the end
Results
of June 2015 by the same nephrologist with more than a The 21% of our cohort HD patients (n = 12) developed
3-year experience in HD Unit, in order to avoid inter-­ VAS or VAT across the 2-year interval, and age was the
operators variability. From our HD cohort that in June 2015 only parameter that differed significantly between the two
included 104 patients, we evaluated only those with native subgroups, see Table 3.
and well-functioning AVF, that were 60 patients (57.7%), Age and CCI were positively related to SGA-DMS—B
but we excluded three patients because they started to use = 0.06 (95% CI = 0.01; 0.11), p = 0.02 and B = 0.31 (95%
AVF less than 3 months since the start of observational CI = 0.01; 0.63), p = 0.05, respectively—at the univariate
analysis and it did not meet our criteria, in order to avoid linear regression analysis, see Table 4. Other clinical and
complications due to beginning of cannulation. biohumoral variables including malnutrition biomarkers
Patients were evaluated for the combined AVF out- albumin and nPCR were not related to SGA-DMS.
come (VAS or VAT) across a 2-year follow-up from 30 Univariate logistic regression analysis reported that both
June 2015 to 30 June 2017. No one of the patients passed higher albumin and nPCR levels had a protective role
away during the observational analysis. We considered against AVF failure, 0.67 (0.56; 0.81), p = 0.00 and 0.46
VAS the angiographic evidence of greater than 50% ste- (0.32; 0.67), p = 0.00, respectively, see Table 5. Age and
nosis of the vascular access and indications for angiogra- diabetes were not related to AVF failure, p = 0.81 and p =
phy referral were consistent with those of the 2006 0.24, respectively. Cox regression analysis showed that
Kidney Disease Outcomes Quality Initiative (KDOQI) both SGA-DMS and CCI were significant risk factors for
work group guidelines.24 Thus, the indications included: AVF failure in all models: HR = 1.23 (95% CI = 1.01;
physical findings suggestive of stenosis including persis- 1.52), p = 0.05 and HR = 1.38 (95% CI = 1.01; 1.89),
tent arm swelling, prolonged bleeding after needle with- p = 0.04, respectively, in Model A; HR = 1.42 (95%
drawal, collateral veins, altered features of the pulse or CI = 1.04; 1.92), p = 0.02 and HR = 1.48 (95% CI = 1.01;
74
Table 5. Logistic regression model with AVF failure (VAS and/
or VAT) as dependent variable.
Data OR (95% CI) p value
Albumin 0.67 (0.56; 0.81) 0.00
nPCR 0.46 (0.32; 0.67) 0.00
Ht 1.03 (0.86; 1.23) 0.73
Kt/V 0.99 (0.11; 8.30) 0.99
P 2.10 (0.63; 7.00) 0.22
PCR 1.19 (0.19; 1.51) 0.20
Age 1.05 (0.99; 1.12) 0.81
Diabetes 1.34 (0.07; 1.96) 0.24
Hypertension 1.59 (0.39; 6.54) 0.52
nPCR: normalized protein catabolic rate. Ht: hematocrit; P: phosphore-
mia; CRP: C-reactive protein; AVF: arteriovenous fistula; VAT: vascular
access thrombosis; VAS: vascular access stenosis; PCR: protein catabolic
rate.
Both higher Albumin and nPCR values had a protective role against
AVF failure: OR = 0.67 (95% CI = 0.56; 0.81), p = 0.00 and OR = 0.46
(95% CI = 0.32; 0.67), respectively.
2.17), p = 0.04, respectively, in adjusted full Model B, see
Table 6. Differently, mCCI was not related to AVF Failure
in all models (p = 0.12 and p = 0.10). ROC analysis showed
that AUC of SGA-DMS was significantly higher than
those of both CCI and mCCI: 0.70 (95% CI = 0.50; 0.88),
p = 0.05 versus 0.61 (95% CI = 0.41; 0.80), p 0.25 and 0.55
(95 CI% = 0.41; 0.70) p = 0.57, see Figure 1. The best
combination of sensitivity and specificity was found for
SGA-DMS value ≥10.5 with sensitivity 72% and specific-
ity 78%. Kaplan–Meier curves showed that HD patients
with higher SGA-DMS scores as Group 2 (SGA-DMS:
10–12) and Group 3 (SGA-DMS: ≥13) had a worse prog-
nosis on vascular access survival compared to Group 1
(SGA-DMS: 7–9), see Figure 2.
Discussion
Malnutrition biomarkers have been previously related to
poor vascular access outcome in HD patients with native
AVFs.8,25 In fact, Gagliardi et al.8 measured levels of albu-
min, nPCR and total cholesterol in 91 HD patients and
pointed out that decreased nPCR values as well as raised
cholesterol levels increased the risk of AVF failure.
Similarly, Premuzic et al.25 reported that hypoproteinemia
<65 g/L was an independent prognostic marker for AVF
failure at 2 years in 497 HD patients. In logistic regression,
all patients with serum proteins value >65 g/L had adjusted
odds ratio for survival of 5.10 (95% CI = 4.23; 5.92). Our
results in the logistic regression analysis agree with previ-
ous studies when considering both malnutrition biomark-
ers albumin and nPCR.
Malnutrition is reported to promote VAS and/or VAT
through different mechanisms: hypoproteinemia, as a
result of inflammation which decreases the synthesis of
proteins by the liver, is a risk factor for prolonged bleeding
The Journal of Vascular Access 20(1)
Table 6. Cox regression models with AVF failure as
dependent variable. Model A includes dialysis vintage,
gender, diabetes, and hypertension. Model B includes Model
A + albumin, nPCR, Kt/V, P, Ht, and CRP.
Data Model A Model B
HR (95% CI), p value HR (95% CI), p value
SGA-DMS 1.23 (1.01; 1.52), 1.42 (1.04; 1.92),
p = 0.05 p = 0.02
CCI 1.38 (1.01; 1.89), 1.48 (1.01; 2.17),
p = 0.04 p = 0.04
mCCI 1.40 (0.90; 2.16), 1.48 (0.95; 2.33),
p = 0.12 p = 0.10
nPCR: normalized protein catabolic rate. Ht: hematocrit; P: phosphore-
mia; CRP: C-reactive protein; SGA-DMS: Subjective Global Assessment–
Dialysis Malnutrition Score; CCI: Charlson Comorbidity Index; mCCI:
modified Charlson Comorbidity Index; AVF: arteriovenous fistula.
Both higher SGA-DMS and CCI scores are negative prognostic impact
in all models.
and consequently inflammation promotes up-­regulation of
procoagulants and inhibition of fibrinolysis leading to a
prothrombotic state.26 In addition, hypoalbuminemia pre-
disposes to stasis,9 which is one of the Virchow triad factors,
along with endothelial cell injury and hypercoagulability,
that leads to VAT. Inflammation-related malnutrition may
also promote endothelial cell injury that could promote
VAT.26
In our analysis, patients who develop VAS or VAT were
significantly older compared to those who did not, how-
ever, both age and diabetes factors were not positively
related to AVF failure and currently, older age is not con-
sidered a negative prognostic index of AVF survival as
also reported by Kidney Disease: Improving Global
Outcomes (KDIGO) guidelines27 and other evidence-
based literature.28,29 On the contrary, diabetes has been
reported as an independent risk factor for poor VA sur-
vival9,25 because of the endothelial cell injury and raised
oxidative stress. Probably, due to the small number of HD
evaluated and the monocentric effect (only 35% was
affected by diabetes), we were not able to identify a rela-
tion with this complication, even if we detected a positive
association between CCI and AVF failure and diabetes
with end-organ damage (retinopathy, neuropathy, nephrop-
athy, or brittle diabetes) is scored two points in CCI
assessment.
Our analysis involved a prevalent group of maintenance
HD patients with a dialysis vintage (DV) of 4.5, instead of
incident HD patients, because the impact of factors such as
the surgical technique, size and type of anastomosis, size
of vessels selected, and suture materials could have influ-
enced the AVF outcome30 more than malnutrition. The pre-
vious cited factors are well recognized as prognostic
markers of AVF failure as well as a defect in AVF matura-
tion, early cannulation and hematomas due to first cannu-
lation. According to this fact, we preferred to consider,
Spatola et al. 75

Figure 1. AUCs of SGA-DMS, CCI, and mCCI. AUC of SGA-DMS was significantly higher than those of both CCI and mCCI:
0.70 (95% CI = 0.50; 0.88), p = 0.05 versus 0.61 (95% CI = 0.41; 0.80), p = 0.25 and 0.55 (95CI% = 0.41; 0.70), p = 0.57. The best
combination of sensitivity and specificity was found for SGA-DMS value ≥10.5 with sensitivity 72% and specificity 78%.

Figure 2. Kaplan–Meier curves for different groups of SGA-DMS—Group 1: n = 21 patients with SGA-DMS 7–9; Group 2: n = 26
patients with SGA-DMS 10–12; Group 3: n = 10 patients with SGA-DMS ≥13. Log-rank 4.02; p = 0.04.
76
according to the inclusion criteria, only patients with regu-
lar use of AVF in good status for at least 3 months.
Of note, in our study, both SGA-DMS and CCI were
related to poor AVF outcome, and this result has not previ-
ously been reported.
In the past years, SGA-DMS has been shown to be a
reliable malnutrition clinical marker by Janarrdhan et al.22
who pointed out a significant negative correlation between
SGA-DMS and anthropometric measures such as triceps
skin fold thickness, mid arm circumference, and biomark-
ers like albumin, transferring, and ferritin. In this contest,
SGA-DMS was evaluated as a simple and useful index to
identify malnourished HD patients.
In the Janahardan study age, DV and total cholesterol
did not correlate with SGA-DMS, and unfortunately, vas-
cular access survival was not evaluated as an outcome.
On the contrary, in our experience, age was positively
related to SGA-DMS. This relation can be explained due
to the fact that elderly patients have functional impairment
and are prone to protein energy wasting because of the
limitations in food gathering, cooking, and poor appetite,
frequent chronic or acute illnesses, reduced mobility, and
cognition.31 It is important to consider that, unlike former
SGA, SGA-DMS also assesses comorbidity factor and the
DV is added to the comorbidity component, so this index
could appraise both malnutrition and comorbidity at the
same time with a single score. Prior SGA is a 7-point scale
consisting of six components three based on patients’ his-
tory of weight loss, dietary intake, and gastrointestinal
symptoms and three based on the physician’s grading of
muscle wasting, presence of edema, and loss of subcutane-
ous fat.32,33 Patients are subjectively scored from 1
(“severely malnourished”) to 7 (“well nourished”). A fair
intra- and inter-rater reliability and validity of SGA have
been demonstrated in end stage renal disease popula-
tion.32,34 In addition, the role of SGA as a clinical marker
of nutritional status has been clearly demonstrated32 as
well as its prognostic role in all-cause mortality in HD
patients.35,36 Unfortunately, the prognostic weight of both
SGA and SGA-DMS in predicting vascular access failure
has not been evaluated before, being mostly analyzed as
malnutrition and all-cause mortality clinical markers. Due
to this fact, a comparison is still lacking, and new evi-
dences are needed to better assess the role of malnutrition
clinical scores in AVF outcome.
SGA has been related to chronic inflammation in several
studies,37,38 underlying how the pathogenesis of malnutri-
tion in HD patients is multifactorial including anorexia,
altered taste sensation, intercurrent illness, emotional dis-
tress or illness, impaired ability to prepare food, hyperca-
tabolism, reduced anabolism, and also chronic inflammatory
state. Both Kizil et al.37 and Essadik et al.38 reported a posi-
tive relation between SGA and CRP: r = 0.24, p < 0.05 and
r = 0.65, p = 0.00, respectively, in maintenance HD patients
without intercurrent acute inflammatory disease. In this
The Journal of Vascular Access 20(1)
setting, the SGA score could reflect the severity of malnu-
trition better than low serum albumin levels due to high
prevalence from 35% to 75% of malnutrition recognized
by using SGA score.38 SGA in its first formulation by
Detsky et al.39 and in 7-point SGA scale38 did not include
the comorbidity factor and the DV, whereas the SGA-DMS
assesses this parameter because the element of time on HD
may have an impact on the degree of malnutrition and
inflammation as well as severe comorbidities such as
severe coronary artery diseases and congestive heart fail-
ure Class III and IV. However, former SGA has been asso-
ciated not only with all-cause mortality in HD patients but
also with frequency and duration of hospitalization.35
Indeed, HD patients with worse SGA score B and C
showed significantly more frequent and longer hospitali-
zation due to AVF failure or catheter events, cardiovascu-
lar diseases such as acute coronary artery syndrome and
peripheral arterial occlusive disease.
To the best of our knowledge, CCI, that is an already
known prognostic marker for both all-cause and cardio-
vascular mortality in HD23,40 has not previously evaluated
as tool for VAT and/or VAS and our findings support the
thesis that comorbidity status should be assessed when
determining eligibility for AVF creation.29
In fact, comorbidity status appraisal allows for life
expectancy, health status, and quality of life to be esti-
mated more completely than age alone and/or single
comorbidity in order to reach a better decision concerning
preferred vascular access. This evaluation should be car-
ried out especially in elderly patients according to the ris-
ing amount of incident HD patients over 65 years old in the
general population.41 In this setting, both SGA-DMS and
CCI could be proposed as valid tools to integrate in the
decision-making process of AVF creation.
Acknowledgement
All authors thank Professor P. Taylor for the English revision.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect
to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research,
authorship, and/or publication of this article.
References
1. Winkelmayer WC. Tackling the Achilles’ heel of hemodialy-
sis. N Engl J Med 2011; 364(4): 372–374.
2. Murad MH, Elamin MB, Sidawy AN, et al. Autogenous
versus prosthetic vascular access for hemodialysis: a sys-
tematic review and meta-analysis. J Vasc Surg 2008; 48(5,
Suppl.): 34S–47S.
3. National Kidney Foundation-Dialysis Outcomes Quality
Initiative. NKF-DOQI clinical practice guidelines for
Spatola et al.
vascular access. Am J Kidney Dis 1997; 30(4, Suppl. 3):
S150–S191.
4. Papanikolaou V, Papagiannis A, Vrochides D, et al. The natu-
ral history of vascular access for hemodialysis: a single center
study of 2,422 patients. Surgery 2009; 145(3): 272–279.
5. Huber TS, Carter JW, Carter RL, et al. Patency of autog-
enous and polytetrafluoroethylene upper extremity arterio-
venous hemodialysis accesses: a systematic review. J Vasc
Surg 2003; 38(5): 1005–1011.
6. Roy-Chaudhury P, Kelly BS, Zhang J, et al. Hemodialysis
vascular access dysfunction: from pathophysiology to novel
therapies. Blood Purif 2003; 21: 99–110.
7. Chang CJ, Ko YS, Ko PJ, et al. Thrombosed arteriovenous
fistula for hemodialysis access is characterized by a marked
inflammatory activity. Kidney Int 2005; 68: 1312–1319.
8. Gagliardi GM, Rossi S, Condino F, et al. Malnutrition,
infection and arteriovenous fistula failure: is there a link? J
Vasc Access 2011; 12(1): 57–62.
9. Montagnana M, Meschi T, Borghi L, et al. Thrombosis
and occlusion of vascular access in hemodialyzed patients.
Semin Thromb Hemost 2011; 37(8): 946–954.
10. Miller PE, Carlton D, Deierhoi MH, et al. Natural history of
arteriovenous grafts in hemodialysis patients. Am J Kidney
Dis 2000; 36(1): 68–74.
11. Ocak G, Rotmans JI, Vossen CY, et al. Type of arterio-
venous vascular access and association with patency and
mortality. BMC Nephrol 2013; 14: 79.
12. Kuo WH, Lee YT, Ng HY, et al. C-reactive protein vari-
ability is associated with vascular access outcome in hemo-
dialysis patients. J Clin Lab Anal 2018; 32(1): e22213.
13. Alves FC, Sun J, Qureshi AR, et al. The higher mortality
associated with low serum albumin is dependent on sys-
temic inflammation in end-stage kidney disease. PLoS ONE
2018; 13(1): e0190410.
14. Chang CJ, Ko PJ, Hsu LA, et al. Highly increased cell pro-
liferation activity in the restenotic hemodialysis vascular
access after percutaneous transluminal angioplasty: impli-
cation in prevention of restenosis. Am J Kidney Dis 2004;
43: 74–84.
15. Rotmans JI, Pasterkamp G, Verhagen HJ, et al. Hemodialysis
access graft failure: time to revisit an unmet clinical need? J
Nephrol 2005; 18: 9–20.
16. De Mutsert R, Grootendorst DC, Indemans F, et al.
Association between serum albumin and mortality in dialy-
sis patients is partly explained by inflammation, and not by
malnutrition. J Ren Nutr 2009; 19: 127–135.
17. Acchiardo SE, Moore LW and Latour PA. Malnutrition as
the main factor in morbidity and mortality of haemodialysis
patients. Kidney Int 1983; 24: S199–S203.
18. Dwyer JT, Cunniff PJ, Maroni BJ, et al. The Hemodialysis
(HEMO) pilot study: nutrition program and participant
characteristics at baseline. J Renal Nutr 1998; 8: 11–20.
19. Kondrup J, Allison SP, Elia M, et al. Educational and
Clinical Practice Committee, European Society of Parenteral
and Enteral Nutrition (ESPEN). ESPEN Guidelines for
Nutrition Screening 2002. Clin Nutr 2003; 22: 415–421.
20. Combe C, McCullough KP, Asano Y, et al. Kidney Disease
Outcomes Quality Initiative (K/DOQI) and the Dialysis
Outcomes and Practice Patterns Study (DOPPS): nutrition
77
guidelines, indicators, and practices. Am J Kidney Dis 2004;
44(5, Suppl. 2): 39–46.
21. Combe C, Bragg-Gresham JL, Pifer TB, et al. Validation
of a Malnutrition-Inflammation Score (MIS) to pre-
dict mortality and hospitalization in the DOPPS. In:
Proceedings of the XLI congress of the European renal
association–European dialysis and transplant associa-
tion, Lisbon, 15–18 May 2004.
22. Janardhan V, Soundararajan P, Rani NV, et al. Prediction of
malnutrition using modified subjective global assessment-
dialysis malnutrition score in patients on hemodialysis.
Indian J Pharm Sci 2011; 73(1): 38–45.
23. Beddhu S, Bruns FJ, Saul M, et al. A simple comorbid-
ity scale predicts clinical outcomes and costs in dialysis
patients. Am J Med 2000; 108: 609–613.
24. Hemodialysis Adequacy 2006 Work Group. Clinical prac-
tice guidelines for hemodialysis adequacy, update 2006. Am
J Kidney Dis 2006; 48(Suppl. 1): S2–S90.
25. Premuzic V, Hudolin T, Pasini J, et al. Hypoproteinemia
as a prognostic risk factor for arteriovenous fistula failure.
Hemodial Int 2018; 22: 37–44.
26. Sung SA, Ko GJ, Jo SK, et al. Interleukin-10 and tumor
necrosis factor-α polymorphisms in vascular access failure
in patients on hemodialysis: preliminary data in Korea. J
Korean Med Sci 2008; 23: 89–93.
27. Kidney Disease: Improving Global Outcomes (KDIGO)
CKD Work Group. KDIGO clinical practice guidelines for
the evaluation and management of chronic kidney disease.
Kidney Int Suppl 2013; 3: 1–150.
28. Vachharajani TJ, Moossavi S, Jordan JR, et al. Re-evaluating
the fistula first initiative in octogenarians on hemodialysis.
Clin J Am Soc Nephrol 2011; 6: 1663–1667.
29. Lomonte C, Forneris G, Gallieni M, et al. The vascular
access in the elderly: a position statement of the Vascular
Access Working Group of the Italian Society of Nephrology.
J Nephrol 2016; 29(2): 175–184.
30. Bashar K, Conlon PJ, Kheirelseid EA, et al. Arteriovenous
fistula in dialysis patients: factors implicated in early and
late AVF maturation failure. Surgeon 2016; 14(5): 294–
300.
31. Wolfson M. Nutrition in elderly dialysis patients. Semin
Dial 2002; 15(2): 113–115.
32. Steiber A, Leon JB, Secker D, et al. Multicenter study of the
validity and reliability of subjective global assessment in the
hemodialysis population. J Ren Nutr 2007; 17(5): 336–342.
33. Enia G, Sicuso C, Alati G, et al. Subjective global assess-
ment of nutrition in dialysis patients. Nephrol Dial
Transplant 1993; 8(10): 1094–1098.
34. Cooper BA, Bartlett LH, Aslani A, et al. Validity of subjec-
tive global assessment as a nutritional marker in end-stage
renal disease. Am J Kidney Dis 2002; 40(1): 126–132.
35. Fiedler R, Jehle PM, Osten B, et al. Clinical nutrition scores
are superior for the prognosis of haemodialysis patients
compared to lab markers and bioelectrical impedance.
Nephrol Dial Transplant 2009; 24(12): 3812–3817.
36. Santin FG, Bigogno FG, Dias Rodrigues JC, et al.
Concurrent and predictive validity of composite methods to
assess nutritional status in older adults on hemodialysis. J
Ren Nutr 2016; 26(1): 18–25.
78
37. Kizil M, Tengilimoglu-Metin MM, Gumus D, et al. Dietary
inflammatory index is associated with serum C-reactive pro-
tein and protein energy wasting in hemodialysis patients: a
cross-sectional study. Nutr Res Pract 2016; 10(4): 404–410.
38. Essadik R, Msaad R, Lebrazi H, et al. Assessing the preva-
lence of protein-energy wasting in haemodialysis patients:
a cross-sectional monocentric study. Nephrol Ther 2017;
13(7): 537–543.
39. Detsky AS, Baker JP, Mendelson RA, et al. Evaluating
the accuracy of nutritional assessment techniques
The Journal of Vascular Access 20(1)
applied to hospitalized patients: methodology and com-
parisons. JPEN J Parenter Enteral Nutr 1984; 8(2):
153–159.
40. Chae JW, Song CS, Kim H, et al. Prediction of mortality in
patients undergoing maintenance hemodialysis by Charlson
Comorbidity Index using ICD-10 database. Nephron Clin
Pract 2011; 117(4): c379–c384.
41. Lok C and Foley R. Vascular access morbidity and mortal-
ity: trend of the last decade. Clin J Am Soc Nephrol 2013; 8:
1213–1219.