Abstract
Introduction
Nutritional risk assessment, which is necessary for initiating appropriate nutritional therapy to improve prognosis, should be as simple as possible to perform, particularly when direct patient contact is limited. We examined whether simplified Subjective Global Assessment (SGA), a modified SGA method omitting physical examination components, is associated with prognosis in patients with kidney failure starting hemodialysis (HD).
Methods
Retrospectively, we reviewed the medical records of patients who initiated HD between 2016 and 2018. Based on these records, two dialysis nurses independently assessed each patient’s nutritional status using simplified SGA and categorized them as well-nourished (W) or malnourished (M). Patients were classified into 3 groups (W-W, W-M, and M-M) according to the combination of the 2 assessments. The primary composite outcome was death or hospitalization within 2 years.
Results
A total of 138 patients starting HD (mean age 66.8 years; 70% male) were examined. According to simplified SGA, 97 (70%), 25 (18%), and 16 (12%) patients were categorized respectively as W-W, W-M, and M-M. Kaplan-Meier analysis showed significantly lower survival and hospitalization-free rates in the M-M group (p < 0.001 and = 0.024, respectively). Cox proportional hazards analysis demonstrated significant association between the M-M category and the primary outcome (hazard ratio 3.34, 95% confidence interval 1.52–7.36).
Conclusion
Simplified SGA scores might be associated with 2-year adverse outcomes in patients starting HD.
Keywords: Dialysis nurse, Hemodialysis, Hospitalization, Mortality, Subjective global assessment
Plain Language Summary
Good nutrition is important for people starting hemodialysis because it affects their health and survival. However, some common tools for checking nutrition are complicated and require a physical exam. In situations where physical examination is difficult – such as during infection control measures or when medical resources are limited – it can be hard to perform these evaluations. To make this process easier, we tested a simple method called the simplified Subjective Global Assessment (SGA). This method uses only information from the patient’s medical records, without a physical exam. We looked at 138 people who started hemodialysis between 2016 and 2018. Two dialysis nurses independently rated each patient’s nutrition as either “well-nourished” or “malnourished” using the simplified SGA. Patients were then grouped as follows: both nurses said “well-nourished” (W-W), one said “well-nourished” and the other “malnourished” (W-M), or both said “malnourished” (M-M). We followed these patients for 2 years to see who died or was hospitalized. Patients in the M-M group had a much higher chance of death or hospitalization than those in the other 2 groups. This link remained even after considering other factors such as age and laboratory results. Our findings suggest that the simplified SGA, which is easy to perform and does not require physical examination, may be a practical and reliable way to identify high-risk patients, especially in settings where direct patient contact or medical resources are limited.
Introduction
Malnutrition, which is prevalent in 28–54% of patients administered hemodialysis (HD), presents severe health difficulties that increase risks of complications such as infection, cardiovascular disease (CVD), and frailty, leading to impaired quality of life and worse patient survival [1–6]. Early detection and intervention for malnutrition reportedly promote timely nutritional recovery and might contribute to reduced complications, hospitalization, healthcare costs, and mortality [7–9].
As the first step for managing malnutrition, nutritional risk assessment using the appropriate tools at the right time is crucially important. For patients either starting or undergoing HD, attempts have been undertaken to identify malnutrition using existing or modified nutritional assessment tools and newly developed tools. Their clinical significance has been reported. Evaluation of nutritional status using the Subjective Global Assessment (SGA), which comprises 5 medical history and 5 physical examination findings items without laboratory testing, can predict poor outcomes related to malnutrition in patients on HD [10]. Malnutrition assessed using 7-point SGA, a quantitative version of SGA, was also reported to have prognostic value for patients receiving kidney replacement therapy including HD [11, 12]. Results of assessment obtained using the Nutritional Risk Index for Japanese Hemodialysis Patients (NRI-JH), which was developed based on nationwide registry data and calculated from the objective measurements including laboratory data such as serum albumin (Alb), reveal factors associated with mortality in patients on maintenance HD [13]. However, these tools might present several concerns related to their clinical use for patients on HD. The concerns include their time-consuming nature because of their numerous assessment items, suitability for use in situations where obtaining physical findings is difficult, such as during a pandemic, and incapability of performing assessment when measurement tools or data are unavailable. It is preferable to use nutrition assessment tools that rely on a few readily available and simple assessment items and methods.
For this study, we investigated whether simplified SGA, a modified method of SGA that omits physical examination information and evaluates based on medical history, can be associated with adverse outcomes, hospitalization or death within 2 years, in patients starting HD. We also examined whether a relation of the simplified SGA to the outcomes is identified when using the NRI-JH and the Geriatric Nutritional Risk Index (GNRI). Thereby, we clarify differences between simplified SGA and these existing nutritional assessment tools [14].
Methods
Study Design and Patients
We conducted a retrospective cohort study of adult patients initiating maintenance HD in a tertiary care hospital between January 1, 2016, and December 31, 2018. According to our usual clinical practice policy, all patients included in this study started conventional HD under inpatient care. They went to other outpatient HD clinics to continue regular HD treatment after discharge from our hospital. Alternatively, when discharge to home was difficult, they were transferred to another hospital. Exclusion criteria were the following: (1) patients hospitalized for more than 3 months before starting HD (N = 3), (2) patients lost to follow-up within 3 months after HD initiation (N = 6), (3) patients who had only 1 HD session (N = 2), and (4) patients who needed HD temporarily but who were weaned off it successfully (N = 5).
This study was conducted in accordance with the Declaration of Helsinki. The study protocol was reviewed and approved by the Research Ethics Committee of the Faculty of Medicine of the University of Tokyo (Approval No. 2269, December 22, 2008). Because of the retrospective observational characteristics of the study, written informed consent was not required from adult participants; instead, an opt-out consent process was adopted. The opt-out informed consent protocol for the use and collection of participant data for research purposes was approved by the same Ethics Committee. After we provided participants with information of the study (purpose, method, required data, and duration) on the website of our department, we described to each the opportunity for opting out.
Data Collection
Basic information of the patients was obtained from their medical records: age, gender, cause of kidney failure, planned HD initiation, and history of CVD. Clinical and laboratory data before HD initiation were also obtained from medical records: body mass index, systolic and diastolic blood pressure, hemoglobin (Hb), Alb, blood urea nitrogen, serum creatinine (Cr), estimated glomerular filtration rate (eGFR), uric acid, corrected calcium, inorganic phosphate, C-reactive protein (CRP), total cholesterol, triglyceride, and HbA1c.
Evaluation of Nutritional Status at HD Initiation
Each patient’s nutritional status was assessed using simplified SGA, a modified SGA method omitting information obtained from physical examination findings [10]. The simplified SGA comprises 5 items derived from the standard SGA: weight change, changes in dietary intake (compared to usual), gastrointestinal symptoms (lasting more than 2 weeks), functional impairment, and disease related to nutritional requirements [10]. Each component was assessed according to the method of the original SGA described by Detsky et al. [10]. As with the original SGA, simplified SGA set no strict threshold or score for the results of each assessment item or for the determination of malnutrition. For example, regarding body weight change, both the absolute and proportionate changes from baseline weight during the preceding 6 months were evaluated. The rate of weight loss was classified into 3 categories (<5%, 5–10%, and >10%), but patterns of their changes over time were also considered. Recent stabilization or gain was regarded as indicative of better nutritional status. Nutritional status judgment (well-nourished, moderate malnutrition, and severe malnutrition) was performed based on the examiner’s comprehensive evaluation of the results obtained from all items in simplified SGA. A list of evaluation items included in the simplified SGA is presented as online supplementary Table S1 (for all online suppl. material, see https://doi.org/10.1159/000553346).
Based on nursing records kept within 5 days after HD initiation, two dialysis nurses with more than 10 years of experience independently assessed the nutritional status of each patient. According to the results of simplified SGA from the 2 nurses, patients were categorized into 3 groups: W-W when both nurses assessed the patient as well-nourished, W-M when the patient was assessed by 1 nurse as well-nourished and by the other as moderately to severely malnourished, and M-M when both nurses assessed the patient as moderately to severely malnourished. Existing nutritional assessment indices, GNRI and NRI-JH, were also evaluated using the data at HD initiation [13, 14].
Outcomes
The primary outcome was a composite of all-cause mortality and hospitalization because of complications within 2 years after HD initiation. The secondary outcome was all-cause mortality within 2 years after HD initiation. The occurrence and date of the first observed outcomes were investigated after the start of HD. A researcher other than the 2 nurses evaluating simplified SGA was responsible for collecting prognostic data through medical records and telephone surveys. Patient survival was censored at kidney transplantation or on the date of last follow-up.
Statistical Analysis
All statistical analyses were conducted using software (BellCurve for Excel; Social Survey Research Information Co., Ltd., Tokyo, Japan and EZR; Saitama Medical Center, Jichi Medical University, Saitama, Japan) and a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). Continuous data were expressed as mean ± standard deviation or median (interquartile range). Student t tests or Mann-Whitney U-tests were used to compare continuous variables for the 2 groups. Comparison of continuous variables among 3 groups was achieved using one-way analysis of variance or the Kruskal-Wallis test. Tukey or Steel-Dwass tests were used for post hoc analysis of multiple comparisons among 3 groups. The chi-square test or Fisher’s exact test was used to compare categorical variables. The Kaplan-Meier and log-rank tests were used to infer the existence of a difference in the occurrence of primary or secondary outcomes among the W-W, W-M, and M-M groups. The Bonferroni method was used for multiple comparisons. Univariate and multivariate Cox proportional hazards regression analyses were conducted to examine significant factors associated with the primary outcome.
Results
Characteristics of Study Participants
Of the 154 patients who initiated HD during the study period, 138 were included in the analysis. Baseline characteristics and laboratory data of all participants are presented in Table 1. The mean age of the patients was 66.8 ± 14.5 years; 96 (70%) were male. As a primary outcome, 17 deaths and 43 hospitalizations were observed. The overall 2-year mortality and hospitalization rate was 43% (N = 60). The causes of hospitalization were CVDs (N = 14), malignancy (N = 8), infection (N = 7), and others (N = 14). The causes of death were malignancy (N = 4), infection (N = 3), CVDs (N = 2), sudden death (N = 3), and unknown or others (N = 5). Twenty-six deaths were observed as secondary outcomes because of malignancy (N = 6), infection (N = 3), CVD (N = 5), sudden death (N = 3), and unknown or others (N = 9). When patients were divided into 2 groups with and without the primary outcome, the group with the primary outcome had significantly lower systolic blood pressure (155 ± 19 vs. 145 ± 24 mm Hg, p < 0.01), lower serum Cr concentrations (9.2 [7.4, 11.8] vs. 7.5 [6.4, 9.3] mg/dL, p < 0.01), higher eGFR (4.7 [3.6, 5.8] vs. 5.7 [4.6, 6.7] mL/min/1.73 m2, p < 0.01), and higher serum CRP concentrations (0.16 [0.05, 0.81] vs. 0.44 [0.12, 3.17] mg/dL, p < 0.01) compared to the group without primary outcomes. Results from simplified SGA showed that 97 (70%), 25 (18%), and 16 (12%) of patients were classified, respectively, into W-W, W-M, and M-M groups. The nutritional statuses of patients classified using simplified SGA differed between groups with and without the primary outcome, with a significantly higher percentage of patients in the M-M category in the group with the primary outcome than in the group without (Table 1).
Table 1.
Comparison of two groups divided by primary outcome within 2 years after HD initiation
| Variables | All (N = 138) | 2-year hospitalization or death (−) (N = 78) | 2-year hospitalization or death (+) (N = 60) | p |
|---|---|---|---|---|
| Age, years | 66.8±14.5 | 65.2±14.6 | 68.8±14.2 | 0.15 |
| Male n (%) | 96 (70) | 51 (65) | 45 (75) | 0.26 |
| Diabetic kidney disease, n (%) | 56 (41) | 29 (37) | 27 (45) | 0.39 |
| Planned HD initiation, n (%) | 75 (54) | 48 (62) | 27 (45) | 0.06 |
| History of CVD, n (%) | 59 (43) | 30 (38) | 29 (48) | 0.30 |
| Body mass index, kg/m2 | 23.9 (20.9, 27.4) | 23.6 (21.0, 27.1) | 24.5 (20.9, 27.8) | 0.43 |
| Systolic blood pressure, mm Hg | 151±22 | 155±19 | 145±24 | <0.01 |
| Diastolic blood pressure, mm Hg | 77±16 | 77±15 | 77±17 | 0.83 |
| Hb, g/dL | 9.0±1.6 | 8.9±1.5 | 9.0±1.8 | 0.97 |
| Alb, g/dL | 3.1±0.6 | 3.2±0.5 | 3.0±0.6 | 0.12 |
| BUN, mg/dL | 92 (76, 107) | 92 (75, 106) | 92 (76,111) | 0.96 |
| Cr, mg/dL | 8.5 (6.9, 10.5) | 9.2 (7.4, 11.8) | 7.5 (6.4, 9.3) | <0.01 |
| eGFR, mL/min/1.73 m2 | 5.2 (4.0, 6.2) | 4.7 (3.6, 5.8) | 5.7 (4.6, 6.7) | <0.01 |
| UA, mg/dL | 7.3 (6.1, 8.9) | 7.3 (6.1, 8.8) | 7.3 (6.2, 8.9) | 0.58 |
| Corrected Ca, mg/dL | 8.8 (8.4, 9.1) | 8.8 (8.4, 9.3) | 8.7 (8.4, 9.0) | 0.35 |
| Pi, mg/dL | 5.7 (4.9, 6.8) | 5.8 (4.7, 6.9) | 5.7 (5.1, 6.7) | 0.84 |
| CRP, mg/dL | 0.24 (0.07, 1.66) | 0.16 (0.05, 0.81) | 0.44 (0.12, 3.17) | <0.01 |
| T.chol, mg/dL | 155 (126, 183) | 162 (135, 185) | 148 (118, 173) | 0.06 |
| TG, mg/dL | 117 (88, 158) | 117 (86, 158) | 117 (91, 158) | 0.75 |
| HbA1c (%) | 5.7 (5.3, 6.3) | 5.7 (5.3, 6.2) | 5.8 (5.3, 6.3) | 0.93 |
| GNRI score | 91 (85, 101) | 92 (86, 100) | 91 (84, 102) | 0.77 |
| NRI-JH, n (%) | | | | 0.01 |
| Low risk | 68 (49) | 46 (59) | 22 (37) | |
| Medium risk | 61 (44) | 26 (33) | 35 (58) | |
| High risk | 9 (7) | 6 (8) | 3 (5) | |
| Simplified SGA, n (%) | | | | <0.01 |
| W-W | 97 (70) | 61 (78) | 36 (65) | |
| W-M | 25 (18) | 14 (18) | 11 (16) | |
| M-M | 16 (12) | 3 (4) | 13 (19) |
Continuous data are presented as mean ± SD or median (IQR).
HD, hemodialysis; CVD, cardiovascular disease; Hb, hemoglobin; Alb, albumin; BUN, blood urea nitrogen; Cr, creatinine; eGFR, estimated glomerular filtration rate; UA, uric acid; Ca, calcium; Pi, inorganic phosphate; CRP, C-reactive protein; T.chol, total cholesterol; TG, triglyceride; HbA1c, hemoglobin A1c; GNRI, geriatric nutrition risk index; NRI-JH, nutritional risk index for Japanese hemodialysis patients; SGA, subjective global assessment; W-W, rated as well-nourished by both nurses; W-M, rated as well-nourished by one nurse and malnourished by the other; M-M, rated as moderately to severely malnourished by both nurses.
All participants were divided into three groups according to the results of the simplified SGA. Clinical parameters and outcomes were compared among groups (Table 2). As the nutritional status as evaluated by 2 dialysis nurses became worse, the rate of planned HD initiation became significantly lower (p < 0.01). Moreover, the incidence of primary and secondary outcomes was significantly higher (p < 0.01 for both). The M-M group had significantly lower Hb (p < 0.01) and higher CRP (p < 0.01) concentrations than either the W-W or W-M groups, and also had significantly lower Alb (p < 0.01), lower Cr (p = 0.01), higher eGFR (p = 0.02), and lower GNRI score (p < 0.01) than the W-W group.
Table 2.
Correlation of clinical parameters and outcomes among groups divided by simplified SGA
| Variables | W-W (N = 97) | W-M (N = 25) | M-M (N = 16) | p value |
|---|---|---|---|---|
| Age, years | 66.1±14.5 | 68.7±15.3 | 68.0±13.5 | 0.69 |
| Male, n (%) | 68 (70) | 17 (68) | 11 (69) | 0.98 |
| Diabetic kidney disease, n (%) | 39 (40) | 11 (44) | 6 (38) | 0.91 |
| Planned HD initiation, n (%) | 67 (69) | 6 (24) | 2 (13) | <0.01 |
| History of CVD, n (%) | 41 (42) | 11 (44) | 7 (44) | 0.98 |
| Body mass index, kg/m2 | 24.1 (20.9, 27.9) | 23.9 (22.0, 27.2) | 22.0 (20.7, 25.2) | 0.39 |
| Systolic blood pressure, mm Hg | 152±20 | 151±22 | 146±32 | 0.62 |
| Diastolic blood pressure, mm Hg | 77±14 | 78±20 | 74±19 | 0.60 |
| Hb, g/dL | 9.0±1.5 | 9.6±1.5 | 7.9±1.5a,b | <0.01 |
| Alb, g/dL | 3.2±0.5 | 3.0±0.6 | 2.6±0.5a | <0.01 |
| BUN, mg/dL | 92 (78, 108) | 87 (73, 99) | 103 (75, 116) | 0.27 |
| Cr, mg/dL | 8.5 (6.9, 10.4) | 9.1 (7.4, 11.3) | 7.5 (6.4, 9.1)a | 0.01 |
| eGFR, mL/min/1.73 m2 | 5.0 (3.9, 6.0) | 6.0 (4.4, 7.1) | 6.0 (4.6, 8.0) | 0.02 |
| UA, mg/dL | 7.3 (6.2, 8.8) | 7.3 (5.8, 8.4) | 7.8 (6.3, 9.4) | 0.82 |
| Corrected Ca, mg/dL | 8.8 (8.4, 9.1) | 8.7 (8.3, 9.0) | 8.9 (8.5, 9.2) | 0.91 |
| Pi, mg/dL | 5.8 (4.9, 6.6) | 5.5 (4.6, 7.3) | 6.3 (5.5, 7.1) | 0.46 |
| CRP, mg/dL | 0.19 (0.06, 1.02) | 0.14 (0.08, 1.09) | 2.74 (0.74, 6.04)a,b | <0.01 |
| T.chol, mg/dL | 151 (127, 179) | 170 (132, 194) | 138 (108, 181) | 0.17 |
| TG, mg/dL | 117 (91, 156) | 127 (104, 168) | 86 (72, 162) | 0.25 |
| HbA1c (%) | 5.7 (5.3, 6.2) | 5.7 (5.3, 6.3) | 5.6 (5.1, 6.1) | 0.56 |
| GNRI score | 93 (87, 102) | 93 (83, 100) | 78 (74, 89)a | <0.01 |
| NRI-JH, n (%) | | | | 0.05 |
| Low risk | 55 (57) | 8 (32) | 5 (31) | |
| Medium risk | 36 (37) | 16 (64) | 9 (56) | |
| High risk | 6 (6) | 1 (4) | 2 (13) | |
| 2-year mortality/hospitalization | 36 (37) | 11 (44) | 13 (81) | <0.01 |
| 2-year mortality | 11 (11) | 6 (24) | 9 (56) | <0.01 |
Continuous data are presented as mean ± SD or median (IQR).
W-W, rated as well-nourished by both nurses; W-M, rated as well-nourished by one nurse and malnourished by the other; M-M, rated as moderately to severely malnourished by both of two nurses; HD, hemodialysis; CVD, cardiovascular disease; Hb, hemoglobin; Alb, albumin; BUN, blood urea nitrogen; Cr, creatinine; eGFR, estimated glomerular filtration rate; UA, uric acid; Ca, calcium; Pi, inorganic phosphate; CRP, C-reactive protein; T.chol, total cholesterol; TG, triglyceride; HbA1c, hemoglobin A1c; GNRI, geriatric nutrition risk index; NRI-JH, nutritional risk index for Japanese hemodialysis patients; SGA, subjective global assessment.
a p < 0.05 vs. W-W group.
b p < 0.05 vs. W-M group.
Relation between Outcomes and Simplified SGA
Using Kaplan-Meier analysis and the log-rank test, the composite primary outcome of survival or hospitalization-free rate within the first 2 years after initiation of HD as well as the secondary outcome of survival rate alone were compared among the W-W, W-M, and M-M groups (Fig. 1). The M-M group had a significantly lower survival or hospitalization-free rate than either the W-W group (p < 0.001) or W-M group (p = 0.024) (Fig. 1a). The survival rate of the M-M group was significantly lower than that of the W-W group (p < 0.001), but it was not significantly different from that of the W-M group (p = 0.15) (Fig. 1b).
Fig. 1.
Kaplan-Meier analysis for composite outcome (survival or hospitalization-free rate) (a) and survival rate (b) within 2 years after initiating HD. W-W, rated as well-nourished by both examiners; W-M, rated as well-nourished by one examiner and malnourished by the other; M-M, rated as moderately to severely malnourished by both examiners.
Using the Cox proportional hazards model, we evaluated the association between the primary outcome, 2-year mortality and hospitalization, and the results of simplified SGA (online suppl. Tables S2, 3). From univariate analysis, significant associations were found between the outcome and the M-M category in simplified SGA {hazard ratio (HR; 95% confidence interval [CI]) = 3.80 [2.01–7.21], p < 0.01}, systolic blood pressure (HR [95% CI] = 0.98 [0.97–0.99], p < 0.01), serum Cr concentration (HR [95% CI] = 0.86 [0.78–0.95], p < 0.01), eGFR (HR [95% CI] = 1.14 [1.04–1.26], p < 0.01), CRP (HR [95% CI] = 1.09 [1.03–1.14], p < 0.01), total cholesterol (HR [95% CI] = 0.99 [0.986–0.999], p = 0.02), and the medium risk category in NRI-JH (HR [95% CI] = 2.07 [1.22–3.54], p < 0.01) (online suppl. Table S2). Results of multivariate analysis showed that the M-M category in simplified SGA was associated significantly with the primary outcome (Table 3). After adjustment for age, systolic blood pressure, Cr, and CRP, the M-M category in simplified SGA was found to be associated significantly with 2-year mortality and hospitalization; however, the other nutritional assessment indices, NRI-JH classification and GNRI score, were not (Table 4).
Table 3.
Results from multivariate Cox proportional hazards model for 2-year mortality and hospitalization
| Variables | Model 1 | Model 2 | Model 3 | |||
|---|---|---|---|---|---|---|
| HR (95% CI) | p value | HR (95% CI) | p value | HR (95% CI) | p value | |
| Simplified SGA | ||||||
| W-W | Ref | | Ref | | Ref | |
| W-M | 1.20 (0.56–2.57) | 0.63 | 1.20 (0.56–2.58) | 0.95 | 1.22 (0.57–2.62) | 0.61 |
| M-M | 3.34 (1.52–7.36) | <0.01 | 3.07 (1.52–6.19) | <0.01 | 3.18 (1.56–6.50) | <0.01 |
| Age | 1.01 (0.99–1.03) | 0.45 | | | 1.00 (0.98–1.02) | 0.96 |
| Planned HD initiation | 0.95 (0.50–1.79) | 0.87 | | | | |
| Systolic blood pressure | 0.98 (0.97–0.99) | <0.01 | 0.99 (0.98–1.00) | 0.18 | 0.99 (0.98–1.00) | 0.13 |
| Alb | 0.83 (0.49–1.43) | 0.50 | | | | |
| Cr | | | 0.90 (0.80–1.00) | 0.054 | 0.89 (0.79–1.00) | 0.052 |
| CRP | | | 1.04 (0.98–1.12) | 0.21 | | |
| T chol | | | 0.99 (0.99–0.999) | 0.03 | 0.99 (0.99–0.999) | 0.02 |
W-W, rated as well-nourished by both nurses; W-M, rated as well-nourished by one nurse and malnourished by the other; M-M, rated as moderately to severely malnourished by both of two nurses; SGA, Subjective Global Assessment; Alb, albumin; Cr, creatinine; CRP, C-reactive protein; T.chol, total cholesterol; HR, hazard ratio; CI, confidence interval; Ref., reference.
Table 4.
Results of Cox proportional hazards analysis for the association between nutritional risk assessment indices and 2-year mortality or hospitalization
| Variables | Crude model | Adjusted model* | ||
|---|---|---|---|---|
| HR (95% CI) | p value | HR (95% CI) | p value | |
| NRI-JH | ||||
| Low risk | Ref | | Ref | |
| Medium risk | 2.07 (1.22–3.54) | <0.01 | 1.45 (0.76–2.77) | 0.26 |
| High risk | 1.02 (0.31–3.42) | 0.97 | 0.71 (0.20–2.52) | 0.60 |
| GNRI score | 0.99 (0.97–1.01) | 0.45 | 0.99 (0.97–1.01) | 0.34 |
| Simplified SGA | ||||
| W-W | Ref | | Ref | |
| W-M | 1.31 (0.66–2.57) | 0.44 | 1.03 (0.51–2.10) | 0.93 |
| M-M | 3.80 (2.01–7.21) | <0.01 | 3.43 (1.77–6.65) | <0.01 |
*: adjusted for age, systolic blood pressure, serum creatinine, and serum C-reactive protein.
W-W, rated as well-nourished by both nurses; W-M, rated as well-nourished by one nurse and malnourished by the other; M-M, rated as moderately to severely malnourished by both of two nurses; SGA, Subjective Global Assessment; GNRI, geriatric nutrition risk index; NRI-JH, Nutritional Risk Index for Japanese Hemodialysis Patients; HR, hazard ratio; CI, confidence interval; Ref., reference.
Discussion
As described herein, we investigated whether simplified SGA, a nutritional risk assessment using several components of SGA except for physical findings, at HD initiation was associated with mortality and hospitalization within 2 years. Being a member of the M-M group, comprising patients judged as malnourished by both examiners, was associated significantly with the primary outcome. The M-M group had significantly lower planned HD initiation rate and Hb, Alb, and Cr concentrations and higher CRP concentrations than the other groups. Multivariate analysis demonstrated that the M-M category in simplified SGA was associated with the outcome, but results from other nutrition assessment tools, NRI-JH and GNRI, showed no such association.
The M-M group had significantly lower Hb, Alb, and Cr concentrations and higher CRP concentrations than other groups. These clinical characteristics observed in the M-M group might reflect a state of Malnutrition-Inflammation Complex Syndrome. Systemic inflammation causes treatment-resistant anemia through impaired erythropoiesis [15], and reduction in serum Alb and creatinine via suppressed protein synthesis and accelerated muscle catabolism [16]. The discrepancy between low Cr levels and high blood urea nitrogen levels in the M-M group might indicate ongoing muscle wasting [17]. Sohrabi et al. [18] reported that more severe malnutrition tends to be associated with greater inflammatory involvement among patients on HD. Patients classified as M-M group by simplified SGA at HD initiation might be those with malnutrition caused by chronic inflammation.
In our study, the M-M category in simplified SGA was found to be associated significantly with the primary outcome in patients starting HD by the adjusted Cox proportional hazards model, but results obtained using the other nutritional assessment tools (GNRI and NRI-JH) were found to have no such association. During the dialysis initiation period, serum Alb concentrations are prone to fluctuate because of inflammation or fluid overload [19]. Therefore, the utility of nutritional assessments using GNRI and NRI-JH, both of which include serum Alb as a component, might be limited in such a clinical setting. An earlier study demonstrated that low 7-point SGA, a quantitative version of SGA, was associated more strongly with mortality than low GNRI was. Moreover, 7-point SGA was associated significantly with hospitalization but GNRI was not [20]. Prognostic predictive capability of NRI-JH at HD initiation might not be established because it was developed based on data obtained from patients on maintenance HD [13]. Furthermore, unlike GNRI and NRI-JH, which are based on data from a single point in time, the simplified SGA presents the important benefit of being an index that reflects changes, such as weight loss, over time, which is independently associated with mortality in patients with CKD [21].
Simplified SGA does not include the physical examination component. That omission might facilitate its use in telemedicine, isolation, and retrospective evaluations. Moreover, by incorporating weight change, it preserves a key component of nutritional assessment. An earlier study demonstrated that a modified SGA including weight change, despite omission of several original components, exhibited comparable malnutrition-predictive performance to that provided by the original SGA [22].
Our study had several limitations. First, because this is a single-center retrospective study of a small number of patients, it remains unclear whether the results are applicable to patients starting HD at other hospitals. Second, this study did not assess protein-energy wasting (PEW) or sarcopenia, which are conditions known to be associated with poor prognosis. Diagnosing these conditions is somewhat complex because it requires physical measurements such as muscle mass. Therefore, this study, which specifically examined the relation between simplified nutritional assessment without physical examination and prognosis, did not evaluate them. Third, given the simplified SGA’s subjectivity and reduced item count, two dialysis nurses with more than 10 years of experience independently assessed each patient. Their evaluations were integrated for analysis, with inter-rater reliability (Cohen’s kappa = 0.45) indicating moderate agreement [23]. Multi-person assessment might help avoid missing patients at nutritional risk. Fourth, we were unable to do a direct comparison of the prognostic performance of the simplified SGA to that of the original SGA including the physical examination component. To address these limitations, a large, multicenter prospective cohort study should be conducted.
In conclusion, this study demonstrated that the results of simplified SGA at HD initiation are associated significantly with mortality and hospitalization rate within 2 years. A simple, objective nutritional risk assessment based on basic patient information without physical examination by 2 dialysis nurses might be useful for prognostic risk stratification of patients starting HD. Further research must be conducted to validate and extend these findings.
Acknowledgments
We are grateful to Fastek, Ltd. for assistance with English proofreading.
Statement of Ethics
All procedures performed in this retrospective observational study were conducted in accordance with the ethical standards of the Research Ethics Committee of the Faculty of Medicine, the University of Tokyo (Approval No. 2269, December 22, 2008), and with the principles of the Declaration of Helsinki and its later amendments or comparable ethical standards. Written informed consent was waived because of the retrospective observational characteristics of the study; instead, an opt-out consent procedure, which was reviewed and approved by the same Ethics Committee, was adopted. Information regarding the study (including its purpose, methods, required data, and study period) was provided on the website of our department; participants were given the opportunity to opt out.
Conflict of Interest Statement
The authors declare that they have no competing financial or other interest or personal relationship that could have influenced this paper or the study it describes.
Funding Sources
The authors declare that they have no relevant financial interests.
Author Contributions
All authors contributed to the study conception and design. Material preparation, data collection, and analyses were performed by Y.H., K.S., M.I., Y.K., and T.Y. This study was designed by Y.H. The first draft of the manuscript was written by K.D. and Y.H. and was revised by K.S., M.I., Y.K., T.Y., R.M., and M.N. All authors have read and approved the final manuscript.
Funding Statement
The authors declare that they have no relevant financial interests.
Data Availability Statement
Data that support the findings of this study are not publicly available because of privacy reasons but are available from the corresponding author (Y.H.) upon request.
Supplementary Material.
Supplementary Material.
References
- 1. Carrero JJ, Thomas F, Nagy K, Arogundade F, Avesani CM, Chan M, et al. Global prevalence of protein-energy wasting in kidney disease: a meta-analysis of contemporary observational studies from the International society of renal nutrition and metabolism. J Ren Nutr. 2018;28(6):380–92. [DOI] [PubMed] [Google Scholar]
- 2. Cordos M, Martu M-A, Vlad C-E, Toma V, Ciubotaru AD, Badescu MC, et al. Early detection of inflammation and malnutrition and prediction of acute events in hemodialysis patients through PINI (prognostic inflammatory and nutritional index). Diagnostics. 2024;14(12):1273. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Yamada S, Arase H, Yoshida H, Kitamura H, Tokumoto M, Taniguchi M, et al. Malnutrition-Inflammation complex syndrome and bone fractures and cardiovascular disease events in patients undergoing hemodialysis: the Q-Cohort study. Kidney Med. 2022;4(3):100408. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. McLean C, Randall A-M, Ryan M, Smyth B, Thomsett M, Brown MA, et al. The association of frailty and malnutrition with dietary intake and gastrointestinal symptoms in people with kidney failure: 2-year prospective study. J Ren Nutr. 2024;34(2):177–84. [DOI] [PubMed] [Google Scholar]
- 5. Visiedo L, Rey L, Rivas F, López F, Tortajada B, Giménez R, et al. The impact of nutritional status on health-related quality of life in hemodialysis patients. Sci Rep. 2022;12(1):3029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Rashid I, Sahu G, Tiwari P, Willis C, Asche CV, Bagga TK, et al. Malnutrition as a potential predictor of mortality in chronic kidney disease patients on dialysis: a systematic review and meta-analysis. Clin Nutr. 2024;43(7):1760–9. [DOI] [PubMed] [Google Scholar]
- 7. Wilson B, Fernandez-Madrid A, Hayes A, Hermann K, Smith J, Wassell A. Comparison of the effects of two early intervention strategies on the health outcomes of malnourished hemodialysis patients. J Ren Nutr. 2001;11(3):166–71. [DOI] [PubMed] [Google Scholar]
- 8. Lacson E, Ikizler TA, Lazarus JM, Teng M, Hakim RM. Potential impact of nutritional intervention on end-stage renal disease hospitalization, death, and treatment costs. J Ren Nutr. 2007;17(6):363–71. [DOI] [PubMed] [Google Scholar]
- 9. Calegari A, Barros EG, Veronese FV, Thomé FS. Malnourished patients on hemodialysis improve after receiving a nutritional intervention. J Bras Nefrol. 2011;33(4):394–401. [PubMed] [Google Scholar]
- 10. Detsky AS, McLaughlin JR, Baker JP, Johnston N, Whittaker S, Mendelson RA, et al. What is subjective global assessment of nutritional status? JPEN J Parenter Enteral Nutr. 1987;11(1):8–13. [DOI] [PubMed] [Google Scholar]
- 11. Kalantar-Zadeh K, Kleiner M, Dunne E, Lee GH, Luft FC. A modified quantitative subjective global assessment of nutrition for dialysis patients. Nephrol Dial Transpl. 1999;14(7):1732–8. [DOI] [PubMed] [Google Scholar]
- 12. Adequacy of dialysis and nutrition in continuous peritoneal dialysis: association with clinical outcomes. Canada–USA (CANUSA) peritoneal dialysis study group. J Am Soc Nephrol. 1996;7(2):198–207. [DOI] [PubMed] [Google Scholar]
- 13. Kanda E, Kato A, Masakane I, Kanno Y. A new nutritional risk index for predicting mortality in hemodialysis patients: nationwide cohort study. PLoS One. 2019;14(3):e0214524. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Bouillanne O, Morineau G, Dupont C, Coulombel I, Vincent J-P, Nicolis I, et al. Geriatric Nutritional Risk Index: a new index for evaluating at-risk elderly medical patients. Am J Clin Nutr. 2005;82(4):777–83. [DOI] [PubMed] [Google Scholar]
- 15. Rafiean-Kopaie M, Nasri H. Impact of inflammation on anemia of hemodialysis patients who were under treatment of recombinant human erythropoietin. J Ren Inj Prev. 2013;2(3):93–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Kaysen GA, Dubin JA, Müller H-G, Rosales L, Levin NW, Mitch WE, et al. Inflammation and reduced albumin synthesis associated with stable decline in serum albumin in hemodialysis patients. Kidney Int. 2004;65(4):1408–15. [DOI] [PubMed] [Google Scholar]
- 17. Deger SM, Hung AM, Gamboa JL, Siew ED, Ellis CD, Booker C, et al. Systemic inflammation is associated with exaggerated skeletal muscle protein catabolism in maintenance hemodialysis patients. JCI Insight. 2017;2(22):e95185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Sohrabi Z, Eftekhari MH, Eskandari MH, Rezaeianzadeh A, Sagheb MM. Malnutrition-inflammation score and quality of life in hemodialysis patients: is there any correlation? Nephrourol Mon. 2015;7(3):e27445. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Robinson B, Zhang J, Morgenstern H, Bradbury BD, Ng LJ, McCullough K, et al. World-wide, mortality is a high risk soon after initiation of hemodialysis. Kidney Int. 2014;85(1):158–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Rodrigues J, Santin F, Brito FDSB, Lindholm B, Stenvinkel P, Avesani CM. Nutritional status of older patients on hemodialysis: which nutritional markers can best predict clinical outcomes? Nutrition. 2019;65:113–9. [DOI] [PubMed] [Google Scholar]
- 21. Ku E, Kopple JD, Johansen KL, McCulloch CE, Go AS, Xie D, et al. Longitudinal weight change during CKD progression and its association with subsequent mortality. Am J Kidney Dis. 2018;71(5):657–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Nursal TZ, Noyan T, Tarim A, Karakayali H. A new weighted scoring system for subjective global assessment. Nutrition. 2005;21(6):666–71. [DOI] [PubMed] [Google Scholar]
- 23. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics. 1977;33(1):159–74. [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data that support the findings of this study are not publicly available because of privacy reasons but are available from the corresponding author (Y.H.) upon request.
