Accuracy of Quick and Easy Undernutrition Screening Tools—Short Nutritional Assessment Questionnaire, Malnutrition Universal Screening Tool, and Modified Malnutrition Universal Screening Tool—in Patients Undergoing Cardiac Surgery

RESEARCH Research and Practice Innovations

Accuracy of Quick and Easy Undernutrition Screening Tools—Short Nutritional Assessment Questionnaire, Malnutrition Universal Screening Tool, and Modified Malnutrition Universal Screening Tool—in Patients Undergoing Cardiac Surgery LENNY M. W. VAN VENROOIJ, PhD, RD; PAUL A. M. VAN LEEUWEN, PhD, MD; WENDY HOPMANS, MSc; MIEKE M. M. J. BORGMEIJER-HOELEN, MD; RIEN DE VOS, PhD; BAS A. J. M. DE MOL, PhD, MD

ABSTRACT The objective of this study was to compare the quick-andeasy undernutrition screening tools, ie, Short Nutritional Assessment Questionnaire and Malnutrition Universal Screening Tool, in patients undergoing cardiac surgery with respect to their accuracy in detecting undernutrition measured by a low-fat free mass index (FFMI; calculated as kg/m2), and secondly, to assess their association with

L. M. W. van Venrooij is an epidemiologist and dietitian, Department of Cardio-thoracic Surgery and Department of Dietetics, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands. P. A. M. van Leeuwen is a professor, Department of Surgery, VU University Medical Center, Amsterdam, The Netherlands. W. Hopmans is a doctoral student, Department of Public and Occupational Health and Department of Radiation Oncology, VU University Medical Center, Amsterdam, The Netherlands; at the time of the study, she was a student at the Institute of Health Sciences, Faculty Earth and Life Sciences, VU University, Amsterdam, The Netherlands. M. M. M. J. BorgmeijerHoelen is a medical doctor, and B. De Mol is a professor, Department of Cardio-thoracic Surgery, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands. R. de Vos is an associate professor, Department of Clinical Epidemiology and Biostatistics, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands. Address correspondence to: Lenny M. W. van Venrooij, PhD, RD, Department of Cardio-thoracic Surgery and Department of Dietetics, Room G3-152, Academic Medical Center, University of Amsterdam, PO Box 22700, 1100 DE Amsterdam, The Netherlands. E-mail: l.m. [email protected] Manuscript accepted: June 24, 2011. Copyright © 2011 by the American Dietetic Association. 0002-8223/$36.00 doi: 10.1016/j.jada.2011.09.009

1924

Journal of the AMERICAN DIETETIC ASSOCIATION

postoperative adverse outcomes. Between February 2008 and December 2009, a single-center observational cohort study was performed (n⫽325). A low FFMI was set at ⱕ14.6 in women and ⱕ16.7 in men measured using bioelectrical impedance spectroscopy. To compare the accuracy of the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire in detecting low FFMI sensitivity, specificity, and other accuracy test characteristics were calculated. The associations between the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire and adverse outcomes were analyzed using logistic regression analyses with odds ratios and 95% confidence intervals (CI) presented. Sensitivity and receiver operator characteristicbased area under the curve to detect low FFMI were 59% and 19%, and 0.71 (95% CI: 0.60 to 0.82) and 0.56 (95% CI: 0.44 to 0.68) for the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire, respectively. Accuracy of the Malnutrition Universal Screening Tool improved when age and sex were added to the nutritional screening process (sensitivity 74%, area under the curve: 0.72 [95% CI: 0.62 to 0.82]). This modified version of the Malnutrition Universal Screening Tool, but not the original Malnutrition Universal Screening Tool or Short Nutritional Assessment Questionnaire, was associated with prolonged intensive care unit and hospital stay (odds ratio: 2.1, 95% CI: 1.3 to 3.4; odds ratio: 1.6, 95% CI: 1.0 to 2.7). The accuracy to detect a low FFMI was considerably higher for the Malnutrition Universal Screening Tool than for the Short Nutritional Assessment Questionnaire, although still marginal. Further research to evaluate the modified version of the Malnutrition Universal Screening Tool, ie, the cardiac surgery– specific Malnutrition Universal Screening Tool, is needed prior to implementing. J Am Diet Assoc. 2011;111:1924-1930.

I

n cardiac surgery, disease-related undernutrition is associated with an increased occurrence of postoperative infectious and noninfectious complications, mortality, prolonged intensive care unit (ICU) and hospital

© 2011 by the American Dietetic Association

stay, and lower quality of life (1-5). If measured by a low body mass index (BMI; calculated as kg/m2), unintended weight loss, a low fat-free mass index (FFMI; calculated as kg/m2) or postoperative loss of muscle mass, approximately 10% to 25% of patients undergoing cardiac surgery are already undernourished or become so (1-5). Thus, close observation to identify and treat the undernourished is important in order to reduce the risk of complications after cardiac surgery. To identify the undernourished, the use of quick and easy screening tools is recommended (6). Without nutritional screening tools, 50% of undernourished patients remain unidentified (7). The currently used screening tools, such as the Malnutrition Universal Screening Tool (8) and the Short Nutritional Assessment Questionnaire (7,9) have not been tested in the cardiac surgery population (10). Both tools include unintended weight loss in the months before the operation and a question referring to nutritional intake. The main difference between the Malnutrition Universal Screening Tool and the Short Nutritional Assessment Questionnaire is that the Malnutrition Universal Screening Tool scores for low BMI and the Short Nutritional Assessment Questionnaire does not. In the cardiac surgery population, both low BMI and unintended weight loss have been associated with adverse outcomes (3). Thus, it can be hypothesized that the Malnutrition Universal Screening Tool screens patients who are undernourished prior to cardiac surgery more accurately because it includes both low BMI and weight loss. The main aim of this study was to compare the accuracy of the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire to detect undernutrition measured by a low FFMI. Previous studies show that a low preoperative FFMI is associated with adverse outcomes after cardiac surgery also independent from unintended weight loss and low BMI (4,5). Second, both tools were tested for their association with the occurrence of postoperative infections, death, and a prolonged length of ICU or hospital stay. Last, if accuracy of the Malnutrition Universal Screening Tool or Short Nutritional Assessment Questionnaire was low, it was assessed if the tool could be improved by the integration of specific patient characteristics known to be associated with a low FFMI such as age, sex, operative risk, or severity of heart failure (4,11). METHODS Patients and Design A prospective observational cohort study was performed to compare the Malnutrition Universal Screening Tool and the Short Nutritional Assessment Questionnaire. Between February 2008 and December 2009, patients (18 years of age and older) undergoing coronary artery bypass graft and/or heart valve surgery with extracorporeal circulation admitted to the cardiothoracic surgical ward at the Academic Medical Center, Amsterdam, were consecutively asked to participate. Those patients who were not willing or able to give written informed consent, who had a pacemaker, a congenital heart abnormality, or who had undergone open-heart surgery within the preceding 3 months were excluded. This study was approved by the

Medical Ethical Committee of the Academic Medical Center, Amsterdam, The Netherlands. Measurement of Short Nutritional Assessment Questionnaire, Malnutrition Universal Screening Tool, and FFM On admission to the cardiothoracic surgical ward, information on preoperative nutritional data (Short Nutritional Assessment Questionnaire and Malnutrition Universal Screening Tool–related questions, body weight 1 month and 6 months before surgery) were obtained from the patients. Actual body weight (kg), height (cm), and FFM (kg) were measured with patients barefoot and in their underwear. Body weight was measured using an electronic beam scale with digital read-out to the nearest 0.1 kg (Inventum Scala PW200, Veenendaal, The Netherlands). Height was measured to the nearest 0.5 cm using a stadiometer (Seca, Hamburg, Germany). FFM was determined by bioelectrical impedance spectroscopy (BodyScout Fresenius Kabi, Bad Homburg van der Höhe, Germany) (12). Bioelectrical impedance spectroscopy measurements were done 10 minutes after patients were lying in supine position with legs apart and arms abducted (13). The measurement was performed on the right side of the body using four electrodes (3M Red Dot; 3M Health Care, Nauss, Germany); two electrodes were placed on the dorsum of the hand and two on the dorsum of the foot. A 5- to 1,000-kHz electrical current was introduced into the body and tissue conductivity was measured. At low frequency, the electrical signal travels predominantly through the extracellular space, whereas high-frequency signals travel through the extra- and intracellular space. Thereby, the multiple frequencies enable a difference between intracellular fluid and extracellular fluid to be measured. Subsequently, FFM was calculated (12,14). To adjust for differences in body height, FFMI was calculated by dividing FFM (kg) by squared height (m2) (15). Malnutrition Universal Screening Tool, Short Nutritional Assessment Questionnaire, and Low FFMI Scores The reference standard low FFMI was set at ⱕ14.6 in women and ⱕ16.7 in men in accordance with the literature (4,16).The Malnutrition Universal Screening Tool score for undernutrition is ⱖ1 (8), and the Short Nutritional Assessment Questionnaire score for undernutrition is ⱖ2 (7). Patient Characteristics and Postoperative Adverse Outcomes Data describing patient-, cardiac-, and operation-related baseline characteristics and postoperative adverse outcomes were extracted from the standard electronic database of the cardiac-thoracic surgery department and from medical files. The database included the European System for Cardiac Operation Risk Evaluation (17,18). This preoperative risk evaluation includes patient and cardiacrelated factors, ie, age, sex, and comorbidities such as chronic obstructive pulmonary disease or recent myocardial infarction, and also operation-related factors. A European System for Cardiac Operation Risk Evaluation of ⱖ6 refers to patients with a high operative risk; a score of December 2011 ● Journal of the AMERICAN DIETETIC ASSOCIATION

1925

⬍6 refers to a patient with a low or medium operative risk. The postoperative adverse outcomes were occurrence of postoperative infection and/or death; prolonged ICU stay; and prolonged hospital stay, all during the period of hospitalization for surgery. Postoperative infection was defined as the composite of septicemia, respiratory tract infection, mediastinitis, deep sternal wound, and leg wound infection. These infectious complications were in accordance with the definitions of The Society of Thoracic Surgeons (19). Prolonged postoperative ICU stay was defined as ⱖ48 hours (including readmission hours, patients who died were excluded) and prolonged hospital stay was defined as a postoperative hospital stay ⱖ7 days (patients who died were excluded). Postoperative death was defined as death during hospital admission. Statistical Analyses First, to compare the accuracy of the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire in detecting a low FFMI sensitivity and specificity rates were calculated. Also, other accuracy test characteristics such as the positive predictive value, negative predictive value, positive likelihood ratio, and the receiver operating characteristic– based area under the curves were utilized (20,21). Second, the associations of the Malnutrition Universal Screening Tool and the Short Nutritional Assessment Questionnaire with the occurrence of infection and/or death, prolonged ICU or hospital stay after cardiac surgery were analyzed using logistic regression analyses. Odds ratios (ORs) and 95% confidence intervals (CI) were calculated. Last, if the level of accuracy was low, it was assessed if the accuracy of the Malnutrition Universal Screening Tool or Short Nutritional Assessment Questionnaire could be improved by the integration of high-risk preoperative patient characteristics. High-risk patients for a low FFMI are women, elderly patients (65 years of age and older), high operative risk patients (European System for Cardiac Operation Risk Evaluation ⱖ6), patients undergoing heart valve surgery, and patients suffering from severe heart failure (N-terminal prohormone brain natriuretic peptide [marker for heart failure] ⱖ600 ng/L) (4,11). Backward multivariate logistic regression modeling with Malnutrition Universal Screening Tool or Short Nutritional Assessment Questionnaire, sex, age, operative procedure, heart failure, and operative risk as independent variables and low FFMI as dependent variable was conducted. Variables were removed stepwise if not statistically significant. To assess its sensitivity and specificity, a score to define undernutrition was calculated using the ␤-regression coefficients of the logistic regression analysis (7). Pⱕ0.05 was considered statistically significant. All statistical analyses were performed with SPSS (version 16.0, 2007, SPSS Inc, Chicago, IL). RESULTS Patients Between February 2008 and December 2009, a total of 396 patients were asked to participate. Of these patients 17.9% (n⫽71) refused. These patients were older and,

1926

December 2011 Volume 111 Number 12

Table 1. Sociodemographic profile, operative risk, severity of heart failure, inflammatory status, and operation-related factors of patients undergoing cardiac surgery (n⫽325) Sociodemographic profile Female, n (%) Age (y), mean⫾SDa Age ⱖ65 y, n (%) BMI,b mean⫾SD BMIⱖ30, n (%) BMIⱕ21, n (%) Weight lost ⱖ10% in preceding 6 months, n (%) European System for Cardiac Operation Risk Evaluation, n (%) 0-2 (low risk) 3-5 (medium risk) ⱖ6 (high risk) Laboratory, n (%) NT pro BNPc ⱖ600 ng/L CRPd ⱖ5 mg/L Albumin ⱕ39 g/L Operative procedure, n (%) CABGe Heart valve surgery Both Duration extracorporeal circulation, n (%) CPBf ⱖ130 min ACCg ⱖ95 min

90 (27.7) 65.7⫾10.1 186 (57.2) 27.2⫾4.1 63 (19.4) 13 (4.0) 17 (5.2) 89 (27.4) 109 (33.5) 127 (39.1) 107 (32.9) 53 (16.3) 12 (3.7) 168 (51.7) 105 (32.2) 52 (16.0) 106 (32.6) 96 (29.5)

a

SD⫽standard deviation. BMI⫽body mass index; calculated as kg/m2. NT pro BNP⫽N-terminal prohormone brain natriuretic peptide, marker for heart failure. d CRP⫽C-reactive protein. e CABG⫽coronary artery bypass grafting. f CPB⫽cardiopulmonary bypass time. g ACC⫽aortic cross-clamp time. b c

when scored by the European System for Cardiac Operation Risk Evaluation, these patients had a higher operation risk than those patients who gave informed consent (P⫽0.01 and P⬍0.005, respectively). The remaining data from 325 patients were analyzed. Preoperative baseline characteristics are summarized in Table 1. Accuracy of Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire in Detecting a Low FFMI The prevalence of undernutrition measured by a low FFMI was 8.3% (n⫽27). According to the Malnutrition Universal Screening Tool, 20.9% of patients (n⫽67) were undernourished and according to the Short Nutritional Assessment Questionnaire, 7.5% (n⫽24) were undernourished.The sensitivity rate in detecting a low FFMI was three times higher for the Malnutrition Universal Screening Tool than for the Short Nutritional Assessment Questionnaire (59.3% and 18.5%, respectively), but had a lower specificity (82.7% and 93.6%, respectively) (Table 2). In addition, the patients described by both the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire as undernourished had more often a low FFMI (P⫽0.04) and more severe heart

Table 2. Accuracy of the Short Nutritional Assessment Questionnaire, Malnutrition Universal Screening Tool, and the cardiac surgery–specific version of the Malnutrition Universal Screening Tool in detecting low fat-free mass index Prevalencea MUST, undernourishedg SNAQ, undernourishedh CSSM, undernourishedi

Sensitivity

Specificity

PPVb

NPVc

4™™™™™™™™™™™™™™™™™™™™™™™™™™™™ % (n) ™™™™™™™™™™™™™™™™™™™™™™™™™™™™3 8.3 (27/325) 59.3 (16/27) 82.7 (243/294) 23.9 (16/67) 95.7 (243/254) 8.3 (27/325) 18.5 (5/27) 93.6 (275/294) 20.8 (5/24) 92.6 (275/297) 8.3 (27/325) 74.1 (20/27) 70.1 (206/294) 18.5 (20/108) 96.7 (206/213)

LRⴙd

AUCe (95% CIf)

3.4 2.9 2.5

0.71 (0.60-0.82) 0.56 (0.44-0.68) 0.72 (0.62-0.82)

Undernourished according to the reference standard; fat-free mass index: ⱕ14.6 (women); ⱕ16.7 (men). PPV⫽positive predictive value. c NPV⫽negative predictive value. d LR⫹⫽positive likelihood ratio. e AUC⫽area under the curve. f CI⫽confidence interval. g Undernourished according to the Malnutrition Universal Screening Tool. h Undernourished according to the Short Nutritional Assessment Questionnaire. i Undernourished according to the cardiac surgery–specific version of the Malnutrition Universal Screening Tool. a

b

failure than the well-nourished (P⬍0.005). The patients described as undernourished by the Malnutrition Universal Screening Tool, but not by the Short Nutritional Assessment Questionnaire, included more females, had lower levels of albumin, and included more patients at higher operative risk (P⬍0.005) than did the well-nourished. In both the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire undernourished no differences were seen in age, operative procedure, C-reactive protein levels, or operative time on comparison with the well-nourished. Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire in Relation to Postoperative Adverse Outcomes During the stay at the operating hospital, the incidence of postoperative infection was 5.8% (n⫽19) and mortality was 2.5% (n⫽8). The cumulative incidence of postoperative infection and/or death was 7.7% (n⫽25). A total of 36.4% (n⫽116) had a prolonged postoperative ICU stay and 33.1% (n⫽106) had a prolonged postoperative hospital stay. No associations with the occurrence of infection and/or death or a prolonged length of ICU or hospital stay were observed in either the Malnutrition Universal Screening Tool or the Short Nutritional Assessment Questionnaire (P⬎0.10). Accuracy of a Modified Version of the Malnutrition Universal Screening Tool—The Cardiac Surgery–Specific Malnutrition Universal Screening Tool Backward logistic regression modeling with low FFMI as dependent variable showed that sex and age, but not operative procedure, operative risk or severity of heart failure, further improved the detection of a low FFMI in addition to the original Malnutrition Universal Screening Tool (Table 3). To assess the sensitivity and specificity rates of this modified version of the Malnutrition Universal Screening Tool, the cardiac surgery–specific Malnutrition Universal Screening Tool, a new score was calculated (Table 3): The Malnutrition Universal Screening

Table 3. The Malnutrition Universal Screening Tool extended by other predictive patient characteristics to detect low fat-free mass indexa ORb (95% CIc)

Regression P valued coefficient Scoree

Constant ⫺4.22 MUST,f undernourished 6.10 (2.55-14.62) 0.000* 1.81 Sex, female 3.30 (1.39-7.86) 0.007* 1.19 Age ⱖ65 y 2.94 (1.1-8.01) 0.035* 1.08

2 1 1

a The initial model included the independent variables Malnutrition Universal Screening Tool, sex, age, operation procedure, operation risk, and heart failure. No additional value to detect a low fat-free mass index was found for operation procedure, risk, or heart failure (P⬎0.05). b OR⫽odds ratio. c CI⫽confidence interval. d P value ⱕ0.05 was considered statistically significant using backward logistic regression analysis. e To get round numbers for the cardiac surgery–specific Malnutrition Universal Screening Tool scores, the ␤-coefficients of the logistic regression analyses were rounded to the nearest integer. f MUST⫽ Malnutrition Universal Screening Tool. *Pⱕ0.05.

Tool accorded undernourishment a score of 2 points and being female and being 65 years or older 1 point each. If the cut-off score to detect a low FFMI was set at ⱖ2, the sensitivity in detecting undernutrition increased from 59.3% for the original Malnutrition Universal Screening Tool to 74.1% for the cardiac surgery–specific Malnutrition Universal Screening Tool (Table 2). The specificity decreased from 82.7% to 70.1%. In contrast to the original Short Nutritional Assessment Questionnaire and Malnutrition Universal Screening Tool, the cardiac surgery⫺specific Malnutrition Universal Screening Tool was associated with a prolonged ICU and hospital stay at the operating hospital (OR: 2.1, 95% CI: 1.3 to 3.4; P⬍0.005 and OR: 1.6, 95% CI: 1.0 to 2.7; P⫽0.05, respectively). Undernutrition was present in 33.6% (n⫽108) of patients

December 2011 ● Journal of the AMERICAN DIETETIC ASSOCIATION

1927

according to the cardiac surgery⫺specific Malnutrition Universal Screening Tool. DISCUSSION As hypothesized, the accuracy in detecting undernutrition prior to cardiac surgery measured by a low FFMI was higher for the Malnutrition Universal Screening Tool than for the Short Nutritional Assessment Questionnaire. The accuracy of the Malnutrition Universal Screening Tool further improved when the patient characteristics of sex and age were integrated into the screening process. This modified version of the Malnutrition Universal Screening Tool, the cardiac surgery–specific Malnutrition Universal Screening Tool, but not the original Malnutrition Universal Screening Tool or Short Nutritional Assessment Questionnaire, was associated with a higher risk for a prolonged postoperative stay in the ICU and hospital. According to the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire, 20.9% and 7.5% of the 325 cardiac surgery patients were preoperatively classified as undernourished, respectively.These rates were consistent with those found in other surgical populations (8,22). Using BMI ⱕ21.0 and/or weight loss ⱖ10% in the preceding 6 months, 9.0% of our cardiac surgery patients were found to be undernourished. This was equal to the prevalence found in a previous study in cardiac surgery patients (3). If a low FFMI was added to weight loss and BMI to define undernutrition, 13% of cardiac surgery patients were classified as undernourished. The suggestion that the Malnutrition Universal Screening Tool overestimates and the Short Nutritional Assessment Questionnaire underestimates the presence of disease-related undernutrition is in accordance with the low sensitivity found for the Short Nutritional Assessment Questionnaire and low specificity found for the Malnutrition Universal Screening Tool. The lower sensitivity of the Short Nutritional Assessment Questionnaire compared to the Malnutrition Universal Screening Tool can be logically explained by the fact that the Short Nutritional Assessment Questionnaire does not include low BMI (16). The lower specificity of the Malnutrition Universal Screening Tool may be caused by the high variability of interpreting and high scoring of its question about nutritional intake (23). In comparison to subjective global assessment—incorporating unintended weight loss, physical examination, and clinical history—similar sensitivity of 61% and specificity of 76% was observed for the Malnutrition Universal Screening Tool in a mixed hospital population at admission (21). Compared to unintended weight loss and/or a BMI ⬍18.5 a sensitivity of 79% and a specificity of 83% was observed for the Short Nutritional Assessment Questionnaire (7). This large difference in sensitivity of the Short Nutritional Assessment Questionnaire between the Short Nutritional Assessment Questionnaire study (7) and this study in cardiac surgery patients (79% and 18%, respectively) can be explained by the fact that the reference standard of the Short Nutritional Assessment Questionnaire study— unintended weight loss and/or BMI ⬍18.5— does not fit the reference standard of this study, ie, a low FFMI (3,4). In a cardiac surgery population, a reference standard for undernutrition, including a low FFMI seems more appropriate (4).

1928

December 2011 Volume 111 Number 12

In contrast to recently published studies in other hospital populations (24-26), no association was found between Malnutrition Universal Screening Tool or Short Nutritional Assessment Questionnaire and adverse outcomes after cardiac surgery. In light of their relatively low sensitivity rates, this is easily understood. The accompanying misclassifications resulted in both the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire not being sensitive enough to predict adverse outcomes, although their reference standard, ie, low FFMI, is (4). Furthermore, it should be emphasized that nutritional screening tools that include disease severity will have an advantage on the unadjusted association with adverse outcomes (23-25). Their higher predictive values do not necessarily mean that they better identify those patients benefiting from nutritional intervention. This study demonstrated that the sensitivity of both the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire in detecting low FFMI in patients undergoing cardiac surgery was low (59% and 19%, respectively). It is clear that misclassifying should be prevented. Ideally, all cardiac surgery patients with low FFMI should be identified and receive intervention. Because actually measuring FFMI in all cardiac surgery patients increases costs, workload, and level of expertise required, a quick and easy nutritional screening tool to detect those cardiac surgery patients who are highly likely to have a low FFMI is strongly recommended. This study demonstrated that the Malnutrition Universal Screening Tool extended with the highrisk patient characteristics of age and sex adds additional benefit. The fact that the original Malnutrition Universal Screening Tool was already associated with severity of heart failure and operative risk could explain why the accuracy of a cardiac surgery–specific Malnutrition Universal Screening Tool did not further improve with the addition of a score for severity of heart failure, operative risk, or procedure. Larger studies are now needed to evaluate the accuracy of this cardiac surgery–specific nutritional screening tool. For now, in clinical practice it remains to be decided whether FFMI needs to be measured in all patients. This study had some limitations. FFM was measured with bioelectrical impedance. In healthy populations, bioelectrical impedance measures FFM accurately (27). There are three different bioelectrical impedance techniques, namely single-frequency biolectric impedance analysis, multifrequency biolectrical impedance analysis, and bioelectrical impedance spectroscopy. All three bioelectrical impedance techniques make assumptions (eg, a constant hydration factor of FFM of 0.738) that can be violated in disease states including obesity (12,27,28). Of the three bioelectrical impedance techniques, bioelectrical impedance spectroscopy has most potential to provide accurate FFM measurements in illness and fluid shifts (27,28). Instead of fixed frequencies and population-dependent regression equations, bioelectrical impedance spectroscopy measures impedance at a range of nonfixed frequencies and uses physical models. In disease states, the optimal frequency of the electrical signal required to travel through total body water (ie, extra- and intracellular space) differs (27). Using a range of nonfixed fre-

quencies results in a subject-specific characteristic frequency, which is most likely better at estimating total body water and thereby FFM. However, in cardiac surgery, bioelectrical impedance spectroscopy results should be interpreted with caution because these patients tend to have higher BMI and extracellular fluid imbalances are expected because of underlying cardiac disease. It was observed in this specific population that severe obesity (BMI ⬎35) was present in 5% and decompensatio cordis in 8%. This may have resulted in some bias and overestimation of the metabolically active part of FFM and thereby underestimation of undernutrition. In contrast, no further bias was expected in the stable chronic heart failure patients because others have observed similar extracellular water: intracellular water ratios compared to healthy controls (0.73 vs 0.75) (29). Compared to dual-energy x-ray absorptiometry scanning, it was observed in cardiac surgery patients that bioelectrical impedance spectroscopy assessed FFM showed very good association with dual-energy x-ray absorptiometry (r⫽0.97, P⬍0.01) and only slightly overestimated FFM by 2.3 kg but with a wide inter-individual variation (95% CI: ⫺3.5 to 8.1 kg) (14). Whether bioelectrical impedance spectroscopy or dual-energy x-ray absorptiometry is the instrument of choice to measure FFMI is still the subject of debate (14). Another limitation of this study may be that the reference standard for undernutrition does not include unintended weight loss and low BMI (3,4,30). However, post-hoc analyses showed similar sensitivity rates for the Malnutrition Universal Screening Tool and Short Nutritional Assessment Questionnaire aggregating low FFMI with unintended weight loss and low BMI (57% and 29%, respectively). In line with this, it should be emphasized that, at present, agreement among experts about the optimal definition and operationalism of undernutrition is lacking (31-33). Moreover, BMI is only a blunt tool for measuring overall body fatness. CONCLUSIONS Accuracy in detecting undernutrition measured by low FFMI prior to cardiac surgery was considerably higher for the Malnutrition Universal Screening Tool than for the Short Nutritional Assessment Questionnaire, although still marginal, at 59% and 19%, respectively. This means that approximately 80% of cardiac surgery patients with a low FFMI will be misclassified as wellnourished if screened using the Short Nutritional Assessment Questionnaire. These misclassified patients do not receive nutritional intervention. In addition, although less, still 40% of cardiac surgery patients will be misclassified as well-nourished if screened using the Malnutrition Universal Screening Tool. Nevertheless, although absolutely not optimal, even screening for undernutrition with the Short Nutritional Assessment Questionnaire remains valuable compared to no screening at all. The Short Nutritional Assessment Questionnaire does identify patients with unintended weight loss, which is an important well-established parameter of undernutrition in both surgical and chronic heart failure patients (30,34,35). However, to reduce misclassifications and postoperative adverse outcomes, it is recommended to integrate FFMI measurements in addition to unintended weight loss and low BMI only to identify and to refer the

undernourished. This screening should take place several weeks prior to cardiac surgery. To reduce extra workload and costs of implementing the bioelectrical impedance spectroscopy, further research on quick and easy undernutrition screening tools, such as the cardiac surgery– specific Malnutrition Universal Screening Tool, integrating age and sex, is recommended. Preliminary results demonstrated a relatively high sensitivity of 74% for the cardiac surgery–specific Malnutrition Universal Screening Tool. Ideally, nurses screen for undernutrition. If positively screened, referral to the registered dietitian for further diagnostic assessment and intervention takes place. Screening for undernutrition should be repeated at the time of hospital admission and weekly after surgery in combination with monitoring nutritional intake. Consequently, the waiting period and also the early- and longterm postoperative phases can be used to increase patients’ reserve capacity of metabolically active FFM by means of nutritional interventions combined with exercise programs to optimally respond to operative stress. Studies assessing the cost-effectiveness of pre- and postoperative exercise programs combined with nutritional interventions in the undernourished cardiac surgery patients on body composition and adverse outcome level are now needed. STATEMENT OF POTENTIAL CONFLICT OF INTEREST: No potential conflict of interest was reported by the authors. References 1. Engelman DT, Adams DH, Byrne JG, Aranki SF, Collins JJ Jr, Couper GS, Allred EN, Cohn LH, Rizzo RJ. Impact of body mass index and albumin on morbidity and mortality after cardiac surgery. J Thorac Cardiovasc Surg. 1999;118:866-873. 2. van Straten AH, Bramer S, Soliman Hamad MA, van Zundert AA, Martens EJ, Schonberger JP, de Wolf AM. Effect of body mass index on early and late mortality after coronary artery bypass grafting. Ann Thorac Surg. 2010;89:30-37. 3. van Venrooij LM, de Vos R, Borgmeijer-Hoelen AM, Haaring C, de Mol BA. Preoperative unintended weight loss and low body mass index in relation to complications and length of stay after cardiac surgery. Am J Clin Nutr. 2008;87:1656-1661. 4. van Venrooij LM, de Vos R, Zijlstra E, Vulperhorst L, BorgmeijerHoelen AM, van Leeuwen PA, de Mol BA. The impact of low preoperative fat-free body mass on infections and length of stay after cardiac surgery: A prospective cohort study. J Thorac Cardiovasc Surg. 2011;142:1263-1269. 5. van Venrooij LM, Verberne HJ, De Vos R, Borgmeijer-Hoelen AM, Van Leeuwen PA, De Mol BA. Postoperative loss of skeletal muscle mass, complications and quality of life in patients undergoing cardiac surgery [published online ahead of print May 27, 2011]. Nutrition. doi:10.1016/j.nut.2011.02.007. 6. Kondrup J, Allison SP, Elia M, Vellas B, Plauth M. ESPEN Guidelines for Nutrition Screening 2002. Clin Nutr. 2003;22:415-21. 7. Kruizenga HM, Seidell JC, de Vet HC, Wierdsma NJ, van Bokhorst-de van der Schueren MA. Development and validation of a hospital screening tool for malnutrition: The Short Nutritional Assessment Questionnaire (SNAQ). Clin Nutr. 2005;24:75-82. 8. Stratton RJ, Hackston A, Longmore D, Dixon R, Price S, Stroud M, King C, Elia M. Malnutrition in hospital outpatients and inpatients: Prevalence, concurrent validity and ease of use of the ’malnutrition universal screening tool’ (’MUST’) for adults. Br J Nutr. 2004;92:799808. 9. Kruizenga HM, Van Tulder MW, Seidell SJ, Thijs A, Ader HJ, van Bokhorst-de van der Schueren MA. Effectiveness and cost-effectiveness of early screening and treatment of malnourished patients. Am J Clin Nutr. 2005;82:1082-1089. 10. van Venrooij LM, de Vos R, Borgmeijer-Hoelen AM, Kruizenga HM,

December 2011 ● Journal of the AMERICAN DIETETIC ASSOCIATION

1929

11.

12. 13. 14.

15.

16. 17. 18.

19. 20. 21. 22.

Jonkers-Schuitema CF, de Mol BA. Quick-and-easy nutritional screening tools to detect disease-related undernutrition in a hospital in- and outpatient setting: A systematic review of sensitivity and specificity. Eur J Clin Nutr Metab. 2007;2:21-37. Oroepoulos A, Ezekowitz JA, McAlister FA, Kalantar-Zadeh K, Fonarow GC, Norris CM, Johnson JA, Padwal RS. Association between direct measures of body composition and prognostic factors in chronic heart failure. Mayo Clin Proc. 2010;85:609-617. BodyScout. Xitron Hydra ECF/ICF (Model 4200) Bio-Impedance Spectrum Analyzer. Operating Manual Revision 1.03.2007. San Diego, CA: Xitron Technologies, Inc. Lukaski HC, Johnson PE, Bolonchuk WW, Lykken GI. Assessment of fat-free mass using bioelectrical impendance measurements of the human body. Am J Clin Nutr. 1985;41:810-817. van Venrooij LM, Verberne HJ, de Vos R, Borgmeijer-Hoelen AM, van Leeuwen PA, de Mol BA. Preoperative and postoperative agreement in fat free body mass (FFM) between the bioelectrical impedance spectroscopy (BIS) method and the dual energy x-ray absorptiometry (DXA) in patients undergoing cardiac surgery. Clin Nutr. 2010;29: 789-794. van Itallie TB, Yang MU, Heymsfield SB, Funk RC, Boileau RA. Height-normalized indices of the body’s fat-free mass and fat mass: Potentially useful indicators of nutritional status. Am J Clin Nutr.1990;52:953-959. Kyle UG, Schutz Y, Dupertuis YM, Pichard C. Body composition interpretation: Contributions of the fat free mass index and the body fat mass index. Nutrition. 2003;19:597-604. Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg. 1999;16:9-13. Nashef SA, Roques F, Hammill BG, Peterson ED, Michel P, Grover FL, Wyse RK, Ferguson TB. Validation of European System for Cardiac Operative Risk Evaluation (EuroSCORE) in North American cardiac surgery. J Thorac Cardiovasc Surg. 2002;22:101-105. The Society of Thoracic Surgeons. STS Data Collection. version 2.61. 2007. http://www.sts.org/sites/default/files/documents/pdf/Adult CVDataSpecifications2.61.pdf. Accessed May 2008. Altman DG. Diagnostic tests. In: Practical Statistics for Medical Research. London, UK: Chapman and Hall; 1999:409-419. American Dietetic Association. Nutrition Screening. 2011. www.adaevidencelibrary.com/topic.cfm?format_tables⫽0&cat⫽3958. Accessed March 2011. Leistra E, Neelemaat F, Evers A, Kruizinga HM. Prevalence of undernutrition in Dutch hospital outpatients. Eur J Int Med. 2009;20: 509-513.

1930

December 2011 Volume 111 Number 12

23. Kyle UG, Kossovsky MP, Karsegard VL, Pichard C. Comparison of tools for nutritional assessment and screening at hospital admission: A population study. Clin Nutr. 2006;25:409-417. 24. Raslan M, Gonzalez MC, Goncalves MC, Nascimento M, Castro M, Marques P, Segatto S, Torrinhas RS, Cecconello I, Waitzberg DL. Comparison of nutritonal risk screening tools for predicting clinical outcomes in hospitalized patients. Nutrition. 2010;26:721-726. 25. Schiesser M, Kirchhoff P, Muller MK, Schafer M, Clavien P. The correlation of nutritional risk index, nutritional risk score, and bioimpedance analysis with postoperative complications in patients undergoing gastrointestinal surgery. Surgery. 2009;145:519-526. 26. Ozkalkanli MY, Ozkalkanli DT, Katircioglu K, Savaci S. Comparison of tools for nutrition assessment and screening for predicting the development of complications in orthopedic surgery. Nutr Clin Pract. 2009;24:274-280. 27. Earthman C, Traughber D, Dobratz J, Howell W. Bioimpedance spectroscopy for clinical assessment of fluid distribution and body cell mass. Nutr Clin Pract. 2007;22:389-405. 28. Kyle UG, Bosaeus I, De Lorenzo AD, Deurenberg P, Elia M. Bioelectrical impedance analysis—Part 2: Utilization in clinical practice. Clin Nutr. 2004;23:1430-1453. 29. Uszko-Lencer NH, Bothmer F, Pol van PE, Schols AM. Measuring body composition in chronic heart failure: A comparision of methods. Eur J Heart Fail. 2006;8:208-214. 30. Anker SD, Ponikowski P, Varney S, Chua TP, Clark AL, Webb-Peploe KM, Harrington D, Kox WJ, Poole-Wilson PA, Coats AJ. Wasting as independent risk factor for mortality in chronic heart failure. Lancet. 1997;349:1050-1053. 31. Jensen GL, Mirtallo J, Compher C, Dhaliwal R, Forbes A, Figueredo Grijalba R, Hardy G, Kondrup J, Labadarios D, Nyulasi I, Castillo Pineda J, Waitzberg D. Adult starvation and disease-related malnutrition: A proposal for etiology-based diagnosis in the clinical practice setting from the International Consensus Guideline Committee. JPEN. 2010;34:156-159. 32. Meijers JM, Van Bokhorst-de van der Schueren MA, Schols JM, Soeters PB, Halfens RJ. Defining malnutrition: Mission or mission impossible? Nutrition. 2010;26:432-440. 33. Soeters PB, Schols AM. Advances in understanding and assessing malnutrition. Curr Opin Clin Nutr Metab Care. 2009;12:487-494. 34. Detsky AS, Smalley PS, Chang J. Is this patient malnourished? JAMA. 1994;271:54-58. 35. Van Bokhorst-de van der Schueren MA, van Leeuwen PA, Sauerwein HP, Snow GB, Quak JJ. Assessment of malnutrition parameters in head and neck cancer and their relation to postoperative complications. Head Neck. 1997;19:419-425.