Ten days before being treated, a 65-year-old man experienced a Wallenberg stroke involving the lateral medulla, which left him with difficulty swallowing. Since then, he had been treated with intravenous fluids, as attempts at eating led to mild aspiration with pneumonia. In that period, he lost 6% of his usual body weight and was continuing to lose weight. He felt weak and was able to ambulate only with difficulty because of his stroke-related ataxia and generalized weakness. On physical examination, there was an obvious squared-off appearance to his shoulders from subcutaneous tissue and muscle wasting. There was no edema.
A 63-year-old man was admitted to the hospital for gastric resection of an obstructing gastric carcinoma. He was well until 6 weeks before admission, when he began to notice the rapid onset of early satiety. This progressed to the point where he began to vomit virtually all food and fluids. He had lost 15% of his body weight and was continuing to lose weight. He was ambulatory but felt weak and was no longer able to carry on his usual daily activities because of this weakness. On physical examination, there was muscle wasting. There was obvious subcutaneous tissue loss in the triceps and thoracic regions, as well as muscle loss in the deltoids. There was edema in his ankles but no ascites.
A 70-year-old man was admitted to the hospital for resection of his descending colon because of an adenocarcinoma detected on investigation for bright-red blood in his bowel movements. Between 6 and 3 months before admission, he had lost 10% of his body weight for reasons that he could not explain. However, his weight had stabilized in the 2 months before admission, and in fact, he had gained back 4% of his weight. His dietary intake had been slightly below normal but had recently improved. He reported no significant gastrointestinal (GI) symptoms other than the bleeding and a mild change in his bowel habits. He had his usual level of energy. On physical examination, there was no evidence of subcutaneous tissue loss, muscle wasting, edema, or ascites.
Why Perform Nutritional Status Assessment?
Malnutrition occurs among patients either because of their primary diseases (eg, malignancy) or because the procedures they undergo to treat the primary disease prevent them from receiving adequate nutritional intake for prolonged periods (eg, surgery).
There are 2 components of nutritional status assessment. The first is body composition analysis, which is the determination of the mass of body components, such as total body protein and total body fat. These components are measured by in vivo neutron activation analysis and tritiated water dilution technique, which represents the criterion standard (also known as the gold standard) for measures of body composition. The second component is physiologic function, defined by some as changes in cellular and organ function, measured in a variety of ways, such as skeletal muscle strength, respiratory function, protein synthesis, and tissue repair.
During the past 3 decades, clinicians have become increasingly aware of the prevalence of malnutrition among hospitalized patients.1-4 Clinicians have recognized that malnourished patients are at a higher risk of developing complications while undergoing treatment. These complications include death, sepsis, abscess formation, other infections such as pneumonia, wound healing difficulties postoperatively, and respiratory failure. Some have used the term nutrition-associated complications5, 6 to highlight the relationship between malnutrition and these adverse events. The increased risk for malnourished patients is thought to be caused more by functional impairment than changes in body composition,7 although in studied subjects there is clearly a correlation between the 2 components of nutritional status.
Investigations in the 1970s1, 2 estimated that the prevalence of malnutrition among hospitalized patients was as high as 40%. Studies4, 8 on patients undergoing general GI surgery showed that the prevalence of either mild or severe malnutrition was 48%3 and 31%, respectively. Detsky et al4 confirmed the relationship between malnutrition and the risk of nutrition-associated complications. In their series of 202 patients undergoing general GI surgery at 2 Toronto (Ontario, Canada) teaching hospitals, 10% of the total series of patients had major nutrition-associated complications, including 6 deaths related to sepsis, 2 nonfatal episodes of sepsis, 3 subphrenic or intra-abdominal abscesses, 2 anastomotic breakdowns, 2 wound dehiscences, and 5 major wound abscesses. However, among those who were assessed to be severely malnourished preoperatively, this major complication rate was 67%. Windsor and Hill,7 using a slightly different system of nutritional status assessment in 102 patients undergoing major GI surgery, also showed that severely malnourished patients had a higher risk of major complications than patients designated as having normal nutritional status. These results confirm the usefulness of nutritional status assessment as a predictor of high risk for postoperative complications. Thus, it becomes both a method of assessing prognosis and a method of diagnosing a particular health state. Furthermore, assessing nutritional status identifies patients who may benefit from enteral or parenteral nutritional repletion to reduce the risk of these complications.9-11 Although patients with chronic medical conditions also are thought to be at higher risk of developing complications, such as respiratory failure or infection, most of what we know comes from patients undergoing surgical procedures.
The Anatomic/Physiologic Origin of Findings in This Area
Syndromes of undernutrition of calories and protein have been studied most extensively in children of developing nations and are not frequently observed in North America. Two extremes of protein-energy malnutrition have been defined: marasmus, caused primarily by deficiency of calories, resulting in stunted growth in children, loss of adipose tissue, and generalized wasting of lean body mass without edema; and kwashiorkor, a primary deficiency of protein manifested by edema but in which adipose tissue is preserved.
Many individuals who are malnourished will have elements of both protein and calorie deficiencies. The complex metabolic processes that result from protein-energy malnutrition are beyond the scope of this overview. However, in North America, nutritional assessment is used as a predictor of future complications in patients and therefore may go beyond the traditional measurement of pure malnutrition resulting from inadequate intake of protein, calories, or micronutrients. Nutritional assessment, particularly if it encompasses or focuses on physiologic function, may be an overall marker of illness that is not caused solely by inadequate intake or reversed by nutritional supplementation. This may explain why the clinical trials of total parenteral nutrition in a variety of clinical circumstances have in some cases produced disappointing results in improving outcomes.12
Nutritional deficiency syndromes involving vitamins and micronutrients evolve through 3 stages because most micronutrients are stored in tissues, and a temporary reduction in intake is buffered by a reduction in body stores. The second stage involves metabolic changes without symptoms, whereas severe depletion will result in the final stage of clinical signs and symptoms. They will not be discussed in this article.
How to Perform Nutritional Assessment
This article primarily describes features of the medical history and physical examination for assessing overall nutritional status.
The relevant features of a patient's medical history and physical examination can be elicited by a technique known as the subjective global assessment (SGA) of nutritional status.8 The application of this technique divides patients into 3 classes: class A, well nourished; class B, moderately (or suspected of being) malnourished; and class C, severely malnourished. The components of this technique are described in Table 28-1. There are 4 elements of the medical history.
Table 28-1.Features of Subjective Global Assessmenta |Favorite Table|Download (.pdf) Table 28-1. Features of Subjective Global Assessmenta
|Medical History |
|1. Weight change |
| Overall loss in past 6 months: amount = ________ kg; ________ % |
| Change in past 2 weeks: ||________ increase || || || |
| ||________ no change || || || |
| ||________ decrease || || || |
|2. Dietary intake change (relative to normal) |
| ________ no change || || || || |
| ________ change ||________ duration = ||________ weeks || || |
| ||________ type: ||________ suboptimal solid diet ||________ full liquid diet || |
| || ||________ hypocaloric liquids ||________ starvation || |
|3. Gastrointestinal symptoms (that persisted for > 2 weeks) |
| ________ none ||_______ nausea ||________ vomiting ||________ diarrhea ||________ anorexia |
|4. Functional capacity |
| ________ no dysfunction (eg, full capacity) || || || |
| ________ dysfunction ||________ duration = ||________ weeks || || |
| ||________ type: ||________ working suboptimally || || |
| || ||________ ambulatory || || |
| || ||________ bedridden || || |
|Physical (for each trait specify: 0 = normal, 1+ = mild, 2+ = moderate, 3+ = severe) |
| ________ loss of subcutaneous fat (triceps, chest) |
| ________ muscle wasting (quadriceps, deltoids) |
| ________ ankle edema |
| ________ sacral edema |
| ________ ascites |
|Subjective global assessment rating (select one)a |
| ________ A = well nourished |
| ________ B = moderately (or suspected of being) malnourished |
| ________ C = severely malnourished |
1. Weight Loss in the 6 Months Before the Examination, Expressed as a Proportionate Loss From Previous Weight
A weight loss of less than 5% is considered small. A weight loss between 5% and 10% is considered potentially significant, and a weight loss of more than 10% is considered definitely significant. In addition to considering the amount of weight loss, it is important to note the pattern of the weight loss. For example, suppose a patient lost 12% of his or her weight in the 6 months to 1 month before the examination and then regained half of that weight in the subsequent month, resulting in a net loss of 6% for the entire period. This patient would be considered better nourished than a patient who had lost 6% progressively in the 6 months, with continued weight loss in the recent weeks, before the examination. Patients can be considered well nourished despite significant proportions of weight loss if there has been a recent stabilization or increase in weight. In eliciting the history of weight pattern from patients, we recommend asking the patient what his or her maximum weight was and what it was 1 year ago, 6 months ago, 1 month ago, and at present. If patients report substantial weight loss that we cannot confirm with prior records, we ask for confirming history of a change in clothing size or whether their clothes now fit very loosely. Finally, we ask for the pattern of the weight loss during the past few weeks (continued loss, stabilization, or gain).
2. Dietary Intake in Relation to the Patient's Usual Pattern
Patients are classified as having either normal or abnormal (decreased) intake in the weeks to months before the examination. The duration and degree of abnormality are also noted (eg, starvation, hypocaloric liquids, full liquid diet, or suboptimal solid diet). For example, patients with strokes resulting in swallowing difficulties may have been starved, simply receiving intravenous or hypocaloric fluids for several weeks before the examination. Patients with lesions that obstruct the outflow from the stomach, such as cancer or severe ulcers, may have been receiving pure liquid diets. In eliciting this history, we recommend asking patients whether their eating patterns have changed during the past few weeks and then ask if their pattern has changed during the past few months. Has the amount of food eaten decreased? If so, by how much? Are there certain kinds of foods that they used to eat that they can no longer eat? Why are they eating less (intentional reduction, unintentional reduction, ordered by clinician)? What happens if they try to eat more? Ask for an example of a typical breakfast, lunch, and dinner and a comparison with typical meals 6 to 12 months ago.
3. Presence of Significant Gastrointestinal Symptoms: Anorexia, Nausea, Vomiting, and Diarrhea
By significant we mean that these symptoms must have persisted on virtually a daily basis for a period longer than 2 weeks. Short-term diarrhea or intermittent vomiting is not considered significant. Daily or twice-daily vomiting secondary to obstruction is considered significant.
4. The Patient's Functional Capacity or Energy, Ranging From Full Capacity to Bedridden
Patients who are unable to eat will often complain of fatigue and weakness to the point at which they are bedridden.
There are 3 features of the physical examination that are recorded as normal (0), mild (1+), moderate (2+), or severe (3+).
1. Loss of Subcutaneous Fat
There are several locations where one can look for loss of subcutaneous fat, and the best are the triceps region of the arms, the midaxillary line at the costal margin, the interosseous and palmar areas of the hand, and the deltoid regions of the shoulder (Figures 28-1 and 28-2). Positive findings are loss of fullness or 1 or more areas where the skin fits too loosely over the deeper tissues; this latter sign may be falsely positive in elderly individuals who may appear to have lost subcutaneous tissue without being clinically malnourished.
Loss of Subcutaneous Tissue in the Arm and Chest Wall
Loss of Subcutaneous Tissue Overlying the Fifth Metacarpal
Hand with tissue loss (left) vs healthy hand (right).
The best muscles to examine are the quadriceps femoris and deltoids. In the deltoid region, malnourished patients have a squared-off appearance to their shoulders from the combination of muscle and subcutaneous tissue loss (Figure 28-3). In severe malnutrition, the quadriceps will have loss of bulk and tone. Obviously, neurologic lesions (that may present with unilateral wasting) may produce false-positive findings here.
Loss of Subcutaneous Tissue in the Shoulders, Giving a Squared-off Appearance
3. Loss of Fluid From the Intravascular to Extravascular Space, Namely, Ankle or Sacral Edema and Ascites
The first 2 signs are best assessed by inspection and then by palpation, remembering that some features are best inspected from a distance, eg, squared-off shoulders. Edema is assessed by pressing the ankle (leg) or sacrum, feeling the fluid move out of the subcutaneous tissue, and then observing “pitting,” persistent depression of the area pressed (more than 5 seconds).
There is no explicit numeric weighting scheme described for combining these features of the history and physical examination into an SGA. Rather, they are combined subjectively into an overall or global assessment. In the study that established the precision and accuracy of SGA,4, 8 clinicians placed greatest importance on the following variables: weight loss of more than 10%, poor dietary intake, loss of subcutaneous tissue, and muscle wasting. Patients suspected of being malnourished or judged to have moderate malnourishment (class B) had lost at least 5% of their body weight in the weeks before examination without stabilization or weight gain, had a definite history of reduction in dietary intake, and exhibited mild (1+) loss of subcutaneous tissue. When patients had considerable edema, ascites, or tumor mass, less attention was paid to the amount of weight loss. The other historical features helped the clinicians confirm the patient's self-report of weight loss or dietary change but received less weight in the ranking system.
If, on the other hand, a patient had a recent weight gain that did not appear to be merely fluid retention, clinicians designated that patient well nourished (class A), even if the net weight loss was between 5% and 10% and there was mild loss of subcutaneous tissue. The assignment of a class A rank also should occur in settings in which the patient has had an improvement in the other historical features of SGA, such as appetite.
To be classified as severely malnourished (class C), patients should demonstrate obvious physical signs of malnutrition, such as severe (3+) subcutaneous tissue loss and muscle wasting, often with edema, in the presence of a clear and convincing pattern of ongoing weight loss of at least 10%.
By design, this system is less sensitive and more specific. That is, few well-nourished patients will receive a false-positive diagnosis of malnourishment, but some patients with mild degrees of malnutrition may be missed.
Windsor and Hill7 describe a slightly different system of nutritional status that focuses more on physiologic function. Their system has 2 components: weight loss and functional status. Preoperative percentage of weight loss is defined as (recalled well weight minus current measured weight) divided by well weight. A weight loss of more than 10% during the preceding 3 months was considered significant. Confirmation of weight loss is sought in the physical examination by palpating skin folds for loss of fat and muscles in a manner similar to that just described, functional impairment of overall activity levels (by observing the patient on the ward), overall mood (alertness, ability to concentrate, and irritability), skeletal muscle function (having the patient squeeze the examiner's hand), respiratory function (effort and sound of coughing and shortness of breath), wound healing (unhealed wounds and sores or scratches or skin sepsis), and serum albumin level of less than 3.2 g/dL. If patients have weight loss of less than 10%, with no evidence of abnormal physiologic function, then they are placed in group 1. With weight loss of more than 10% but no abnormal physiologic function, patients are placed in group 2, and with both features, they are placed in group 3.
Before you read further, we suggest that you return to the patient scenarios that opened this overview and decide whether you judge them to be well nourished, moderately malnourished, or severely malnourished using SGA. After doing so, read on.
The patient in case 1 was moderately malnourished (class B). This ranking was determined by his continuing loss of weight, the limitation of nutritional intake to hypocaloric fluids for 2 weeks, and the mild loss of subcutaneous tissue and muscle.
The patient in case 2 was severely malnourished (class C). This judgment was most influenced by his continuing large weight loss, change in dietary intake, and positive physical findings.
The patient in case 3 was well nourished (class A). Although he had experienced considerable weight loss at some time before admission, his weight had stabilized and increased just before admission.
Precision of the Assessment of Nutritional Status
Investigators at the University of Toronto studied 202 patients at 2 teaching hospitals who were undergoing major GI surgery.8 A nurse and 3 residents learned the technique of nutritional status described herein by examining a series of patients and reviewing their assessments with those of a senior clinician. The emphasis was on combining the symptoms and signs of malnutrition to minimize the false-positive diagnosis of malnutrition (high specificity) at the expense of increasing false-negative results (lower sensitivity). After reviewing several patients together, the nurse and one of the 3 residents performed duplicate, independent assessments of 109 patients. There was perfect agreement in 100 (91%) of 109 patients on the SGA rankings. This was 78% above the agreement that could be expected by chance alone (the κ statistic was 0.78, with SE = 0.08 and 95% confidence interval ranging from 0.62 to 0.94). The κ statistics for the 3 pairings of the nurse with the individual residents were 0.60, 0.81, and 1.0, respectively, revealing some variation in agreement between different clinicians. Hirsch et al13 also documented 79% concordance between SGA rankings of residents and specialists in clinical nutrition.
Accuracy of Nutritional Assessment
Because there is no criterion standard for the diagnosis of malnutrition that incorporates body composition and physiologic function (the in vivo neutron activation analysis and titrated water technique are the criterion standards of body composition alone), studies of the accuracy of techniques of nutritional status assessment have related it to the development of complications judged to result from malnutrition. Therefore, patients are sorted into the columns of the usual 2 × 2 table based on whether they develop malnutrition-associated complications.
The study by Detsky et al4 provides useful data on the accuracy of SGA (Table 28-2). Nineteen patients (10% of the total studied) were classified as severely malnourished (class C), 44 (21%) were classified as moderately (or suspected of being) malnourished (class B), and 139 (69%) were classified as well nourished (class A). The likelihood ratios in this table show that the SGA is a powerful predictor of postoperative complications. The likelihood ratio greater than 4 for severely malnourished patients means that this designation (class C) was more than 4 times as likely to be found in patients with, as opposed to patients without, postoperative complications. Patients designated as moderately (or suspected of being) malnourished (class B) generated a likelihood ratio of close to unity, indicating no clinically important change between the preexamination and postexamination probability of postoperative complications (20/202, or 10%). Finally, well-nourished patients (class A) generated likelihood ratios of 0.66 for their admission SGA and only 0.38 for their minimum SGA, indicating a lower than average risk of postoperative complications.
Table 28-2.Relationship Between Subjective Global Assessment (SGA) and Major Postoperative Complicationsa |Favorite Table|Download (.pdf) Table 28-2. Relationship Between Subjective Global Assessment (SGA) and Major Postoperative Complicationsa
|SGA Class ||Patients Assigned Class on Admission, No. (%) ||Major Complicationsb Occurring in This Class, No. (%) ||Likelihood Ratio for Admission SGA ||Likelihood Ratio for Minimum SGA During Hospitalization |
|Severely malnourished ||19 (10)b ||8 (42) ||4.4 ||4.1 |
|Moderately (or suspected of being) malnourished ||44 (21) ||4 (9) ||0.96 ||0.93 |
|Well nourished ||139 (69) ||8 (5) ||0.66 ||0.38 |
SGA performed better than objective measurements of the physical examination, such as percentage of ideal weight on admission and percentage of body fat calculated from anthropometric measurements. The range of likelihood ratios for these variables displayed considerably less accuracy than those associated with SGA and the combination of SGA (Table 28-3).
Table 28-3.Predictive Properties of Unpromising Techniquesa
Laboratory determination of serum albumin level was also shown to be an accurate predictor of complications, associated with a progression of likelihood ratios that is similar to that of SGA. Moreover, the combination of SGA and albumin provided slightly improved accuracy compared with either method alone. However, other objective methods that are frequently said to be useful techniques of assessing nutritional status (serum transferrin level, creatinine-height index, and total lymphocyte count) were not shown to be accurate predictors of complications.4
The study by Windsor and Hill7 provides similar data demonstrating the predictive validity of their system. Of the 102 patients, 43 (42%) were in group 1 (analogous to SGA class A), 17 (17%) were in group 2 (analogous to SGA class B), and 42 (41%) were in group 3 (analogous to SGA class C). The rate of major complications, septic complications, and pneumonia in the 3 groups was significantly different, and the likelihood ratios for predicting major complications showed a similar progression to SGA of 0.53, 0.69, and 1.8 for groups 1, 2, and 3, respectively.
Finally, the predictive validity of SGA was also reported by the Veterans Affairs perioperative total parenteral nutrition randomized trial11 that enrolled only patients with various degrees of malnutrition. Among the control patients, those in SGA class C had higher rates of major infectious complications and noninfectious complications.
Some have also reported the high correlation between SGA and other measures of nutritional status assessments that are thought to be more objective, such as anthropometry,3, 13 albumin level,3, 13 total serum protein level,3, 13 and criterion standard measures of body composition. Windsor and Hill7 also show good correlations between their system and anthropometry, body composition, and objective measures of the physiologic functions in their method (eg, grip strength and respiratory muscle index).
Are These Symptoms or Signs Ever Normal?
Many individuals are thin, and this in itself does not constitute malnutrition. However, we should note that obesity, defined as an excess of adipose tissue or by the degree to which a patient's weight exceeds that which is judged ideal by some anthropometric formula, is also a common problem in hospitalized patients. Epidemiologic studies have shown that a 20% excess over ideal weight imparts a health risk. Similarly, obesity has been shown to place patients at a high risk of experiencing surgical complications, such as poor wound healing and venous thrombosis.
Special Ways to Learn, Test Yourself, and Correct Deficiencies in the Elicitation of These Symptoms and Signs
Clinicians who wish to become competent at nutritional assessment can do so by applying the following strategies: First, they should undergo a training period with other learners, in which they discuss each of the features of the technique together and review a series of patients for each of the findings. In particular, the group should review methods of eliciting the medical history, performing the inspection, and standardizing terms such as normal, mild, moderate, and severe. Next, they should rank several patients together and reach consensus about what constitutes an A, B, or C ranking. Finally, they should perform their own tests of clinical reproducibility by treating a series of (perhaps 10) patients independently and comparing their rankings. To improve the precision and validity of their elicitation of the individual features of the SGA, they should consider verification strategies, such as asking the patient's spouse about the features of the history, examining physician records for previous weights, asking whether the patient's clothes now fit loosely, and examining recent and old pictures of the patient.
Clinicians can learn to perform SGA of nutritional status with precision. The features of the medical history and physical examination are shown in Table 28-1. We recommend the group approach to standardize the definitions of the features of the history and physical examination contained in SGA and to gain competency in their application. In doing so, we recommend that clinicians train themselves to be less sensitive and more specific in labeling patients as malnourished. Because there is no criterion standard for malnutrition that incorporates body composition and physiologic function, this clinical skill should be used as a prognostic instrument to identify patients who are at high risk of developing complications and who may benefit from nutritional repletion and support. The technique is an accurate predictor of patients who are at higher risk of developing complications such as infection or poor wound healing.
Author Affiliations at the Time of the Original Publication
From the Department of Medicine (Drs Detsky and Smalley) University of Toronto, Toronto, Ontario, Canada; Division of General Internal Medicine and Clinical Epidemiology, The Toronto Hospital, Toronto, Ontario, Canada (Dr Detsky); Department of Medicine, Oshawa General Hospital, Oshawa, Ontario, Canada (Dr Chang); and Division of General Internal Medicine, The Wellesley Hospital, Toronto, Ontario, Canada (Dr Smalley).
This study was supported in part by a National Health Research Scholar Award from Health and Welfare Canada, No. 6606-2849-48, to Dr Detsky.
et al.. Nutritional assessment: a comparison of clinical judgment and objective measurements. N Engl J Med.
et al.. Predicting nutrition-associated complications for patients undergoing gastrointestinal surgery. JPEN J Parenter Enteral Nutr.
KN. Evaluating the accuracy of nutritional assessment techniques applied to hospitalized patients: methodology and comparisons. JPEN J Parenter Enteral Nutr.
KN. The choice to treat all, some, or no patients undergoing gastrointestinal surgery with nutritional support: a decision analysis
approach. JPEN J Parenter Enteral Nutr.
GL. Weight loss with physiologic impairment: a basic indicator of surgical risk. Ann Surg.
et al.. What is subjective global assessment of nutritional status? JPEN J Parenter Enteral Nutr.
et al.. A randomized clinical trial of total parenteral nutrition
in malnourished surgical patients: the rationale and impact of previous clinical trials and pilot study on protocol design. Am J Clin Nutr.
Veterans Affairs Total Parenteral Nutrition
Cooperative Study Group. Perioperative nutrition in surgical patients. N Engl J Med.
AS. Parenteral nutrition: is it helpful? N Engl J Med.
et al.. Subjective global assessment of nutritional status: further validation. Nutrition.