§2_ _ nutrition survey
Hospital malnutrition A prospective evaluation of general medical patients during the course of hospitalization 1-3 Roland L. Weinsier, M.D., Dr.P.H., Edie M. Hunker, R.N., Carlos L. Krumdieck, M.D., Ph.D., and C. E. Butterworth, Jr., M.D.
ABSTRACT Nutrition status was evaluated in 134 consecutive admissions to a general medical service and throughout hospitalization among patients hospitalized 2 weeks or longer. Likelihood of malnutrition was determined using eight nutrition-related parameters: serum folate and vitamin C, triceps skinfold, weight/height, arm muscle circumference, lymphocyte count, serum albumin. and hematocrit. On admission 48% of patients had a high likelihood of malnutrition, which correlated with a longer hospital stay (20 versus 12 days for patients with a low likelihood of malnutrition) and an increased mortality rate (13 versus 4%). Likelihood of malnutrition increased with hospitalization in 69% of patients with paired determinations. Compared to admission, at final follow-up a greater proportion of patients fell into the depleted range of values for folate, triceps skinfold, weight/height, arm muscle circumference, lymphocyte count, and hematocrit. These parameters worsened in over 75% of patients admitted with normal values. Hematocrit fell in all patients with normal admission levels. These findings demonstrate an association between nutrition status and hospital course and a worsening trend during hospitalization. Am. J. Clin. NUlr. 32: 418-426, 1979.
Nutrition surveys in the United States and England have documented the existence of malnutrition among hospitalized patients (1-7). One report concluded that 44% or more patients on a general medical service had protein-calorie malnutrition as evidenced by the presence of one or more substandard nutrition parameters (4). Other studies indicated that hypovitaminemia occurred commonly among hospitalized patients (I, 5). Inferences drawn from such studies have been that nutrition support is neglected (2, 3, 6), that little attempt is made to reverse malnutrition (4), and that physicians lack concern for the nutritional status oftheir patients (7). Although some reports have suggested that malnutrition actually develops during the course of hospitalization (2, 6), there have been no prospective studies to indicate 418
whether the reported prevalence of malnutrition is, in fact, hospital-related or simply a reflection of the status of patients upon hospital admission. It was the intent of this study to answer the following questions; 1) does the nutrition status of medical patients deteriorate over the course of their hospitalization? and, 1) does malnutrition adversely affect the hospital 1 From the Department ofNutrition Sciences, Schools of Public and Allied Health, Medicine, and Dentistry, University of Alabama in Birmingham, Birmingham, Alabama 35294. 2 Supported by Grant 5 R18-CA-17928-01, awarded by the National Cancer Institute, Department of Health, Education and Welfare. 3 Address reprint requests to: Roland L. Weinsier, M.D., Dr.P.H., Department of Nutrition Sciences, University of Alabama, Birmingham, Alabama 35294.
The American Journal of Clinical Nutrition 32: FEBRUARY 1979, pp. 418-426. Printed in U.S.A.
Downloaded from https://academic.oup.com/ajcn/article-abstract/32/2/418/4666289 by St Bartholomew's & the Royal London School of Medicine and Denistry user on 26 March 2018
HOSPITAL MALNUTRITION
course? To these ends we estimated likelihood of malnutrition on the basis of eight nutritionrelated parameters in patients on admission to and throughout the hospital course on a general medical service, and compared nutrition status with the length of hospital stay and mortality rate. Materials and methods The nutrition status of 134 consecutive admissions to a teaching hospital general medical service was assessed during a 15-week period in July to October, 1976. The mean age of the patient population was 52 years; 50% were male, approximately 33% were Caucasian, and 66% were black. EssentiaBy all patients admitted to this service were believed to have a serious illness of either an acute or chronic nature. Common admitting problems were congestive heart failure, pulmonary infection, gastrointestinal bleeding, pancreatitis, altered mental status, urinary tract infection, and renal insufficiency. Patients were examined for serum vitamin levels, anthropometric measurements, hematological and serum albumin values within approximately 48 hr of admission. Not every parameter was obtainable or usable on all patients. Among those patients still hospitalized two weeks after admission, nutrition status waS reexamined On a weekly basis and/or several days before discharge, transfer, or death. In order to relate the findings to a single clinical service, patients transferred from the medical service were no longer followed. It was expected that significant changes in most nutrition parameters would not be detectable in less than two weeks. Each admission and follow-up patient was interviewed, examined, and weighed, and' a fasting blood sample taken for analysis of serum folate (by microbiological assay with Lactobacillus easel) (8) and Serum vitamin C (by the method of Roe and Kuether (9».
419
Serum albumin (determined by cellulose acetate electrophoresis), absolute lymphocyte count, and hematocrit are routinely obtained On admission and were reordered for the purpose of this study only if unavailable on the follow-up chart. In cases of obvious hemoconcentration the hematocrit value used was obtained after 24 to 48 hr of rehydration. Triceps skinfold (TSF) was measured using Lange calipers, taking the average of three readings at the midarm. Follow-up measurements were obtained on the same arm. Skinfold data were not used if edema of the arm was present. The mid upper arm circumference (AC) was measured with the arm straight, relaxed, and the arm muscle circumference (AMC) calculated using the standard formula (10): AMC = AC - ,,(TSF). Standing height was measured whenever possible and asked of the patient when not. Body weight measurements were not used if edema was apparent in any area of the body, and AMC data were not used if the patient had or developed arm paresis or paralysis. The guidelines used for standard and substandard values of serum folate, serum vitamin C, serum albumin, and hematocrit were taken from the monograph, Laboratory Tests for the Assessment of Nutritional Status by Sauberlich et al. (II) and are given in Table I. These guidelines are based on data compiled from published experimental studies and large population surveys. The values chosen for vitamin C correspond to the latest revision according to studies of experimental scurvy in humans refered to in this monograph. Lymphocyte count was related to standard guidelines (12). Standard values for triceps skinfold thickness and AMC were based on generally accepted survey criteria (10), and weight/ height standards were adapted from Metropolitan Life Insurance data (13), choosing as a "desirable" body weight the lower range of the medium frame category, adjusted to height without shoes. Established criteria for substandard categories ofthese anthropometric measurements are not available; hence, arbitrary levels were
TABLE I Table of standard values Substandard
Standard
Severe
Male
Moderate
Female
Vitamins Serum folate Serum vitamin C
Hospital malnutrition A prospective evaluation of general medical patients during the course of hospitalization 1-3 Roland L. Weinsier, M.D., Dr.P.H., Edie M. Hunker, R.N., Carlos L. Krumdieck, M.D., Ph.D., and C. E. Butterworth, Jr., M.D.
ABSTRACT Nutrition status was evaluated in 134 consecutive admissions to a general medical service and throughout hospitalization among patients hospitalized 2 weeks or longer. Likelihood of malnutrition was determined using eight nutrition-related parameters: serum folate and vitamin C, triceps skinfold, weight/height, arm muscle circumference, lymphocyte count, serum albumin. and hematocrit. On admission 48% of patients had a high likelihood of malnutrition, which correlated with a longer hospital stay (20 versus 12 days for patients with a low likelihood of malnutrition) and an increased mortality rate (13 versus 4%). Likelihood of malnutrition increased with hospitalization in 69% of patients with paired determinations. Compared to admission, at final follow-up a greater proportion of patients fell into the depleted range of values for folate, triceps skinfold, weight/height, arm muscle circumference, lymphocyte count, and hematocrit. These parameters worsened in over 75% of patients admitted with normal values. Hematocrit fell in all patients with normal admission levels. These findings demonstrate an association between nutrition status and hospital course and a worsening trend during hospitalization. Am. J. Clin. NUlr. 32: 418-426, 1979.
Nutrition surveys in the United States and England have documented the existence of malnutrition among hospitalized patients (1-7). One report concluded that 44% or more patients on a general medical service had protein-calorie malnutrition as evidenced by the presence of one or more substandard nutrition parameters (4). Other studies indicated that hypovitaminemia occurred commonly among hospitalized patients (I, 5). Inferences drawn from such studies have been that nutrition support is neglected (2, 3, 6), that little attempt is made to reverse malnutrition (4), and that physicians lack concern for the nutritional status oftheir patients (7). Although some reports have suggested that malnutrition actually develops during the course of hospitalization (2, 6), there have been no prospective studies to indicate 418
whether the reported prevalence of malnutrition is, in fact, hospital-related or simply a reflection of the status of patients upon hospital admission. It was the intent of this study to answer the following questions; 1) does the nutrition status of medical patients deteriorate over the course of their hospitalization? and, 1) does malnutrition adversely affect the hospital 1 From the Department ofNutrition Sciences, Schools of Public and Allied Health, Medicine, and Dentistry, University of Alabama in Birmingham, Birmingham, Alabama 35294. 2 Supported by Grant 5 R18-CA-17928-01, awarded by the National Cancer Institute, Department of Health, Education and Welfare. 3 Address reprint requests to: Roland L. Weinsier, M.D., Dr.P.H., Department of Nutrition Sciences, University of Alabama, Birmingham, Alabama 35294.
The American Journal of Clinical Nutrition 32: FEBRUARY 1979, pp. 418-426. Printed in U.S.A.
Downloaded from https://academic.oup.com/ajcn/article-abstract/32/2/418/4666289 by St Bartholomew's & the Royal London School of Medicine and Denistry user on 26 March 2018
HOSPITAL MALNUTRITION
course? To these ends we estimated likelihood of malnutrition on the basis of eight nutritionrelated parameters in patients on admission to and throughout the hospital course on a general medical service, and compared nutrition status with the length of hospital stay and mortality rate. Materials and methods The nutrition status of 134 consecutive admissions to a teaching hospital general medical service was assessed during a 15-week period in July to October, 1976. The mean age of the patient population was 52 years; 50% were male, approximately 33% were Caucasian, and 66% were black. EssentiaBy all patients admitted to this service were believed to have a serious illness of either an acute or chronic nature. Common admitting problems were congestive heart failure, pulmonary infection, gastrointestinal bleeding, pancreatitis, altered mental status, urinary tract infection, and renal insufficiency. Patients were examined for serum vitamin levels, anthropometric measurements, hematological and serum albumin values within approximately 48 hr of admission. Not every parameter was obtainable or usable on all patients. Among those patients still hospitalized two weeks after admission, nutrition status waS reexamined On a weekly basis and/or several days before discharge, transfer, or death. In order to relate the findings to a single clinical service, patients transferred from the medical service were no longer followed. It was expected that significant changes in most nutrition parameters would not be detectable in less than two weeks. Each admission and follow-up patient was interviewed, examined, and weighed, and' a fasting blood sample taken for analysis of serum folate (by microbiological assay with Lactobacillus easel) (8) and Serum vitamin C (by the method of Roe and Kuether (9».
419
Serum albumin (determined by cellulose acetate electrophoresis), absolute lymphocyte count, and hematocrit are routinely obtained On admission and were reordered for the purpose of this study only if unavailable on the follow-up chart. In cases of obvious hemoconcentration the hematocrit value used was obtained after 24 to 48 hr of rehydration. Triceps skinfold (TSF) was measured using Lange calipers, taking the average of three readings at the midarm. Follow-up measurements were obtained on the same arm. Skinfold data were not used if edema of the arm was present. The mid upper arm circumference (AC) was measured with the arm straight, relaxed, and the arm muscle circumference (AMC) calculated using the standard formula (10): AMC = AC - ,,(TSF). Standing height was measured whenever possible and asked of the patient when not. Body weight measurements were not used if edema was apparent in any area of the body, and AMC data were not used if the patient had or developed arm paresis or paralysis. The guidelines used for standard and substandard values of serum folate, serum vitamin C, serum albumin, and hematocrit were taken from the monograph, Laboratory Tests for the Assessment of Nutritional Status by Sauberlich et al. (II) and are given in Table I. These guidelines are based on data compiled from published experimental studies and large population surveys. The values chosen for vitamin C correspond to the latest revision according to studies of experimental scurvy in humans refered to in this monograph. Lymphocyte count was related to standard guidelines (12). Standard values for triceps skinfold thickness and AMC were based on generally accepted survey criteria (10), and weight/ height standards were adapted from Metropolitan Life Insurance data (13), choosing as a "desirable" body weight the lower range of the medium frame category, adjusted to height without shoes. Established criteria for substandard categories ofthese anthropometric measurements are not available; hence, arbitrary levels were
TABLE I Table of standard values Substandard
Standard
Severe
Male
Moderate
Female
Vitamins Serum folate Serum vitamin C
