| Sign In to gain access to subscriptions and/or personal tools. |
Subjective Global Assessment in the Clinical SettingFrom the Servicio de Clínica Medica, Complejo Medico-Policial "Churruca-Visca," Buenos Aires, Argentina Correspondence: Fernando G. Baccaro, MD, Servicio De Clínica Medica, Complejo Medico-Policial "Churruca-Visca," Arenales 2896 Pb (1425), Buenos Aires, Argentina. Electronic mail may be sent to fbaccaro{at}2vias.com.ar.
Background: Our goal was nutrition assessment in hospitalized
patients of an internal medicine service. Methods: Ours was a
longitudinal, prospective, and observational study. Four hundred twelve
patients participated in this study using the Subjective Global Assessment
(SGA). We used In spite of remarkable evolution achieved in medicine, malnutrition persists as a widespread problem in hospitalized patients. Causes are multiple, among which admittance pathology, age, and sociocultural level, among others, are to be highlighted. Malnutrition in hospitalized patients is associated with high morbidity and mortality rates, increased length of stays, and costs.1 For several years, different strategies and methods have been developed to evaluate hospitalized patients' nutrition. Yet, malnutrition is still difficult to diagnose and often overlooked by the medical team. According to different studies,2,3 prevalence is between 30% and 50%. In one of these studies, 75% of malnourished patients deteriorated in their nutrition state during hospitalization. These studies were carried out in developed and developing countries. In Latin America, assessments on nutrition state in hospitalized patients have been carried out in Brazil, Chile, and Argentina. The Brazilian study IBRANUTRI, performed with 4000 patients, identified malnutrition in 56.6% of the patients.4 The Chilean study, performed with 528 patients, showed malnutrition in 37% of the patients.5 Finally, in Argentina and with 1000 patients, malnutrition was detected in 47% of the assessed population.6 The main objective of this paper is to assess the nutrition state of hospitalized patients in an internal medical service in Argentina and the influences that different variables might have on it.
Patients were assessed in admission to the "Servicio de Clínica Médica del Complejo Médico–Policial Churruca–Visca," in Buenos Aires, Argentina, a general hospital with 500 beds, and 70 beds in the Internal Medicine Service. The assessment was carried out from December 3, 2002 to June 1, 2003. According to Service admission conditions, the minimum age of admittance was 16 years old. The chosen assessment method was the Subjective Global Assessment (SGA) described by Detsky et al.7 The SGA screens for rating in 2 broad areas: medical history and physical examination. The first SGA component, the medical history, involves asking questions and evaluating the patient's answers about the following 4 parameters:
There are several body locations to examine for each parameter. The clinician rates each medical history and physical examination parameter as either an A, B, or C on the SGA scoring sheet. According to all of these parameters' ratings, the clinical observer assigns an overall SGA classification that corresponds to the subjective opinion of the patient's nutrition status. It results in a 3-level classification: normal nourished, moderate malnourished (or at risk of malnutrition), and severe malnourished (or poor nutrition status (Figure 1).
The assessment was carried out by one of the authors (F.B.) and 4 resident
physicians. The assessment took place in the first 24 hours after the
patients' admission, and it was not adjoined to their medical history. The
average interview and assessment took 9 minutes (range, 6–14 minutes).
Pathologies presented were classified into groups according to affected
systems or medically prevalent conditions upon admission, without taking into
consideration previous pathologies or comorbidities. A trial to determine the
reliability between 5 observers was conducted to internally validate the SGA.
The sample included 75 patients before the study started. Correlation between
interviews was good, with a coefficient
Interviewers neither followed patients during their hospitalization nor did
they participate in clinical history assessment. As weight, anthropometric
measures, total lymphocyte count, and albumin were not available in all
records, those variables were not included in the study to evaluate or define
patient's nutrition status or to compare it with the SGA. Admission days were
divided between <12 days and >12 days, and age <60 years old
(non-old) and >60 years old (old), dividing population in old and non-old
people according to the World Health
Organization.8,9
Statistic analysis was done with SPSS 10 package (SPSS Inc, Chicago, IL). This
included distributional frequencies of all variables. For the comparison of
ordinal and nominal variables
The evaluation consisted of 412 patients: 236 men and 176 women. The average age was 65.7 years old, ranging between 16 and 94 years old. The distribution of pathologies per group is shown in Table I. The results obtained with the application of SGA were 52.4% of the patients (n = 216) were normonourished (group A), whereas 47.6% (n = 196) were malnourished (Figure 2). From them, 160 patients (38.8%) presented with moderate malnutrition (group B) and 36 patients (8.73%) presented with severe malnutrition (group C). The average length of hospitalization was 12.2 days (± 2.4 days), ranging between 2 and 133 days, with a value of 8.7 days for the normonourished group and 15.7 for the malnourished (p < .05). The univariate analysis showed a significant correlation between malnutrition and sex, age older than 60 years, and infectious and oncologic pathology diagnosis (p < .001).
Men older than 60 years presented higher predisposition to malnutrition (odds ratio [OR] = 2.23; confidence interval [CI] = 1.57–2.76; p < .001). Patients with oncologic and infectious pathology also presented a higher risk of malnutrition (OR = 2.31, CI = 1.91–2.80 for oncologic patients, p < .001; and OR = 1.91, CI = 0.97–1.46 for infectious, p < .05). Sex, age older than 60 years, and infectious and oncologic pathologies raise the probability of malnutrition, although oncologic pathology and age older than 60 have major probabilities (OR = 2.37, CI = 1.63–2.87 years; OR = 2.06, CI = 1.46–2.51, respectively). Malnutrition correlates significantly with a longer length of hospitalization (>12 days; p < .01; average of 15.7 days ± 4.3). Population characteristics according to these variables are shown in Table II. From the studied population, 8 patients died (1.94%): 1 patient corresponding to group A (0.24%), 2 to group B (0.05%), and 5 to group C (1.2%). In the 412 clinical histories of assessed patients, weight and height were not included; only 23% of medical histories included these parameters. The same occurred with the nutrition state of the patients because only in 18% of the clinical histories was this mentioned. It must be remembered that physicians in charge of clinical histories and patients' monitoring did not have access to the nutrition assessment effected.
Many factors contribute to malnutrition, such as socioeconomic status, age, pathology of illness, and hospitalization history.10,11 The results from this paper were similar to others carried out in developed countries, and ones performed in Latin America and Argentina respectively. The number of our patients from group C is similar to the Argentinean study. It is remarkable that fewer than 20% of the clinical histories refer to the nutrition state of patients. Although the nutrition contribution received by the patients was not assessed, a systematic register where the nutrition aspect is part of the daily care does not exist. Either way, it is well known that malnutrition is associated with longer hospitalization. The use of SGA allows one to distinguish between well-nourished patients and malnourished ones without the need to use anthropometric or laboratory methods (total lymphocyte count and albumin) and with adequate certainty of fidelity of results.12–14 Anthropometric measurements, such as measurement of triceps skinfold, arm circumference measurement, etc (as Jelliffe and Frisancho charts) were developed from a male population in service with the U.S. Army in Greece, poor American women, and men and women participating from the U.S. Health and Nutrition Survey between 1971 and 1974.15,16 According to some authors, up to 30% of the healthy controls should be considered malnourished if these charts are used.17,18 However, many recent studies (Malnutrition Screening Tool, Nutrition Screening Tools, Nutritional Risk Index, etc) combine SGA, considered in general as the gold standard, with certain anthropometric parameters (body mass index [BMI], measurement of triceps skinfold, etc). In any case, there are no validation studies between the studies mentioned before and SGA.19–24 SGA has been used by the ELAN (Estudio Lationamericano sobre Nutrición) survey in Latin America, the IBRANUTRI (Investigación Brasileña sobre Nutrición) in Brazil, and the malnutrition study in Argentina, to determine the prevalence of clinical malnutrition. Although anthropometric measurements were registered, lymphocyte blood cell count and albumin levels were recorded in the former but not in the present study. According to this research, malnutrition in hospitalized patients in an internal medicine service in Argentina presents a high incidence, similar to the ones described in previous works. The relationship between malnutrition and socioeconomic factors has been mentioned before, particularly in Latin America, relating to health governmental policies resources, which is beyond the reach of the present study. According to some authors, a "common illness" is one with >10% prevalence. In the beginning of the third millennium, malnutrition constitutes one of the most common illnesses in the hospital field. As this determines outcome conditions, malnutrition should concern the whole medical team.25 The lack of simple economic actions to determine the nutrition state of the hospitalized population contrasts with the high costs derived as a consequence.26–28 Received for publication April 27, 2006. Accepted for publication March 28, 2007.
Journal of Parenteral and Enteral Nutrition, Vol. 31, No. 5,
406-409 (2007)
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
2 for univariate and logistic regression.
Results: Of 412 patients, 47.6% presented with malnutrition: 38.8%
with moderate malnutrition (group B), and 8.58% with severe malnutrition
(group C). Malnutrition was related to male patients older than 65 years,
oncologic and infectious diseases, and length of hospitalization.
Conclusions: Malnutrition incidence in an internal medical service is
high. There is remarkable lack of interest in hospitalized patients' nutrition
state. Results show similarities to other studies from Latin America. 
Top
MATERIALS AND METHODS
RESULTS
= 0.75. In case information
could not be retrieved from the patient, direct relatives were asked. 