ArticlePDF Available

Energy Expenditure in Mechanically Ventilated Patients: Indirect Calorimetry vs Predictive Equations

Authors:
7
https://doi.org/10.33381/dcbybd.2019.1951
ORIGINAL INVESTIGATION
ÖZGÜN ARAŞTIRMA
Yoğun Bakım Derg 2019;10(1):7−12
Energy Expenditure in
Mechanically Ventilated Patients:
Indirect Calorimetry vs Predictive
Equations
Mekanik Ventilasyon Uygulanan Hastalarda Enerji Tüketimi:
İndirek Kalorimetri Tahmin Ettirici Eşitlik Karşılaştırması
Hülya SUNGURTEKİN
1
, Serdar KARAKUZU
1
, Simay SERİN
1
ABSTRACT
Background & Objectives: Indirect calorimetry(IC) is used in the calculation of energy consumption (EE) in
critical care patients. In this study, it was aimed to compare the frequently used equations with IC in different
body weight and disease classes and to determine relationship between them and disease severity.
Materials & Methods: 100 mechanically ventilated critical care patients were prospectively included in the
study. Measurements were done on 3th, 4th and 5th days of ICU stay with IC and Harris Benedict (HB),
Penn State 2003(PS), Schofield(SCH), Swinamer (SW) and Ireton-Jones(IJ) equations were calculated and
APACHE II and SAPS II scores were determined. Bland-Altman limits of agreement analysis was done to
determine the range of error with each predictive equation compared to the measured IC.
Results: The mean age±standard deviation was 66,10 ± 14,98 years and mean body mass index was 24,91
± 4,45 kg.m-2 for the study group. Mean±standard deviation for APACHE II score and SAPS II were 23,42
± 8,47 and 42,23 ± 10,62. Measured EE was 1828, 580 ± 436, 272 kcal/day. Correlation analysis between
equations and IC showed that all equations were moderately correlated with IC. For all weight categories
and equations , the limits-of agreement range was large. For the patient group, the bias was lowest with the PS
predictive equation (mean error 14 kcal/ day). HB and PS equations have better agreement with IC than others
do. No correlation was observed between severity scores and EE.
Conclusion: Predictive formulas for EE is not reliable in determining the energy, confidence intervals are wide
in ICU patients necessitating mechanical ventilation.
Key words: indirect calorimetry, energy expenditure, predictive equations, intensive care, nutrition
ÖZ
Giriş ve Amaçlar: Yoğun bakım hastalarında enerji tüketiminin (EE) hesaplanmasında indirek kalorimetre (IC)
kullanılır. Bu çalışmada, farklı vücut ağırlıkları ve hastalık sınıflarında olan hastalarda sık kullanılan eşitliklerin
IC ile karşılaştırılması ve hastalık şiddeti ile bunların arasında ilişkinin belirlenmesi amaçlanmıştır.
Gereç ve Yöntem: Çalışmaya mekanik ventilasyon uygulanan 100 yoğun bakım hastası prospektif olarak dahil
edildi. Yoğun bakım ünitesi yatışının 3. 4. ve 5. günlerinde IC ölçümleri yapıldı ve Harris Benedict (HB), Penn
State 2003 (PS), Schofield (SCH), Swinamer (SW) ve Ireton-Jones (IJ) denklemleri hesaplandı, APACHE II ve
SAPS II skorları belirlendi. Bland-Altman limit analizi, ölçülen IC'ye kıyasla her bir tahmin denklemi ile hata
aralığını belirlemek için kullanıldı.
Bulgular: Çalışma grubunun yaş ortalaması ± standart sapması 66,10 ± 14,98 yıl ve vücut kitle indeksi ortalaması
24,91 ± 4,45 kg.m-2 idi. APACHE II skoru ve SAPS II için ortalama ± standart sapma 23,42 ± 8,47 ve 42,23 ±
10,62 idi. Ölçülen EE 1828, 580 ± 436, 272 kcal / gündü. Denklemler ve IC arasındaki korelasyon analizi, tüm
denklemlerin IC ile orta derecede ilişkili olduğunu göstermiştir. Tüm ağırlık kategorileri ve denklemleri için,
anlaşma aralığının sınırları büyüktü. Hasta grubu için bias PS tahmin denklemi ile en düşüktü (ortalama hata
14 kcal / gün). HB ve PS denklemleri, IC ile diğerlerinden daha iyi bir uyuma sahiptir. Hastalık şiddet skorları
ile EE arasında bir korelasyon gözlenmedi.
Sonuç: Enerji tüketiminin belirlenmesinde tahmin edici formüller güvenilir değildir, mekanik ventilasyon
gerektiren yoğun bakım hastalarında güven aralıkları geniştir.
Anahtar kelimeler: indirek kalorimetre, enerji tüketimi, tahmin ettirici denklemler, yoğun bakım, beslenme
7
1
Department of Anesthesiology, School
of Medicine, Pamukkale University,
Denizli, Turkey
Cite this article as: Sungurtekin
H, Karakuzu S, Serin S. Energy
Expenditure in Mechanically Ventilated
Patients: Indirect Calorimetry vs
Predictive Equations. Yoğun Bakım Derg
2019; 10(1):7-12.
Corresponding Author /
Sorumlu Yazar: Hülya Sungurtekin
E mail: hsungurtekin@yahoo.com
©Copyright 2019 by Turkish Society
of Medical and Surgical Intensive Care
Medicine - Available online at www.
dcyogunbakim.org
©Telif Hakkı 20198 Türk Dahili ve Cerrahi
Bilimler Yoğun Bakım Derneği - Makale
metnine www.dcyogunbakim.org web
sayfasından ulaşılabilir
Received/Geliş: 13.02.2019
Accepted/Kabul: 04.03.2019
Available online/
Çevrimiçi yayın: 19.03.2019
Sungurtekin H, et al. Indirect calorimetry in ICU
8
Yoğun Bakım Derg 2019;10:7−12
Introduction
Patients with critical illness requiring hospitalization in intensive
care units (ICU) are at high risk for malnutrition. Patients in ICU
have undesirable effects of both overnutrition and malnutrition.
Disease state creates a hypermetabolic state and increases EE. If
the patient’s energy need does not meet, patient’s lean body mass
losts very quickly. Lost of lean body mass in intensive care patients
has been reported to reduce the chances of survival. On the other
hand, overnutrition can be harmful and cause complications such as
hyperglycemia, azotemia and hypercapnia (1). The determination
of the energy requirements associated with a clinical assessment of
nutritional support is an important condition in these patients (2).
Total EE includes resting energy expenditure (REE), physical
activity, disease process and growing issue. Many methods and
equations have been defined to calculate EE in criticall care
patients; on the other hand, all of these methods have some
weakness. The development of a more practical tool for identifying
the EE have emerged in an effort to predictive equations
(3). Most predictive equations come from studies carried out
typically healthy, non-hospitalized patients, but only a few
studies has been approved in patients on mechanical ventilation.
There is large variability in EE regardless of which weight and
equation are used, so many predicted values will differ from the
measured values (4,5). Indirect calorimetry (IC) is a noninvasive
gold standart method that calculate EE via oxygen expenditure
(VO2) and carbon dioxide production (VCO2) (6). However, IC
devices for determining estimated energy requirement are not still
commonbecause of being expensive and being time consuming.
The aim of this study was to compare predictive equations (Harris-
Benedict (HB), IretonJones (IJ), Schofield (SCH), Swinamer (SW)
and Penn State (PS) 2003 predictions) with IC measurements in a
mechanically ventilated critically ill patient population.
Materials and Methods
This prospective study was done after written informed consent
was obtained from the patient or an authorized legal guardian.
The Local Ethics Committee of the, Medical School approved the
study. 114 mechanically ventilated patients admitted to ICU aged
over 18 were included in the study.
Patients were excluded if they were younger than 18 years old, had
hyperthemia (>380C) or hypothermia (<35°C), had intolerance
to IC procedures, were pregnant, had an FiO2 requirement >
60% or a positive end-expiratory pressure requirement greater
than 20 cm H2O, intolerant to mechanical ventilation, had a
bronchopleural fistula or chest tube leak, required continuous renal
replacement therapy or continuous ambulatory peritoneal dialysis,
or were receiving neuromuscular blockade, were intoxicated,
hemodynamically unstable. Patients were not admitted to the
study with possible edema, cardiac and/or renal failure.
The variables for each patient were obtained such as admission
height, admission weight, primary diagnosis, calculated body
mass index (BMI), age and sex by standardized chart abstraction.
Severity of disease scores (eg. Acute Physiology and Chronic
Health Evaluation II (APACHE II), Simplified Acute Physiology
Score II (SAPS II)) were calculated from patient's data within 24
hours of ICU admission. The Subjective Global Assessment (SGA)
was recorded at admission. Enteral and/or parenteral nutrition
were given during the study period according to the patient’s
status. Total parenteral nutrition (TPN) were administered via
multilumen central catheters or peripheral catheters. Enteral
nutrition were given via nasogastric tube or gastrostomy tube. We
use standard nutritional protocol in our clinic. For starting enteral
or parenteral feeding, physician should fill standardized order
sheet. Enteral or parenteral nutrition should be started at low
rates and gradually advanced to an hourly goal rate. We did not
accounted nutrition for feeding interruptions. Patients receiving
enteral nutrition are assessed for gastric residual volume every 4
hours. Gastrointestinal motility agents are received only displays
of delayed gastric emptying. Immune-enhancing formulas such
as the amino acids glutamine or lipids like omega-3 fatty acids;
micronutrients, such as antioxidant vitamins A, C and E and the
minerals selenium and zinc has not been used during the study
period.
The predictive equations: Resting energy expenditure has been
measured from actual EE measurements using IC. At the same
time with IC measurements, the predictive equations were
recorded during study period (3th, 4th and 5th days of ICU stay).
In addition, patients’ nutritional values and temperature also
recorded. The prediction accuracy of these equations was also
accepted as prediction values that were inside the range of 80%
to 110% of the measured value by IC. All other predictions that
outside this range were considered inaccurate.
Indirect Calorimetry protocol: Energy expenditure was measured
after a 30-minute rest, with no movement by the patient in
a thermoneutral environment for the 30-minute duration.
According to the protocol, in the early morning (05:30 to 07:30),
30 minutes IC measurements were taken and the average of the
measurements recorded from monitor. Patients have waited at
least 2 hours after the administration of general anesthetic agents
or hemodialysis, and have an administered fraction of inspired
oxygen of 0.6 or less. Patients should be hemodynamically
stable during measurements. During measurements, patients
should not be given drugs like vasopressors, inhaler steroid, and
bronchodilatators. Enteral or parenteral nutrition was stopped
during the IC measurements.
Patients ventilated-assisted controlled as a pressure or volume
controlled mode to be stable and comfortable in accordance with
the cause of respiratory failure via multiprocessed ventilators
(Savina or Evita XL, Dräger medical). Indirect calorimetry was
performed using Datex Ohmeda M-CAiOVX module (Datex-
Ohmeda, Finland) in accordance with IC protocol.
Body mass index (kg/m2) was calculated for all patiens using their
admission height and weight. Patients were classified into weight
categories such as BMI less than <19,9 kg/m2, 20-24.9, 25-29.9
and 30 or greater. The ideal body weight (IBW) and adjusted body
weight (AdjBW) was calculated for predictive equations(7). Data
for Long correction factor was recorded at the patient’s file.
Sungurtekin H, et al. Indirect calorimetry in ICU
9
Yoğun Bakım Derg 2019;10:7−12
Statistical Analysis
A statistical software program (SPSS 16.0 for Windows, SPSS Inc.,
Chicago, IL, USA) and MedCalc12.7.0.0 were used for statistical
analysis. Descriptive statistical methods (frequency, percentage,
mean, standard deviation) were used for data analyzing.
Correlation analysis conducted to determine the relationship
between equations and IC and between equations itself. Pearson
correlation coefficient was calculated to assess agreement inside
the method. Bland-Altman limits of agreement analysis was
undertaken to determine the extent of error with each predictive
equation compared to the measured IC. The limits of agreement
show the range of differences between the IC measurement
and the EE predicted by the equations. Data are presented as
mean±SD, with P values, and 95% confidence intervals. P values
<0.05 were accepted statistically significant.
Results
In the present clinical study, 114 patients were included.
Fourteen patients were eliminated from the study (having tube
thoracostomy, exitus during the study, need high FiO2 and having
out of range of RQ ratio (<0.7 or> 1.3)). The study was conducted
with 100 patients (58 men and 42 women). Underweight patients
accounted for 12%, whereas 45% were normal-weight, 30% were
over weight, and 13% were obese. According to SGA, 60% of the
patients were in A, 27% B and the others were in class C. Patients
were given 78% enteral nutrition where as 10% parenteral and
12% combined nutrition. Since we have a mixed intensive care
unit, we accept all types of patients. Diagnoses of acute lung injury
(n = 28, 28%), cardiac failure (n=19, 19%) and malignancies (n =
30, 30%) predominated in this population. Demographic variables
for the patient population are given in Table 1.
All of the equations were moderately correlated among themselves
and showed good agreement with each other (p <0.05). Based on
the correlation analysis SW (r=0,913, r2= 0.834) and PS (r=0,897,
r2=0.805) equations were found to be most correlated with the
IC (Table 2).
The limits-of agreement range was large for all equations and in all
weight categories. Bias is the predicted value (by equation) minus
measured value (by IC). For the patient group, the bias was lowest
with the PS predictive equation. HB and PS equations have better
agreement with IC than others. SCH, IJ and SW did not show
agreement with IC in this patient population (Table 3, Figure 1).
Best prediction among the equations was 86% with the SCH.
For the study patients, HB and PS predicted accuracy in 83%
while IJ predicted accuracy in 78%. Correlation analysis between
prediction equations and measured IC data in the patient groups
according to their BMI showed that all predicted equations were
correlated with IC (p <0.05). PS equation was found to be well
correlated with the IC in the overweight patients (r2=0,667) and
obese patients (r2=0,605). HB, SW and PS equations moderately
correlated with IC data in thin patients. According to the Blandt-
Altman analysis, PS equations have well agreement in overweight
and obese patients (Table 4). According to the correlation analysis,
Table 1. Demographic variables for the patient
Mean SD Min. Max.
Age (year) 66,1 15,0 20 89
BMI (kg.m-2)24,91 4,45 16,6 40,6
APACHE II 23,42 8,47 6 40
SAPS2 42,23 10,62 18 70
IC (kcal.day-1)1 828,58 436,27 908 3711
HB (kcal.day-1)1 716,97 404,19 1090 3393
SCH (kcal.day-1)1 692,25 340,44 1167 2758
IJ (kcal.day-1)1 577,61 262,9 1053 2510
SW (kcal.day-1)1 792,04 346,36 955 3280
PS (kcal.day-1)1 842,7 409,43 1075 3400
BMI: body mass index, APACHE II: Acute Physiology and Chronic Health
Evaluation II, SAPSII: Simplied Acute Physiology Score II, IC: indirect
calorimetry, HB: Harris Benedict, SCH: Schoeld, IJ: Ireton –Jones, SW:
Swinamer, PS: Penn-State, Min: minimum, Max: maximum, SD: standard
deviation
Table 2. Correlation analysis between equations and IC, and
between equations itself
  IC HB SCH IJ SW PS
IC r21,00 0,526 0,598 0,540 0,834 0,805
p0,00 0,00 0,00 0,00 0,00 0,00
HB r20,726 1,00 0,760 0,659 0,733 0,929
p0,00 0,00 0,00 0,00 0,00 0,00
SCH r20,598 0,760 1,00 0,605 0,612 0,773
p0,00 0,00 0,00 0,00 0,00 0,00
IJ r20,540 0,659 0,605 1,00 0,602 0,658
p0,00 0,00 0,00 0,00 0,00 0,00
SW r20,834 0,733 0,612 0,602 1,00 0,806
p0,00 0,00 0,00 0,00 0,00 0,00
PS r20,805 0,929 0,773 0,658 0,806 1,00
p0,00 0,00 0,00 0,00 0,00 0,00
IC: indirect calorimetry, HB: Harris Benedict, SCH: Schoeld, IJ: Ireton –Jones,
SW: Swinamer, PS: Penn-State,
Table 3. Summary of limits of agreement, bias and p value for
predicted energy expenditure (by equations listed), and measured
energy expenditure (by IC)
BIAS ± SD
(%95 CI)
Limits of agreement
(upper-lover) p
HB 111,61±45,82
(65,79-157,43) 341,00-564,22 0,1518
SCH 136,33±54,94
(81,39-191,27) -406,33-678,98 0,0002
IJ 250,97±59,77
(191,20- 310,74) -339,42-841,36 <0,0001
PS -14,12±38,46
(-52,58-24,34) -394,07-365,83 0,1583
SW 36,54±36,77
(-0,23-73,31) -326,71-399,79 <0,0001
HB: Harris Benedict, SCH: Schoeld, IJ: Ireton –Jones, SW: Swinamer,
PS: Penn-State,
Sungurtekin H, et al. Indirect calorimetry in ICU
10
Yoğun Bakım Derg 2019;10:7−12
SW equations were found to be suitable in groups with respiratory
failure (r2 = 0.71), malignancy (r2 = 0.69) and cardiac disease (r2
= 0.94). In the neurological and trauma group, PS equations (r2 =
0.96) were the highest correlations.
Mean±standard deviation for APACHE II score and SAPS II were
23,42 ± 8,47 and 42,23 ± 10,62. Measured EE was 1828, 580 ±
436, 272 kcal/day. The energy consumption measured by the IC
method was not correlated with disease severity scores such as the
APACHE II and SAPS II scores (p> 0.05).
Discussion
Optimal nutrition in criticallcare patients is a vital part of
intensive care unit therapy in reducing morbidity and mortality
(8). It is important to know that the actual EE to decide the most
appropriate calorie need (9). IC is accepted like the gold standard
for the assessment of the EE of patients who receive mechanical
ventilation therapy (1). For patients undergoing mechanical
ventilation therapy, 3 systems can be used in the IC measurement
(10). The Deltatrac metabolic monitor (Datex-Ohmeda, Finland)
system is the most extensive used system for IC in recent years.
The other two new systems are Quark RMR (Cosmed, Rome,
Italy) and CCM Express (Medgraphics Corp., St Paul, Minneapolis,
USA). Previous studies comparing different systems to predict
EE in patients undergoing mechanical ventilation with IC have
concluded to be the Deltatrac superiority (11). In recent years,
measurement of gas exchange technologies have been integrated
in both mechanical ventilators and patients monitors and in our
study, M-CAiOVX module was used which a new technology is
belonging to the same manufacturer of Deltatrac Monitor.
Indirect calorimetry can be performed intermittently or
continuously. The number of studies in which EE is continuously
measured for 24 hours is very limited. Most of the study showed
that measurements done in 30 minute. In clinical practice, it is
not feasible to perform 24-hour measurements with an IC in
each patient, so shorter measurements are being made. Most
measurements of studies in the literature have been accepted
to reflect the 24-hour EE for these 30-minute measurements
(12,13). We made 30-minute measurements because it was more
feasible to work with it. Our measured values were consistent
with the literature.
Subramaniam et al. (14) compared HB and SCH equations with
EE measured using the Weir equation in 60 ventilated patients
(systemic inflammatory response syndrome and sepsis). Bland–
Table 4. Correlation analysis between equations and IC for subgroup patients for BMI
IC underweight (n=12) normal (n=45) overweight (n=30) obese (13)
r2p r2p r2p r2p
HB 0,578 0,014 0,411 0,000 0,585 0,000 0,507 0,002
SCH 0,297 0,005 0,547 0,000 0,262 0,000 0,372 0,008
IJ 0,078 0,049 0,448 0,000 0,365 0,001 0,303 0,040
SW 0,548 0,001 0,584 0,000 0,612 0,000 0,549 0,004
PS 0,533 0,002 0,548 0,000 0,667 0,000 0,605 0,000
IC: indirect calorimetry, HB: Harris Benedict, SCH: Schoeld, IJ: Ireton –Jones, SW: Swinamer, PS: Penn-State,
Figure 1. Bland-Altman plot for all patients using HB equation and PS equation compared with measured energy expenditure by IC
The thick straight line in the center: Bias, dotted lines above and below Bias ± SD. Dashed lines: limits of agreement. Solid lines above and below dashed lines: 95% CI for
limits of agreement.
Sungurtekin H, et al. Indirect calorimetry in ICU
11
Yoğun Bakım Derg 2019;10:7−12
Altman analysis were done to compare the two equations. They
found that measured energy expenditure was strongly correlated
with HB among severe sepsis patients (r=0.9) and moderately
correlated with HBE among septic shock patients (r=0.43).
Correlation was found to be better in patients with severe sepsis
and an APACHE 2 score below 25. They reported that, HB and
SCH equations have sufficient validity for use in clinical practice.
Our result also agreement with their sudy. We found that the
bias was lowest with the PS predictive equation. HB and PS
equations have better agreement with IC than others. Long et al
(15) emphasized that variables such as fever or type of the injury
or illness influence the energy expenditure of surgery patients. In
another study conducted by Faisy et al. (16), predicted equations
was compared with the IC measurements. The values measured
with IC were found to exceed the values calculated with 25%
HB, and it was emphasized that the addition of long factors
significantly reduced this difference but also they reported that
values obtained by adding long factors to patients with mechanical
ventilator treatment were not reliable. The addition of long factors
has been emphasized in numerous studies that have improved
compatibility (14-16). In our study, the correlation between the
predicted equation values obtained by adding long factors and the
IC measurements was good.
Reid et al (7) studied the accuracy of equations used to predict
energy expenditure in 192 days of measurements, in 27 critically
ill patients. They compared IC and equations with Bland–Altman
analysis. The HB, SCH and American College of Chest Physicians
equations provided estimates within 80% and 110% of EE values
(66%, 66% and 65%, respectively). They suggested none of
the prediction equations were sufficiently accurate for use in
critically ill patients and would have resulted in approximately
35% of patients receiving excessive or inadequate energy
consumption. Agreement between the equations and measured
values was poor in their study. They reported Bias±SE (95%CI)
for HB, IJ, SCH and American College of Chest Physicians
equations are 111±27.7 (56–166), 141±33.2 (76–207), 85±27.9
(30–140) and 183±26.5 (131–236). In another study, EE was
prospectively measured by IC and calculated with the HB in 70
mechanically ventilated patients (16). They used Bland Altman
analysis to measure agreement between the HB and EE. The
mean bias was 73 (±502) kcal/day with limits of agreement of
932 - 1078 kcal/day. The authors reported that there is a wide
limit of agreement between the two methods and that the HB
equation was unreliable in estimating energy expenditure in
mechanically ventilated patients. We found a range of accurate
energy prediction 78%-86% except SW equation (38%). We
have different specialties of patients and older patients than Reid
at al (7) and Faisy et al. (16) studies. In our study, according to
Bland–Altman analysis Bias±SD (95%CI) for HB, IJ and SCH
111,61±45,82 (65,79-157,43), 250,97±59,77 (191,2-310,74)
and 136,33±54,94 (81,39-191,27). Since the studies conducted
in the literature are usually performed in a few critical cases, the
disease diagnoses cannot be analyzed by subdividing the patients
into subgroups such as old, young, obese, and weak. Frankenfield
et al. (17) reported a comparative study of 202 intensive care
patients was one of the most extensive investigations we have
found. According to their study, the PS equation is the definite
equation across all subgroups. It was reliable in trauma patients
(r = 0.77), in surgical patients (r = 0.66), and in medical patients
(r = 0.62). The correlation rate (r = 0.67) was the same in febrile
and non-febrile patients. In a retrospective analysis, De Waele et
al. (18) examined the agreement between EE measured by IC
and predicted by equations in mechanically ventilated critically
ill patients. They used eleven different prediction equations.
This study had the largest and most different sample among
the studies; there was also great variability in measured EE
and prediction equations they used. They concluded that ten
widely used equations for calculating resting EE failed to achieve
acceptable accordance with resting energy expenditure measured
by IC. In a comprehensive study conducted by Maday (1) in
2013, the agreement of IJ, PS 2003, SW, Brandi, Faisy, HB and
MifflinSt.Jeor equations to IC measurements in obese and non-
obese patients was examined. Patients were grouped as young
obese, young non-obese, old obese and old nonobese. According
to this study, in the whole population, the high accuracy of Penn
State 2003 was found in young obese and elderly non-obese
patients. In our study PS 2003 equation was used. SW equations
were found to be suitable in groups with respiratory failure (r2 =
0.71), malignancy (r2 = 0.69) and cardiac disease (r2 = 0.94). In
the neurological and trauma group, PS equations (r2 = 0.96) were
the highest correlations. We may suggest SW or PS equations for
accurate assessment of metabolic rate in critically ill patients if
IC is not available. Best prediction among the equations was 86%
with the SCH whereas HB and PS predicted accuracy in 83%
while IJ predicted accuracy in 78% for the study patients with
BlandAltman limits of agreement analysis.
IC measurements of patients receiving mechanical ventilator
therapy were not affected by mechanical ventilator modes
(19). In our study, we performed measurements using the most
comfortable mechanical ventilation modes of the patients who
were fit. Flancbaum et al. (3) reported that the difference in
baseline EE might be due to disease severity. However, Brandi et
al. (20) did not found any correlation between disease severity
score and the basal EE measured by IC. In our study, in agreement
with the Brandi et al study (20) no correlation was found between
the APACHE II, SAPS II scores and the IC masurement.
There are some limitations of our study. The study was
conducted on a heterogeneous group of patients and that the
patient group consisted predominantly of elderly patients. The
study is rather small to profound detailed subgroup analysis. An
additional limitation is that several patient populations were not
shown in the current study, spinal cord injury with quadriplegia
or paraplegia, subgroups of children, burn injury, and penetrating
trauma. The clinical benefit of these results should also be
demonstrated.
In conclusion, in the intensive care patients who need mechanical
ventilation, predictive equations are not reliable in determining
EE; the confidence intervals are very high and can lead to
inadequate feeding or overfeeding. In critically ill mechanically
ventilated patients, there is a need for larger scale studies of
which formula should be used in which group of patients in the
absence of IC.
Sungurtekin H, et al. Indirect calorimetry in ICU
12
Yoğun Bakım Derg 2019;10:7−12
References
1. Maday KR. Energy Estimation in the Critically Ill: A Literature
Review. Universal Journal of Clinical Medicine 2013;1(3):39-43.
[CrossRef]
2. Barak N, Wall-Alonso E, Sitrin MD. Evaluation of stress factors
and body weight adjustments currently used to estimate EE in
hospitalized patients. JPEN 2002 Jul-Aug;26(4):231-8. [CrossRef]
3. Flancbaum L, Choban PS, Sambucco S, et al. Comparison of indirect
calorimetry, the Fick method, and prediction equations in estimating
the energy requirements of critically ill patients. Am J Clin Nutr
1999; 69: 461-6. [CrossRef]
4. Dickerson RN.Specialized nutrition support in the hospitalized
obese patient. Nutr Clin Pract 2004; 19(3):245-54. [CrossRef]
5. Krenitsky J. Adjusted body weight, pro: evidence to support use of
adjusted body weight in calculating calorie requirements. Nutr Clin
Pract 2005;20(4):468-73. [CrossRef]
6. Miles-Chan JL, Dulloo AG, Schutz Y. Fasting substrate oxidation at
rest assessed by indirect calorimetry: is prior dietary macronutrient
level and composition a confounder? Int J Obes (Lond) 2015
Jul;39(7):1114-7. [CrossRef]
7. Reid CL. Poor agreement between continuous measurements of
energy expenditure and routinely used prediction equations in
intensive care unit patients. Clin Nutr 2007; 26:649-57. [CrossRef]
8. Strack van Schijndel RJ, Weijs PJ, Koopmans RH, Sauerwein
HP, Beishuizen A, et al. Optimal nutrition during the period of
mechanical ventilation decreases mortality in critically ill, long-term
acute female patients: a prospective observational cohort study. Crit
Care 2009;13(4):R132. [CrossRef]
9. Singer P, Anbar R, Cohen J, et al. The tight calorie control study
(TICACOS): a prospective, randomized, controlled pilot study
of nutritional support in critically ill patients. Intensive Care Med
2011;37:601–9. [CrossRef]
10. Sundström M, Tjader I, Rooyackers O, et al. Indirect calorimetry in
mechanically ventilated patients. A systematic comparison of three
instruments. Clin Nutr 2013; 32:118-21. [CrossRef]
11. Graf S, Karsegard VL, Viatte V, et al. Evaluation of three indirect
calorimetry devices in mechanically ventilated patients: which device
compares best with the Deltatrac II(®)? A prospective observational
study. Clin Nutr 2015 Feb;34(1):60-5. [CrossRef]
12. Wooley JA. Indirect calorimetry: applications in practice. Respir
Care Clin N Am 2006;12:619-33.
13. Swinamer DL, Phang PT, Jones RL, et al. Twenty-four hour energy
expenditure in critically ill patients. Crit Care Med 1987; 15: 637-
43. [CrossRef]
14. Subramaniam A, McPhee M, Nagappan R. Predicting energy
expenditure in sepsis: Harris-Benedict and Schoeld equations
versus the Weir derivation. Crit Care Resusc 2012 Sep;14(3):202-
10.
15. Long CL, Schaffel N, Geiger JW, et al. Metabolic response to injury
and illness: estimation of energy and protein needs from indirect
calorimetry and nitrogen balance. JPEN J Parenter Enteral Nutr
1979;3:452–6. [CrossRef]
16. Faisy C, Guerot E, Diehl JL, et al. Assessment of resting energy
expenditure in mechanically ventilated patients. Am J Clin Nutr
2003;78(2):241-9. [CrossRef]
17. Frankeneld DC, Coleman A, Alam S, et al. Analysis of estimation
methods for resting metabolic rate in critically ill adults. JPEN
2009;33(1): p. 27-36. [CrossRef]
18. De Waele E, Opsomer T, Honoré PM, et al. Measured versus
calculated resting energy expenditure in critically ill adult patients.
Do mathematics match the gold standard? Minerva Anestesiol. 2015
Mar;81(3):272-82.
19. Clapis FCM, Auxiliadora-Martins M, et al. Mechanical ventilation
mode (volume × pressure) does not change the variables obtained by
indirect calorimetry in critically ill patients. JCrit Care 2010;25, 659.
e9-16. [CrossRef]
20. Brandi LS, Santini L, Bertolini R, et al. Energy expenditure and
severity of injury and illness indices in multipl trauma patients. Crit
Care Med 1999; 27: 2684-89. [CrossRef]
21. Ireton-Jones C. Adjusted Body Weight, Con: Why Adjusted Body
Weight in Energy Expenditure Calculations? NutrClin Pract
2005;20:474-79. [CrossRef]
YAZAR KATKILARI:
Fikir: HS, SK; Tasarım: HS, SK; Denetleme: HS, SK, SS; Kaynaklar: HS, SK, SS;
Malzemeler: HS, SK, SS; Veri Toplanması ve/veya İşlemesi: SK; Analiz ve/veya
Yorum: HS, SK; Literatür Taraması: SK; Yazıyı Yazan: HS
Ethics Committee Approval: Ethics committee approval was received for this
study from the ethics committee of Pamukkale University, Faculty of Medicine
(Approval Date: 2013 / Session No: x/x, Decision No: x).
Informed Consent: Written informed consent was obtained from relatives of
patients or patients who participated in this study.
Peer-review: Externally peer-reviewed.
Conict of Interest: Authors have no conicts of interest to declare.
Financial Disclosure: The authors declared that this study has received no
nancial support.
*16th World Congress of Anaesthesiologists, 28 August-2 September 2016, Hong Kong
Etik Komite Onayı: Bu çalışma için etik kurul onayı Pamukkale Üniversitesi Tıp
Fakültesi etik kurulundan alınmıştır (Onay Tarihi: Nisan, 2013 / Oturum No:
xx/x, Karar No: x).
Hasta Onamı: Yazılı hasta onamı bu çalışmaya katılan hasta veya hastaların
yakınlarından alınmıştır.
Hakem Değerlendirmesi: Dış bağımsız.
Çıkar Çatışması: Yazarlar çıkar çatışması bildirmemişlerdir.
Finansal Destek: Yazarlar bu çalışma için nansal destek almadıklarını beyan
etmişlerdir.
*16. Dünya Anestezi Uzmanları Kongresi, 28 Ağustos - 2 Eylül 2016 - Hong Kong
AUTHOR CONTRIBUTIONS:
Concept: HS, SK; Design: HS, SK; Supervision: HS, SK, SS; BB; Resources: HS,
SK, SS; Materials: HS, SK, SS; Data Collection and/or Processing: SK; Analysis
and/or Interpretation: HS, SK; Literature Search: SK; Writing Manuscript: HS
ResearchGate has not been able to resolve any citations for this publication.
Article
Full-text available
Background: Indirect calorimetry (IC) is considered to be the standard method for estimating energy requirements in intensive care unit (ICU) patients. Hence, most ICU clinicians still rely on various mathematical formulas to calculate caloric requirements in their patients. We assessed whether measurements obtained by IC reached agreement with the results of such commonly used equations. Methods: Retrospective study in consecutively hospitalized patients in a mixed medico-surgical adult ICU. Resting energy expenditure (REE) was measured by IC in all patients as a standard procedure within our routine nutritional care planning and simultaneously calculated from 10 distinct predictive equations. IC was performed with the VmaxTM Encore 29n calorimeter (VIASYS Healthcare Inc, Yorba Linda, CA). Bland-Altman plots and regression analysis were used to assess agreement between measured and calculated REE. Results: The study included 259 critically ill patients: 161 subjects (62%) met final analysis criteria (age 63 ± 16 years; 58% males). Measured REE was 1571 ± 423.5 kcal/24 h with VO2 0.23 ± 0.06 L/min and VCO2 0.18 ± 0.05 L/min. Calculated values correlated very weakly with IC-derived measurements. Only the Swinamer equation and the Penn State 2010 reached an R² > 0.5. Widely used formulas in daily ICU practice such as the adjusted Harris Benedict, Faisy-Fagon, and ESICM '98 statement equations, reached R² values of respectively only 0.44, 0.49, and 0.41. Calculation resulted in under- as well as overestimation of REE. Global formulas reached no acceptable correlation in elderly or obese critically ill patients. Conclusion: In critically ill adult patients, measured REE poorly correlated with calculated values, regardless what formula was used. Our findings underscore the important role of IC to adequately estimate energy requirements in this particularly frail population.
Article
Full-text available
Critically ill patients have higher energy expenditure and increased nutritional needs compared to non-critically patients. This can lead to medical malnutrition if nutritional interventions are not implemented quickly and accurately in the disease course. Obtaining an accurate estimation of nutritional needs is paramount in establishing nutritional support. The literature review was designed to look at the numerous predictive equations and determine the accuracy among different critically ill patient populations. Data Sources: Using keywords energy estimation, nutritional predictive equations, and critical illness, systematic reviews, meta-analyses, and validation studies were identified through electronic database searches and citation tracking. Study Selection: Articles were selected that evaluated individual predictive equations in critically ill patients, their accuracy compared to indirect calorimetry, and their use in special patient populations. Data Extraction and Synthesis: A total of 12 articles were selected using the inclusion and exclusion criteria. Accuracy of the predictive equations was separated into overall accuracy among all patients, all ages, non-obese, young obese and non-obese, and elderly obese and non-obese. Conclusions: There are several validated predictive equations for estimating energy expenditure in critical illness. The most accurate equations are the 2003 Penn State equation, for obese and non-obese adult patients, and the 2010 Penn State equation, for elderly obese patients. Given the trend of obesity in the United States, further validated studies are needed to look at the individual classes of obesity and the accuracy with this specific patient population.
Article
Full-text available
Given the difficulties of using indirect calorimetry in many intensive care units, clinicians routinely employ predictive equations (the Harris-Benedict equation [HBE] and Schofield equation are commonly used) to estimate energy expenditure in critically ill patients. Some extrapolate CO(2) production (V CO(2)) and O(2) consumption (V O(2)) by the Weir derivation to estimate energy expenditure. These derivative methods have not been compared with predictive equations. To compare prediction of energy expenditure by the HBE and Schofield equation with energy expenditure as estimated by the Weir derivation in a cohort of critically ill patients. Between June 2009 and May 2010, we conducted a prospective single-centre study of 60 mechanically ventilated patients with sepsis of varying severity in the ICU of a metropolitan hospital. Three groups of patients were compared: those with systemic inflammatory response syndrome (SIRS), severe sepsis and septic shock. The HBE and Schofield equation are age-based, weight-determined, sex-specific derivations that may incorporate stress and/or activity factors. Total energy expenditure (TEE) values calculated from these equations (TEE(HBE) and TEE(SCH), respectively) were compared with the measured energy expenditure (MEE) calculated by the Weir derivation. We derived V CO(2) from end-tidal CO(2) and deduced V O2 assuming a respiratory quotient of 0.8381. Mean (± SD) APACHE II score for the 60 patients was 25.7 ± 8.4. All patients received nutrition (51 enteral, eight parenteral and one combined) in addition to standard management for sepsis and multiorgan supportive therapy. Overall, 45 patients required inotropes and four received continuous renal replacement therapy. TEE derived from both predictive equations correlated well with MEE derived from the Weir equation (mean TEE(HBE), 7810.7 ± 1669.2 kJ/day; mean TEE(SCH), 8029.1 ± 1418.6 kJ/day; mean MEE, 7660.8 ± 2092.2 kJ/day), being within 8% of each other. Better correlations between TEE and MEE were observed in patients with APACHE II scores < 25 (vs those with scores ≥25) and patients with SIRS or severe sepsis (vs those with septic shock). In a cohort of patients with sepsis, TEE values calculated by the HBE and Schofield equation matched reasonably well with MEE values derived from the Weir equation. Correlation was better in patients with less severe sepsis (SIRS and severe sepsis and APACHE II score < 25). Our results suggest that predictive equations have sufficient validity for ongoing regular use in clinical practice.
Article
Full-text available
To determine whether nutritional support guided by repeated measurements of resting energy requirements improves the outcome of critically ill patients. This was a prospective, randomized, single-center, pilot clinical trial conducted in an adult general intensive care (ICU) unit. The study population comprised mechanically ventilated patients (n = 130) expected to stay in ICU more than 3 days. Patients were randomized to receive enteral nutrition (EN) with an energy target determined either (1) by repeated indirect calorimetry measurements (study group, n = 56), or (2) according to 25 kcal/kg/day (control group, n = 56). EN was supplemented with parenteral nutrition when required. The primary outcome was hospital mortality. Measured pre-study resting energy expenditure (REE) was similar in both groups (1,976 ± 468 vs. 1,838 ± 468 kcal, p = 0.6). Patients in the study group had a higher mean energy (2,086 ± 460 vs. 1,480 ± 356 kcal/day, p = 0.01) and protein intake (76 ± 16 vs. 53 ± 16 g/day, p = 0.01). There was a trend towards an improved hospital mortality in the intention to treat group (21/65 patients, 32.3% vs. 31/65 patients, 47.7%, p = 0.058) whereas length of ventilation (16.1 ± 14.7 vs. 10.5 ± 8.3 days, p = 0.03) and ICU stay (17.2 ± 14.6 vs. 11.7 ± 8.4, p = 0.04) were increased. In this single-center pilot study a bundle comprising actively supervised nutritional intervention and providing near target energy requirements based on repeated energy measurements was achievable in a general ICU and may be associated with lower hospital mortality.
Article
Full-text available
Optimal nutrition for intensive care patients has been proposed to be the provision of energy as determined by indirect calorimetry, and protein provision of at least 1.2 g/kg pre-admission weight per day. The evidence supporting these nutritional goals is based on surrogate outcomes and is not yet substantiated by patient oriented, clinically meaningful endpoints. In the present study we evaluated the effects of achieving optimal nutrition in ICU patients during their period of mechanical ventilation on mortality. This was a prospective observational cohort study in a mixed medical-surgical, 28-bed ICU in an academic hospital. 243 sequential mixed medical-surgical patients were enrolled on day 3-5 after admission if they had an expected stay of at least another 5-7 days. They underwent indirect calorimetry as part of routine care. Nutrition was guided by the result of indirect calorimetry and we aimed to provide at least 1.2 g of protein/kg/day. Cumulative balances were calculated for the period of mechanical ventilation. Outcome parameters were ICU, 28-day and hospital mortality. In women, when corrected for weight, height, Apache II score, diagnosis category, and hyperglycaemic index, patients who reached their nutritional goals compared to those who did not, showed a hazard ratio (HR) of 0.199 for ICU mortality (CI 0.048-0.831; P = 0.027), a HR of 0.079 for 28 day mortality (CI 0.013-0.467; P = 0.005) and a HR of 0.328 for hospital mortality (CI 0.113-0.952; P = 0.04). Achievement of energy goals whilst not reaching protein goals, did not affect ICU mortality; the HR for 28 day mortality was 0.120 (CI 0.027-0.528; P = 0.005) and 0.318 for hospital mortality (CI 0.107-0.945; P = 0.039). No difference in outcome related to optimal feeding was found for men. Optimal nutritional therapy improves ICU, 28-day and hospital survival in female ICU patients. Female patients reaching both energy and protein goals have better outcomes than those reaching only the energy goal. In the present study men did not benefit from optimal nutrition.
Article
Indirect calorimetry, the measurement of O2 consumption and CO2 production, constitutes an invaluable tool as the most common method for analyzing whole body energy expenditure, and also provides an index of the nature of macronutrient substrate oxidation, namely carbohydrate (CHO) versus fat oxidation. The latter constitutes a key etiological factor in obesity since this condition can only develop when total fat oxidation is chronically lower than total exogenous fat intake. The standardization of indirect calorimetry measurements is essential to accurately track the relative proportion of energy expenditure derived from CHO and fat oxidation. Here we analyse literature data to show that the average fasting RQ typically shifts from approximately 0.80-0.90 (indicating a doubling of resting CHO oxidation) in response to a switch in dietary CHO intake (as % energy) from 30-60%. This underscores the importance of taking into account dietary macronutrient composition prior to indirect calorimetry studies in the interpretation of data on substrate utilization and oxidation.International Journal of Obesity accepted article preview online, 16 March 2015. doi:10.1038/ijo.2015.29.
Article
Indirect calorimetry (IC) is the gold standard to measure energy expenditure (EE) in hospitalized patients. The popular 30 year-old Deltatrac II(®) (Datex) IC is no more commercialized, but other manufacturers have developed new devices. This study aims at comparing for the first time simultaneously, two new IC, the CCM express(®) (Medgraphics) and the Quark RMR(®) (Cosmed) with the Deltatrac II(®) to assess their potential use in intensive care unit (ICU) patients. ICU patients on mechanical ventilation, with positive end-expiratory pressure <9 cm H2O and fraction of inspired oxygen <60%, underwent measurements by the three IC simultaneously connected during 20 min to the ventilator (Evita XL(®), Dräger). Patients' characteristics, VO2 consumption, VCO2 production, respiratory quotient and EE were recorded. Data were presented as mean (SD) and compared by linear regression, repeated measure one-way ANOVA and Bland & Altman diagrams. Forty patients (23 males, 60(17) yrs, BMI 25.4(7.0) kg/m(2)) were included. For the Deltatrac II(®), VO2 was 227(61) ml/min, VCO2 189(52) ml/min and EE 1562(412) kcal/d. VO2, VCO2, and EE differed significantly between Deltatrac II(®) and CCM express(®) (p < 0.001), but not between Deltatrac II(®) and Quark RMR(®). For EE, diagrams showed a mean difference (2SD) of 25.2(441) kcal between Deltatrac II(®) vs. the Quark RMR(®), and -273 (532) kcal between Deltatrac II(®) vs CCM express(®). Quark RMR(®) compares better with Deltatrac II(®) than CCM express(®), but it suffers an EE variance of 441 kcal, which is not acceptable for clinical practice. New indirect IC should be further improved before recommending their clinical use in ICU.
Article
The aim of the study was to analyze the difference between the results obtained by indirect calorimetry (IC) using volume-controlled and pressure-controlled mechanical ventilation in 2 different ventilators and to characterize the variables achieved by IC after well-defined changes in minute volume (Vm). Prospective study of 20 critically ill patients under volume-controlled (n = 15) or pressure-controlled (n = 5) mechanical ventilation. Three IC measurements of 45 minutes each were taken; values of oxygen consumption (Vo(2)), carbon dioxide production (Vco(2)), Vm, resting energy expenditure (REE), and respiratory quotient (RQ) were obtained. For the last measurement, Vm was set at 20% above the baseline. There were no differences between the results obtained by IC during volume-controlled and pressure-controlled mechanical ventilation. The most relevant changes in the variables obtained by IC before and after intervention in Vm were a significant increase in Vco(2) (from 165 to 177 mL·min(-1); P < .01), a decrease in Paco(2) (from 38.49 to 28.46 mm Hg; P < .01), and a rise in pH (from 7.41 to 7.49; P < .01). There were no alterations in Vo(2), REE, or RQ. Ventilators and ventilation modes do not influence the IC measurements. The observed changes have no clinical effects and are reversible, provided that increased Vm is maintained for no longer than 45 minutes.
Article
Prediction of metabolic rate is an important part of the nutrition assessment of critically ill patients, yet there are limited data regarding the best equation to use to make this prediction. Standardized indirect calorimetry measurements were made in 202 ventilated, adult critical care patients, and resting metabolic rate was calculated using the following equations: Penn State equation, Faisy, Brandi, Swinamer, Ireton-Jones, Mifflin, Mifflinx1.25, Harris Benedict, Harris Benedictx1.25, Harris Benedict using adjusted weight for obesity, and each of the adjusted weight versions of Harris Benedictx1.25. The subjects were subgrouped by age and obesity status (young nonobese, young obese, elderly nonobese, elderly obese). Performance of each equation was assessed using bias, precision, and accuracy rate statistics. Accuracy rates in the study population ranged from 67% for the Penn State equation to 18% for the weight-adjusted Harris Benedict equation (without multiplication). Within subgroups, the highest accuracy rate was 77% in the elderly nonobese using the Penn State equation and the lowest was 0% for the weight-adjusted Harris Benedict equation. The Penn State equation was the only equation that was unbiased and precise across all subgroups. The obese elderly group was the most difficult to predict. Therefore, a separate regression was computed for this group: Mifflin(0.71)+Tmax(85)+Ve(64)-3085. The Penn State equation provides the most accurate assessment of metabolic rate in critically ill patients if indirect calorimetry is unavailable. An alternate form of this equation for elderly obese patients is presented, but has yet to be validated.