The effect of cisatracurium infusion on the energy expenditure of critically ill patients: an observational cohort study

Cisatracurium is the NMBA of choice for sustained neuromuscular blockade during critical illness in Gelderse Vallei Hospital. Cisatracurium was administered when indicated according to the international clinical practice guidelines for the sustained neuromuscular blockade in the adult critically ill patient [13]. An infusion was started at doses of 3 μg/kg per minute and then adjusted by assessment of the train-of-four (TOF) using a peripheral nerve stimulator (TOF-watch® S, Dublin, Ireland). According to the hospital protocol, TOF measurements were performed every hour, and dosage adjustments were made to achieve a TOF level of 1 or lower. The electrodes of the TOF-watch® were placed on the other wrist daily to prevent skin lesions.

The primary endpoint was the total EE, expressed as kcal/day, which was measured before and during cisatracurium infusion. Indirect calorimetry was not routinely available during the study period. EE was, therefore, estimated by an adjusted version of Weir’s equation using the ventilator-derived VCO₂ (EEVCO₂). EEVCO₂ = 3.941 × VCO₂(L/min) / respiratory quotient + 1.11 × VCO₂(L/min) × 1440. The respiratory quotient was considered to be a fixed value of 0.86 [14‐16]. The mechanical ventilator measured the VCO2 (Hamilton-S1, Hamilton Medical AG, Bonaduz, Switzerland), and every minute, data are automatically sent to our electronic patient data management system (MetaVision; iMDsoft MetaVision®, Tel Aviv, Israel). For each patient, the VCO₂ was collected during the 12 h before and during the 12 h after the start of cisatracurium infusion. When patients were not admitted to the ICU 12 h before the start of cisatracurium infusion, the parameters of the available hours were used. The EEVCO₂ was calculated every 2 h using the mean VCO₂ measurements from the previous 2 h.

Secondary endpoint was hypercaloric feeding (> 110% of EE) after cisatracurium infusion. We also evaluated ICU length of stay (LOS) and in-hospital mortality in patients receiving hypercaloric versus regular or hypocaloric feeding.

The World Health Organization/Food and Agricultural Organization of the United Nations (WHO/FAO) formulas were used to calculate caloric and protein targets by our computerized feeding protocol [17]. According to BMI, the actual (BMI < 27), corrected (BMI 27–30; regression to BMI of 27), or ideal body weight (BMI > 30; regression to BMI 21 in women and BMI 22.5 in men) was used. An addition to the resting EE (REE) of 20% was used to correct for disease activity [18].

Most parameters were routinely collected into an extensive ICU database during standard clinical care. Data extraction was performed using SAS Enterprise Guide queries (version 7.12HF1) searching our Patient Data Management System (MetaVision; iMDsoft MetaVision®, Tel Aviv, Israel, and neoZIS®, Electronic Medical Record, MI Consultancy, Katwijk, The Netherlands). Data to calculate the Charlson Comorbidity Index (CCI) [19] were obtained from the quality management system for hospital mortality registration. Data verification was performed manually. Collected data were de-identified and stored on a secure hospital computer. There were no identifiable paper documents.

Descriptive data are reported as means and standard deviation (SD) or median and interquartile range in case of skewed distributions, or as frequencies and percentages when appropriate. For the primary analysis, comparing the EE before and after cisatracurium infusion, a general linear mixed model analysis for repeated measures was performed with an autoregressive covariance structure. In this analysis, we corrected for body temperature, sedative and noradrenaline dosages, pH, PEEP, and FiO₂ and repeated measurements.

The data analyses were performed using IBM corp. SPSS statistics for Windows (version 24.0, released 2015 New York, USA).

During the study period, 4247 patients were admitted to the ICU, of which 179 received cisatracurium for at least 12 h and therefore were eligible for inclusion. We excluded 57 patients according to the exclusion criteria. In total, 122 patients were enrolled in this study (Fig. 1).

Baseline characteristics and nutritional parameters are shown in Tables 1, 2, and 3. The median age was 65.5 years, and 36.1% were female. The median SOFA and APACHE II scores on admission were 8 and 22, respectively. Most patients were septic (58.2%) and admitted to the ICU because of medical reasons (73%). A median ICU and hospital LOS of 15 and 23 days were found. The in-hospital mortality was 28.7%.

The mean EE was 1974 kcal/day before cisatracurium infusion (= control period) and 1888 kcal/day during cisatracurium infusion resulting in a mean difference of − 85.9 kcal (95% CI − 151.8 to − 20.0; p = 0.011). After correction for body temperature, sedative and noradrenalin dosages, pH, PEEP, and FiO₂ in mixed model multivariable analysis, the significant treatment effect of cisatracurium on EE persisted, with a mean difference of − 132.0 kcal (95% CI − 212.0 to − 52.0; p = 0.001). Cisatracurium significantly lowered EE by 6.6% (95% CI 2.6–10.6%). The results are depicted in Fig. 2.

A significant non-linear positive association between body temperature and EE was found (Spearman’s rho = 0.486, p < 0.001; Fig. 3).

Mean EE was 1805 kcal (95% CI 1721–1888) in non-septic patients and 1909 kcal (95% CI 1838–1978) in septic patients (p = 0.062). In mixed-model multivariable analysis, a significantly lower EE was observed in non-septic patients than in septic patients (mean difference − 120.6 kcal, 95% CI − 200.5 to − 40.8; p = 0.003).

Only seven patients (5.7%) received > 110% of their caloric target (estimated by EEVCO₂) on the first day of cisatracurium infusion. Twenty patients (16.4%) received between 80 and 110% of their caloric target, while 95 patients (77.9%) were fed hypocalorically (< 80% of caloric target). Because of the small number of patients with hypercaloric intake, no associations between hypercaloric intake and ICU LOS or mortality were calculated.

We observed a non-linear positive association between body temperature and EE. Four small previous studies reported an association between body temperature and EE in critically ill patients [4, 24, 25]. A reduction of 6.6% of EE per 1 °C decrease at temperatures below 36 °C and an increase of 8.2% per 1 °C at temperatures above 37 °C have been reported [24, 25].

We observed a higher EE in septic patients than in non-septic patients. This was in line with our expectations based on previous studies in which EE in septic patients was 102–198% of EE in non-septic patients [4]. However, a recent observational study in 205 patients found no differences in EE between septic and non-septic patients (1434 vs. 1430 kcal/day) [20].

This is the largest cohort of critically ill patients in which the effects of NMBAs on EE have been studied. The effects of NMBAs, especially cisatracurium, in this specific patient population have not been studied before. A large number of patient variables were available with few missing data, providing enough data to perform rigorous multivariable and repeated measure analyses.

However, our study has several limitations. Indirect calorimetry was not routinely available during the study period. Therefore, EE was calculated using VCO₂ obtained from the mechanical ventilator. Calculation of EE from VCO₂ has been demonstrated to be more accurate than predictive equations, but less than indirect calorimetry. Finally, limitations related to the retrospective design may potentially have introduced bias and residual confounding.

As cisatracurium reduces EE, reduction of caloric intake after the start of NMBAs should be considered, especially in those patients that are on full feeding or considered to reach this target soon, because they are at risk of hypercaloric feeding and associated harm. Before we designed the study, we expected, due to the drop in EE induced by the NMBA, that some of the patients would be overfed. Based on the results, we noticed that a reduction of EE by NMBA could induce an almost 10% overfeeding risk in individual patients. In daily practice, this did not occur as the patients were not on nutrition target. Thus, for most patients, adjustment may not be necessary as in our analysis the reduction of EE found was only 6.6% and hypercaloric feeding was only present in 5.7%, while most other patients were fed (77.9%) hypocalorically after initiation of cisatracurium infusion.

Although not the focus of our present study, it should be noted that the recent ROSE trial, studying the effect of early neuromuscular blockade (48-h continuous infusion of cisatracurium) with concomitant heavy sedation, compared with usual care, did not result in a significant mortality difference at 90 days in patients with moderate to severe acute respiratory distress syndrome in contrast to an earlier RCT [26, 27]. This trial was stopped early at the second interim analysis for futility. This study may lead to reevaluation of the use of NMBAs in severe respiratory failure.