Nutrition in the intensive care unit and early recovery influence functional outcomes for survivors of critical illness: A prospective cohort study

Abstract

Background:

Patients who are critically ill may receive suboptimal nutrition that leads to weight loss and increased risk of functional deficits.

Methods:

Our overarching hypothesis is that nutrition in the intensive care unit (ICU) and the early recovery phase associates with functional outcomes at short-term follow-up. We enrolled adult patients who attended the University of Kentucky ICU recovery clinic (ICU-RC) from November 2021 to June 2022. Patients participated in muscle and functional assessments. Nutrition intake and status during the ICU stay were analyzed. The Subjective Global Assessment and a nutrition questionnaire were used to identify changes in intake, ongoing gastrointestinal symptoms, and patient’s access to food at the ICU-RC appointment.

Results:

Forty-one patients enrolled with a median hospital length of stay (LOS) of 23 days. Patients with 0 days of nil per os (NPO) status throughout hospitalization had a shorter LOS (P = 0.05), were able to complete the five times sit-to-stand test (P = 0.02), and were less likely to experience ICU-acquired weakness (P = 0.04) at short-term follow-up compared with patients with ≥1 day of NPO status. Twenty (48%) patients reported changes in nutrition intake in early recovery compared with before hospitalization. Eight (20%) patients reported symptoms leading to decreased intake and four (10%) reported access to food as a barrier to intake.

Conclusion:

Barriers to nutrition exist during critical illness and persist after discharge, with almost half of patients reporting a change in intake. Inpatient nutrition intake is associated with functional outcomes and warrants further exploration.

INTRODUCTION

Patients who are critically ill are at high risk for developing malnutrition during their hospital stay.¹ Malnutrition in the intensive care unit (ICU) can lead to muscle atrophy, impaired physical function, and adverse posthospital events.² Specifically, reduced nutrition intake in the ICU increases the risk of posthospital mortality.^3,4 Optimal prescription and delivery of nutrition support in the ICU has been shown to be clinically significant for improving patient-centered outcomes.¹ In patients who are critically ill, achieving the optimal nutrition intake is challenging and rarely successful.^5,6

Current literature and expert consensus vary on the recommended prescription to support adequate nutrition intake.^7,8 Goal energy prescriptions fluctuate based on the patient’s severity of illness, medications, procedures, and many other factors that change daily during an ICU admission. The recommendations emphasize initiating nutrition support within 48 h of ICU admission.⁸ The goal for adequate nutrition for critically ill patients is 50%–65% or greater of the prescribed amount by the end of the first 7 days of ICU stay.⁸ Studies have shown that reaching that goal is easier with continuous feeding,⁹ but it is an overall challenge to meet it even with structured protocols.¹⁰

In practice, it is accepted that nutrition support in the ICU and during recovery is important for improving outcomes; however, survivors of critical illness continue to experience inadequate nutrition intake after hospitalization. The aforementioned, paired with reduced physical activity in the ICU, may explain the long-term muscle and functional impairment.^10,11 Clinical observation, including patient testimony at the University of Kentucky ICU recovery clinic (UKY ICU-RC), suggests that a high percentage of individuals may not be reaching adequate nutrition intake in the ICU or upon returning home. Thus, we sought to examine if nutrition intake in the ICU and the early recovery period is associated with short-term functional and nutrition outcomes. Secondarily, we examined if patients having >1 day of nil per os (NPO) status during hospitalization had worse functional outcomes at the ICU-RC.

METHODS

Study design

A prospective cohort study was conducted at the University of Kentucky ICU Recovery Clinic (UKY ICU-RC).

Ethics

The study was approved by expedited review at the University of Kentucky (MEDXP# 47751). Informed consent was waived, as patients participated in outcome measures considered standard of care in the ICU-RC.

Hypothesis

Our overarching hypothesis is that nutrition in the ICU and the early recovery phase associates with functional outcomes at short-term follow-up. Specifically, we hypothesized that patients with >1 day of NPO status during hospitalization would perform worse on the chair rise test (slower times) at 1 month after hospital discharge.

Patient population

Our population consisted of adult patients admitted for acute respiratory failure who required mechanical ventilation (MV) or high oxygen requirements, such as high-flow nasal cannula (HFNC), in the medical ICU and attended the UKY ICU-RC 4–8 weeks after hospital discharge between November 2021 and June 2022. Referrals for ICU-RC and the standard of care have been previously reported.¹¹ In brief, the ICU-RC provides medical and rehabilitation care for survivors of critical illness who have had a diagnosis of acute respiratory failure or sepsis requiring MV or HFNC in the medical ICU. The inclusion of MV and HFNC is consistent with recent research demonstrating that patients on HFNC have a similar partial pressure of oxygen/fraction of inspired oxygen ratio equating to the same criteria on acute respiratory distress syndrome grades by the Berlin definition.^12,13 Therefore, our inclusion focused on patients with a high acuity of illness that would receive follow-up in the ICU-RC. The requirement and length of time on MV was used to determine the severity of illness in our patient population. Patients participated in nutrition, strength, and functional assessments conducted by one of two physical therapists with critical illness recovery expertise. Our exposure of interest was days of NPO status, defined as ≥1 day of no nutrition consumed in the hospital. Our comparator was 0 days of NPO status during hospitalization. Patients were grouped by days of NPO status: (1) 0 days of NPO status and (2) ≥1 days of NPO status.

Demographic and clinical variables

Demographics extracted from the electronic health record (EHR) included age, sex, body mass index (BMI), and race/ethnicity. The comorbid burden was quantified using the functional comorbidity index (FCI). Clinical data extracted from the EHR included ICU and hospital length of stay (LOS; days); receipt and duration of MV; weight (kg) at hospital admission, discharge, and UKY ICU-RC appointment; and discharge disposition.

Primary outcome

The primary outcome was performance on the chair rise test, that is the five times sit-to-stand (5xSTS) test. The 5xSTS is a timed, functional test that assesses lower extremity strength and power through repeated stands from a chair.¹⁴ To complete the test, patients were asked to sit and stand five times from a chair as quickly as possible without using their arms. Faster times are associated with better functional performance and outcomes.¹⁵

Secondary outcomes

Secondary muscle and physical function outcomes included the Short Physical Performance Battery (SPPB) and the Medical Research Council sum score (MRC-ss) as well as 4-m gait speed, a component of SPPB. The SPPB is calculated based on three components: gait speed, 5xSTS, and balance assessments; a higher score indicates better physical function.¹⁶ The MRC-ss is a measure of the muscular strength of six major muscle groups with a score of <48/60 indicating a clinical diagnosis of ICU-acquired weakness (ICUAW).¹⁷ Nutrition status was assessed by subjective reporting at the ICU-RC appointment using the Subjective Global Assessment (SGA) and a postdischarge nutrition questionnaire. The SGA is a valid and reliable physical and observational assessment for malnutrition.¹⁸ A self-report questionnaire was designed with dichotomous and ordinal scoring to examine nutrition intake in the recovery period. Six questions were presented to understand the following:

• Current appetite and intake of energy and protein in relation to intake prior to hospitalization

• New aversions to food

• Active gastrointestinal symptoms affecting nutrition intake

• Access to food limiting nutrition intake (“access” was defined to patients as the ability to purchase healthy food, ie, financial means and grocery stores nearby)

• If the patient needs assistance to prepare meals and, if assistance is required, does the individual have the social support to prepare meals

Data on inpatient nutrition intake were extracted from the EHR. Days without any intake, defined as orders with NPO, were recorded. Enteral nutrition intake in milliliters was extracted from the nursing flowsheets for every 24-h period of hospital admission. Patients were considered to be NPO if no intake was received during a 24-h period of time. Energy intake in kilocalories was calculated based on the milliliters of enteral nutrition received and the prescribed formula type for 76% of patients (n = 31). Nutrition data were not collected for patients on oral diet orders dbecause of inadequate medical record documentation of meal intake.

Because of the average BMI of our patients, we used the recommended requirements of nutrition for critically ill obese patients of 11–14 kcal/kg/day to calculate how many of our patients were meeting the prescribed goal.⁸ Using the average energy and BMI grouped by days of NPO status, we calculated the amount of energy required for each group compared with what was being delivered with enteral feedings during hospitalization (Table 1). Patients were identified as not meeting adequate nutrition intake if <65% of the prescribed nutrition was not being received.¹⁹

TABLE 1.

Nutrition calculation based on enteral nutrition intake during hospitalization.

	≥1 Day NPO (n = 26)	0 Days NPO (n = 15)
Average weight at admission, kg	112.4	100.4
Recommended energy (11–14 kcal/kg/day)	1236–1573	1104–1405
Average energy consumed by day, kcal	1110	1049
Adequate intake,a n (%)	18 (69)	5 (33)

^aAdequate intake was defined as meeting ≥65% of the lower part of the recommended intake (11 kcal/kg/day).

Statistical analysis

Descriptive statistics were performed to examine the central tendencies and variance of the data. Patients were grouped based on nutrition intake: (1) patients with 0 days of NPO status during admission and (2) patients with ≥1 day of NPO status. A sensitivity analysis was conducted with a second stratification based on literature stating that nutrition support should be initiated within 48 h of admission to the ICU: (1) patients with >2 days NPO and (2) ≤2 days of NPO status. Days were counted as NPO if no intake was received throughout the 24-h period. Independent t tests were performed to examine group differences. Exploratory multivariate regressions were completed to understand the relationship between candidate variables, selected by the team in concert with a statistician (S.S.), and functional outcomes. Multivariate linear regressions were performed to explore the relationship between demographics (age, sex, and BMI) and clinical data (diagnosis of coronavirus disease 2019 [COVID-19] and the number of days of NPO status as continuous variables) with performance on gait speed, MRC-ss, and 5xSTS. Because only 31 (76%) of the patients were able to complete the 5xSTS test, multivariate logistic regression models were run to calculate the odds that a patient could complete the test. Finally, multivariate logistic regression was performed to explore the diagnosis of ICUAW: MRC-ss was dichotomized into clinical weakness (<48) or no weakness (48+). Spearman rho correlations were performed to assess the association between independent clinical and nutrition variables in the hospital with outcomes at short-term follow-up in the entire cohort.

RESULTS

Forty-one patients were enrolled and participated in the study with a median age of 48 (range = 23–83) years, with 61% being male with a median BMI of 32.6 (range = 19.5–56.9) kg/m². The median hospital LOS was 23 (range = 5–170) days, and the ICU LOS was 21 (range = 4–97) days. The median FCI was 1 (range = 0–8) for the entire patient population. Six patients never required MV during ICU stay but required HFNC. To note, of these six patients without MV, only two were included in grouped analysis, with one patient having no days of NPO status and one patient with >1 day of NPO status. For days on MV, the median duration was 9 (range = 0–92) days.

The average short-term follow-up was 7 ± 4 weeks after hospital discharge for patients attending the ICU-RC. All the patients completed the SGA and nutrition assessment. Twenty-one patients (51%) had a grade of B on the SGA, which indicates mild-to-moderate malnourishment and some progressive nutrition loss¹⁸ (Table S1). On the nutrition questionnaire, eight patients (20%) reported gastrointestinal symptoms and food aversions limiting nutrition intake after discharge. Twenty-two patients (52%) reported the same intake prior to hospitalization. Four (10%) patients reported barriers to food access that affected their nutrition after discharge (Table S1).

Thirty-one (76%) patients included in this study received enteral nutrition, with the remaining 10 not requiring enteral nutrition. Recommended inpatient nutrition intake goals were met for 23 of the 31 patients requiring enteral feedings (Table 1).

Of the 41 patients, 63% of patients had ≥1 day of NPO status, and 34% of patients had ≥3 days of NPO status. There were no differences in age, BMI, FCI, ICU, and hospital LOS and weight at all the time points between the two groups (no days of NPO status and >1 day) (Table 2). Differences were consistent when using 48 h of NPO status as a cutoff between groups (Table S2).

TABLE 2.

Patient characteristics.

	Cohort (n = 41)	≥1 Day NPO (n = 26)	0 Days NPO (n = 15)	P
Age, median (range), y	48 (38–62)	49.5 (39–63)	43 (35–53)	0.21
Male sex, n (%)	25 (61%)	18 (69%)	7 (47%)	0.15
BMI, median (range), kg/m2	32.6 (26.8–38.3)	32.7 (27.4–39.0)	32.6 (24.1–37.1)	0.41
FCI, median (range)	1 (1–2)	2 (1–2)	1 (1–2)	0.63
Diagnosis, n (%)
ARF caused by				0.89
COVID-19	33 (81)	20 (77)	13 (87)
PNA	5 (12)	4 (15)	1 (7)
AKI	1 (2)	1 (4)	0 (0)
Other	2 (5)	1 (4)	1 (7)
Hospital LOS, median (range), days	23 (16–50)	29.5 (21–61)	22 (11–34)	0.05a
ICU LOS, median (range), days	21 (11–44)	23.5 (11–49)	18 (9–31)	0.37
MV duration, median (range), days	9 (3–22)	9.5 (5–32)	8.5 (0–16)	0.32
Weight at admission, median (range), kg	111.4 (92–122)	112.5 (92–123)	105.5 (80–114.7)	0.16
Weight at discharge, median (range), kg	95.2 (75–112)	91.5 (75–119)	96.9 (71–111.9)	0.89
Weight at clinic, median (range), kg	99.4 (82–118)	103.5 (84.4–123)	97.1 (71.9–108)	0.27
Weight change, median (range), kg
Admission to DC	−8.9 (−14.9 to −3.0)	−12.6 (−21.2 to −5.0)	−3.4 (−9.0 to 0.7)	0.02
DC to ICU-RC	0.9 (−3.6 to 7.1)	5.0 (−1.0 to 10.0)	−2.7 (−7.0 to 0.9)	0.003a
DC disposition, n (%)
LTACH	3 (7)	2 (8)	1 (7)	0.37
Rehab facility	23 (56)	17 (65)	6 (40)
Home with HH	4 (10)	2 (8)	2 (13)
Home	11 (27)	5 (19)	6 (40)

Abbreviations: AKI, acute kidney injury; ARF, acute respiratory failure; BMI, body mass index; COVID-19, coronavirus disease 2019; DC, discharge; FCI, functional comorbidity index; HH, home health; ICU-RC, intensive care unit recovery clinic; LOS, length of stay; LTACH, long-term acute care hospital; MV, mechanical ventilation; PNA, pneumonia.

^aSignificant difference between groups.

Grouped statistics

Patients with ≥1 day of NPO status had a longer median hospital LOS (P = 0.05) when compared with patients with no days of NPO status (Table 2). Weight change was a significant finding between the groups. Patients with no days of NPO status had less weight change during hospital admission and from discharge to ICU-RC appointment (P = 0.02 and 0.003) compared with patients with ≥1 day of NPO status. Patients with ≥1 day of NPO status gained an average of 5 lb from discharge to ICU-RC follow-up (Table 2). Patients with no days of NPO status had faster 5xSTS time, faster gait speed, greater functional strength (t = 0.04, P = 0.04) and higher SPPB score than patients with ≥1 day of NPO status (Table 2 and Figure 1). Similar trends were observed for hospital LOS (P = 0.05) and weight change from discharge to ICU-RC appointment (P = 0.003) between groups in our secondary analysis. Patients with >48 h of NPO also had a longer ICU stay compared with patients with <48 h of NPO (t = 0.22, P = 0.02) (Table S1). There were no significant differences in the results of the nutrition assessments at short-term follow-up when examined between the two groups.

FIGURE 1.

A violin plot depicting the difference in functional outcomes scores collected at the ICU Recovery Clinic for patients with no days of NPO status (0_days) versus patients with at least one day of NPO status (24_days). MRC-ss, Medical Research Council sum score; SPPB, Short Physical Performance Battery; 5xSTS, five times sit to stand.

Cohort statistics

For the entire cohort (n = 41), a shorter LOS was associated with less weight loss during hospital admission (r = −0.49, P < 0.001) and a greater likelihood to discharge home (r = 0.34, P < 0.001). Discharge home was also associated with fewer days receiving enteral feedings (r = −0.52, P < 0.001).

Multivariate regressions (Table 3)

TABLE 3.

Multivariate models examining functional outcomes of ICU survivors.

	Dependent outcomes at short-term follow-up
Candidate variables Parameter	Inability to stand on 5xSTSa		5xSTS		ICUAWb		MRC-ss		Gait speed
	Odds ratio	P	Estimate	P	Odds ratio	P	Estimate	P	Estimate	P
Intercept			12.34	0.003			61.83	<0.001	1.36	<0.001
Age	1.05	0.168	0.03	0.363	1.09	0.080	−0.02	0.794	−0.01	0.025
BMI	1.09	0.167	−0.09	0.260	1.15	0.046	−0.11	0.331	−0.003	0.416
Sex	2.43	0.366					−0.89	0.686	−0.06	0.453
COVID-19	0.18	0.125	−1.66	0.298			−6.69	0.026	−0.24	0.036
Days of nothing by mouth	1.92	0.019	0.67	0.040	2.21	0.039	−0.81	<0.001	−0.02	0.009

Note: Highlighted green values are significant with P values < 0.05.

Abbreviations: BMI, body mass index; COVID-19, coronavirus disease 2019; ICUAW, intensive care unit–acquired weakness; MRC-ss, Medical Research Council sum score; 5xSTS, five times sit to stand.

^aInability to stand on 5xSTS test as a binary outcome was examined with multivariate logistic regression.

^bICUAW, a binary diagnosis of clinical weakness due to critical illness (based on a cutoff score of <48/60 on the MRC-ss) was analyzed with multivariate logistic regression. Because of the small sample of patients with ICUAW, adding the sex and COVID-19 variables led to unreliable odds in the model and were, therefore, excluded.

Patients with ≥1 day of NPO status were 9.5 times more likely to not be able to complete the 5xSTS assessment, although the difference did not reach statistical significance (P = 0.07). However, if the number of days of NPO status is considered as a continuous measure, statistical significance is reached (P = 0.02), and the odds of not being able to complete the 5xSTS almost doubles for each additional day of NPO status (odds ratio = 1.92; 95% CI = 1.12–3.30). Models focusing on the completion time for the 5xSTS (n = 31) demonstrate similar results. Because the MRC-ss was highly skewed to the right, ICUAW as measured by the MRC-ss, was predicted. Of the 39 patients who completed the MRC-ss, only 6 qualified for diagnosis of ICUAW (score <48/60 on MRC-ss). Similar to the 5xSTS results, patients with ≥1 day of NPO status vs those with no days of NPO status did not significantly predict weakness, but the number of days of NPO status did. The diagnosis of COVID-19 as an independent variable was excluded from the weakness model because of separation of the data for COVID-19 status (Table 3). Each additional day of NPO status increased the odds of clinical weakness by 121% (P = 0.04) (odds ratio = 2.21; 95% CI = 1.04–4.73). BMI was also trending in relation to strength measured by the MRC-ss (P = 0.08). For gait speed, each additional day without food associated with a decreased gait speed by 0.02 m/s (P = 0.01).

DISCUSSION

Our study demonstrates that nutrition support and intake in early recovery after critical illness may influence weight changes and functional status. Patients in our study experienced significant weight loss and reported altered nutrition intake caused by early satiety or altered appetite after hospital discharge. Of clinical significance, we demonstrate that one in five patients surviving critical illness have persistent symptoms that are barriers to nutrition one month after discharge. Our data emphasize that nutrition status, as well as access to nutrition, should be considered and examined for ICU survivors.

Meeting nutrition goals throughout hospitalization is challenging and often leads to poor nutrition intake during a hospital stay.^20,21 Our data reflect similar trends with eighteen (44%) patients receiving <65% of the prescribed nutrition intake. We also know that physical function is decreased after a longer hospital stay because of immobility and severity of illness.^22,23 Better functional outcomes were associated with more nutrition consumed throughout the entirety of hospitalization measured by fewer days of NPO status. Similar findings have been reported for length of stay and how severity of illness impacts weight loss and nutrition status similar to our results.²⁴ Although there is clinical significance to our findings, our study is underpowered and, therefore, has a low probability of rejecting the null hypothesis when it is actually false.

We hypothesized that patients with >1 day of NPO status during hospitalization would have worse functional outcomes at short-term follow-up. Our second hypothesis was that inpatient and early recovery phase nutrition is associated with patient outcomes. Patients reporting to the ICU-RC demonstrate impaired physical strength and function at short-term follow-up, which is consistent with other research in ICU survivors.²⁵ Additional days of NPO status were associated with greater odds of having ICUAW at short-term follow-up. Functional outcomes clinically differed between groups (Figure 1) at short-term follow-up; however, many confounders, such as rehabilitation during hospitalization, prior level of function, postacute rehabilitation and severity of illness, also impact function and were not collected in our study. Weight change was significantly different between groups, with patients having no days of NPO status continuing to lose weight from discharge to ICU-RC appointment. The patients in the no days of NPO status group lost less weight during their hospitalization compared with the patients with at least 1 day of NPO status; therefore, less weight gain would be expected. Some of the patients with no days of NPO status also had longer hospitalization and more time spent in facilities after discharge before returning home or attending the ICU-RC. Another aspect of short-term follow-up assessed by our study was changes in nutrition intake after discharge. Twenty percent of patients in our study reported symptoms such as new food aversions and gastrointestinal issues impacting nutrition intake, which have also been reported in patients surviving critical illness.²¹ Adequate nutrition is required to mitigate the negative effects of nutrition deficit during hospitalization.^26,27 During critical illness, the body’s metabolic rate rises in response to the increase in physiological stress.²⁸ Current literature reflects the heterogeneity of nutrition support and the difficulty meeting the prescribed amounts of enteral feedings during the acute phases of a critical illness.^20,29 Nutrition data gathered from this study support the evidence that nutrition intake in the ICU rarely meets or exceeds daily intake goals.^8,19 Patients admitted to the ICU require timely, life-saving care that limits the ability for nutrition intake early on.

Nutrition support should not end at hospital discharge for ICU survivors. Issues with difficulty accessing food, or the inability to consume adequate nutrition, need to be addressed in the early recovery phase. Ten percent of the patients in our study reported that access to food was a barrier to adequate nutrition intake, and we suggest that this number is underreported in our study and this patient population because of the vulnerability of the information. Hospitals should incorporate screening tools to identify a patient’s social needs. Patients should be connected with social workers and appropriate resources prior to discharge if unmet needs are assessed. Postdischarge weight loss, changes in appetite, and self-reported symptoms limiting adequate intake warrant further nutrition support through outpatient nutrition referrals. Future studies should investigate how dietitians can support ICU survivors as well as the impact adequate inpatient and postacute nutrition has on short-term and long-term patient outcomes.

Our patient population was enrolled upon attendance to the UKY ICU-RC. On average, one to two recently discharged patients attend clinic per week thus limiting our enrollment and number of patients during the study period. The majority of our patient population were survivors of critical COVID-19 illness. The short-term and long-term effects of COVID-19 are continuously evolving. The effects of COVID-19 on nutrition intake, gastrointestinal symptoms, appetite, and taste may have impacted our subjective results, although many patients did not report food aversions as a barrier to adequate nutrition. The subjective questionnaire allowed patients to report their symptoms, intake, perceived weight loss, and access to food, which all proved to be valuable information to understand postacute nutrition status in ICU survivors. A barrier to assessing social needs is the limited time spent building trust and rapport for the patient to share vulnerable information. Hence, the observed changes in body language, tone, and delayed response time may warrant further exploration through a trusted relationship and qualitative methodology. The nutrition assessment took <5 min to complete and could be easily added to follow-up care for ICU survivors without burdening patients and healthcare providers. This study was the first step in assessing nutrition in these early phases of recovery in our ICU-RC; therefore, reliability and validity measures were not assessed. Only two providers completed the nutrition assessment with patients during the study period. Patient demographics and inpatient nutrition intake data pulled from the EHR were for patients receiving enteral feeding only, limiting the generalizability to all ICU survivors. Medications that impact macronutrient amounts and energy were not considered in the amount consumed for each patient but may be valuable in describing inpatient nutrition status in future research.

CONCLUSION

Our study elicits the nutrition intake barriers after discharge for ICU survivors and the relationship to short-term functional outcomes. Strength deficits existed in patients who had more days of NPO status. Our patients reported access to nutrition and symptoms limiting intake in the early recovery phase. We identified areas for further exploration for nutrition support after an ICU admission and how inpatient nutrition is associated with postacute outcomes, specifically physical function. Ensuring adequate nutrition should not stop at hospital discharge. Providers should continue to address nutrition deficiencies and barriers after discharge to enhance patient recovery and post-discharge outcomes.

CLINICAL RELEVANCY STATEMENT.

Patients experiencing critical illness may not receive adequate nutrition in the intensive care unit as well as at home after hospital discharge. Findings from our observational study demonstrate that nutrition from admission to the early recovery phase is associated with functional outcomes. Furthermore, we demonstrate that survivors of critical illness have nutrition needs at home with 10% of patients reporting reduced access to healthy food and roughly 50% reporting altered intake compared to pre-hospital baseline. The findings suggest that nutrition intake should not only be addressed in the ICU, but the early recovery period when patients are first returning home.