The global pandemic caused by the novel coronavirus, SARS-CoV-2 (COVID-19) (World Health Organization (WHO), 2022a) has posed unprecedented challenges for healthcare systems worldwide (Barazzoni et al, 2020). There is significant variance in the clinical sequelae associated with COVID-19, ranging from no to moderate (fever, cough, shortness of breath, muscle ache, confusion, sore throat, headache, pneumonia, diarrhoea, nausea and vomiting, loss of taste and smell) to severe symptoms (respiratory failure from COVID-19 pneumonia, requiring ventilation and intensive care). Moderate symptoms may adversely impact nutritional intake, resulting in loss of weight and muscle and malnutrition (Wierdsma et al, 2021), while severe symptoms may deteriorate to multi-organ failure (Australasian Society of Parenteral and Enteral Nutrition (AuSPEN), 2020; Barazzoni et al, 2020; Handu et al, 2021).
Numerous studies show COVID-19 impacting on malnutrition risk, weight loss, and nutritional intake, with malnutrition prevalence ranging from 35% on admission (Bedock et al, 2020; Li et al, 2020; Rouget et al, 2020; Di Filippo et al, 2021; Wierdsma et al, 2021) to 53% during hospitalisation (Li et al, 2020). Additionally, a recent study (Wierdsma et al, 2021) found that 93% of patients hospitalised with COVID-19 experienced symptoms affecting nutritional intake, and 22% lost weight during hospital admission. In general, in hospitalised populations, malnutrition negatively affects morbidity and mortality, and studies show that patients, particularly those with a long intensive care unit (ICU) admission, often require prolonged hospitalisation and rehabilitation, with many not returning to baseline levels of function (Singer et al, 2019; Barazzoni et al, 2020). No studies were identified that establish the prevalence of nutrition impact symptoms among those managed in the community with COVID-19, in contrast to those unwell enough to be hospitalised, in countries with high COVID-19 case numbers.
At the commencement of the pandemic, Australia was advantaged in having small case numbers, with all patients initially hospitalised regardless of symptom severity due to a local state government direction (Queensland Government, 2021). This directive provided a unique opportunity for assessing the type and severity of symptoms across all people with COVID-19, which contrasts with currently published literature comprising those unwell enough to require hospitalisation. With the majority of COVID-19 cases remaining in the community, an understanding of symptom type and prevalence is important for care provision, whether by families or community-based health services.
Among all people who tested positive for COVID-19, this study aimed to 1) identify malnutrition risk, nutritional status and weight change, 2) assess nutritional intake and nutrition impact symptoms, and 3) describe dietitian review frequency and mode of dietetic consult.
Materials and methods
A retrospective observational study was completed over 9 months, from the first positive case identified on the ward (3 March 2020) until 30 November 2020, at the Princess Alexandra Hospital, a large tertiary teaching hospital in Brisbane, Australia. Due to local government direction at that time (Queensland Government, 2021), all COVID-19 positive people were required to be admitted to an infectious diseases ward. The protocol was approved by the Human Research Ethics Committee of Metro South Health (HREC/LNR/2020/QMS/67902).
Participants and timeline of admission
All adult patients with a confirmed COVID-19 diagnosis admitted to the infectious diseases ward from any source, including ICU, another ward, emergency department or directly from the community, were included as participants. The admission timeline explored was categorised into days from admission as follows: Day 0–4; Day 5–10; Day 11–16; Day 17–22; and Day ≥23.
Malnutrition screening
As per hospital wide procedures, malnutrition risk screening was documented within the Electronic Medical Record (EMR) Interactive View—Adult Risk Assessment. Protocol specifies that the Malnutrition Screening Tool (MST) (Ferguson et al, 1999) is completed within 8 hours of transfer or admission, and weekly or on change in the patient's condition thereafter. An MST score of 2 or more indicates that the patient is at risk of malnutrition and generates an automatic referral to the dietitian for full assessment.
Nutritional status
During hospital admissions, nutritional assessments are performed using a Subjective Global Assessment (SGA) (Detsky et al, 1987) (SGA A: well-nourished; SGA B: mild to moderately malnourished; or SGA C: severely malnourished), which incorporates an assessment of fat and muscle mass. If a full SGA was unable to be completed, nutritional status was captured using the Body Mass Index (BMI), with <18.5 kg/m2 being underweight as per WHO criteria (WHO, 2022b), and weight loss (reported or recorded) during or prior to hospital admission, with >5% weight loss considered significant (Detsky et al, 1987).
Nutritional intake and nutrition impact symptoms
Reports of reduced oral intake documented by clinical staff in the EMR during admission was recorded as evidence of reduced oral intake for data analysis purposes. For patients reviewed by the dietitian during their admission, an assessment of estimated energy and protein intake was compared to estimated requirements. The nutrition impact symptoms of reduced appetite, cough, sore throat, nausea, vomiting, diarrhoea, loss of smell and loss of taste were self-reported on admission and at any stage during admission. All reports of an nutrition impact symptom by clinical staff were recorded.
Dietitian consultation
The frequency and mode of dietitian consultation (face-to-face or telephone) was recorded. A dietitian referral could be either automatically generated via an MST triggered electronic health record order, or clinician-initiated by nursing or medical staff due to the need for a modified diet, clinical concerns about reduced oral intake or per protocol when admitted to ICU before or during their admission to the infectious diseases ward.
Due to precautions to minimise contact exposure, as directed by hospital processes, only essential physical care was performed. The wearing of personal protective equipment (PPE) was recommended when undertaking any face-to-face contact with COVID-19 patients. As a consequence, dietitian consultations were performed via chart review, utilising the EMR and a telephone call as required. Patient contact was attempted via the bedside telephone, a portable phone delivered by nursing staff, or the patient's mobile telephone. Face-to-face contact with dietetics was only performed in those patients that had a lengthy admission and had subsequently tested negative following their infectious period.
Statistical analysis
Statistical analysis was performed using SPSS software (version 23.0), with a significance level of 5%. Continuous data was tested for normality, with parametric data presented as mean ± standard deviation, with independent t-tests applied, and non-parametric presented as median (range), with Mann-Whitney U tests applied. Fisher's exact tests was used with categorical data. Pearson or Spearman correlation was used for associations depending on normality.
Results
During the study period, 60 patients (52% female) were admitted to the infectious diseases ward. There were 58 people admitted directly to the infectious diseases ward (one later requiring transfer to ICU and then returning to the ward) and two people being transferred to the ward after their initial admission to the ICU. Average length of stay was 10.7±5.3 days. Mean age was 42.5±16.8 years, and median BMI was 26.8 (range 20.2; 60.6) kg/m2, with 69% (n = 38) considered overweight or obese according to WHO BMI criteria (2022b) (Table 1), which is reflective of the overweight and obesity prevalence in the Australian population (Australian Government, 2020).
Table 1. Patient demographics
| Characteristic | Number (%) | Mean (standard deviation) | Median (range) |
|---|---|---|---|
| Gender | 31 (52) | ||
| Female | |||
| Age (years) | 42.5 (16.8) | ||
| ≤65 | 54 (90) | ||
| Length of stay (days) | |||
| In the infectious diseases ward | 60 (100) | 10.7 (5.3) | |
| In the intensive care unit | 3 (5) | 17 (3−40) | |
| When infectious disease ward and intensive care unit are combined | 60 (100) | 10 (3−40) | |
| BMI (kg/m2) 1 | 55 | 26.8 (20.2–60.6) | |
| 18.5 – 24.9 | 17 (31) | ||
| 25 – 29.9 | 20 (36) | ||
| ≥30 | 18 (33) | ||
| Weight (kg) 1 | 58 | 78.0 (52.8–173.0) |
Note:
1On admission to the infectious disease ward
Malnutrition risk, nutritional status, and weight change
There were 12 episodes (in ten patients) of malnutrition risk identified during the study period. Risk of malnutrition was higher as days since admission to the ward increased: Day 0–4 (5/60; 8%), Day 5–10 (2/51; 4%), Day 11–16 (1/22; 4%), Day 17–22 (2/9; 22%) and Day ≥23 (2/4; 50%).
No malnutrition or underweight status was identified according to the SGA or BMI. One SGA was completed following screening referral, with the individual confirmed as well-nourished. SGAs were unable to be completed the majority of the time, due to patient isolation requirements and hospital procedures. No BMI was recorded at <18.5 kg/m2.
Weight was recorded on admission to the infectious diseases ward for 97% (58/60) of patients, with 35% (21/60) having more than one weight recorded during their admission. Of those with a documented weight change, 57% (12/21) recorded weight loss with a median loss of 2.6kg (range of 0.2kg to 8.6 kg weight loss), and 25% (3/12) had weight loss of >5% (8.4%; 8.8%; 8.8%). Of the three patients (3/12; 25%) who recorded a weight loss of >5% during their admission, two patients (2/3; 67%) required an ICU admission.
Diet type and nutritional intake
Oral diet accounted for ~90% of nutritional intake from admission to day 22, and 75% in patients that had an admission of >22 days. No parenteral nutrition support was required. No enteral or parenteral nutrition was required during admission to the infectious disease ward. Enteral nutrition was required for the three individuals in the ICU.
Among those admitted, reduced oral intake remained steady throughout length of admission. On four occasions, the dietitian could calculate an estimated oral intake and compare to the estimated nutritional requirements and was, therefore, able to verify perceptions of reported reduced oral intake.
Nutrition impact symptoms
The median number of nutrition impact symptoms per patient was 3 (range of 0–12). On admission, 80% experienced at least one symptom, and 58% experienced at least two or more nutrition impact symptoms, with symptom number reducing as length of stay increased (Figure 1). The most common nutrition impact symptoms documented on admission was cough (60%; 36/60) followed by sore throat (35%; 21/60) and reduced appetite (28%; 17/60) (Figure 2). Cough remained one of the most common nutrition impact symptoms reported throughout the duration of admission.
There was a statistically significant association between reduced appetite and reduced oral intake (p=0.000) (Table 2). Of eight documented symptoms, reduced appetite, nausea, vomiting and cough were associated with reduced oral intake early in the admission (Table 2). Having two or more nutrition-impacting symptoms on admission was associated with reduced oral intake (p=0.006) (Table 3).
Table 2. Associations between reduced oral intake and different nutrition impact symptoms
| Nutrition impact symptom | On admission: day 0−4 (N=60) | Day 5−10 (N=51) | ||
|---|---|---|---|---|
| % (n) | Significance | % (n) | Significance | |
| Vomiting | 100 (5/5) | 0.0011 | 100 (1/1) | 0.2041 |
| Reduced appetite | 88 (15/17) | 0.0001 | 80 (8/10) | 0.0001 |
| Nausea | 69 (9/13) | 0.0011 | 83 (5/6) | 0.0011 |
| Diarrhoea | 50 (6/12) | 0.1511 | 50 (2/4) | 0.1801 |
| Cough | 39 (14/36) | 0.042 | 26 (6/23) | 0.4831 |
| Sore throat | 38 (8/21) | 0.2682 | 56 (5/9) | 0.0111 |
| Loss of taste | 30(3/10) | 1.001 | 0 (0/4) | 0.5691 |
| Loss of smell | 20 (2/10) | 0.7071 | 14 (1/7) | 0.5591 |
Note:
1Fishers exact 2 sided;
2Pearson Chi-Square
#Day >10 unable to be completed due to small sample size
Table 3. Associations between reduced oral intake and the frequency of nutrition impact symptoms*
| On admission | % (number) | Significance |
|---|---|---|
| >2 symptoms | 50 (17/34) | 0.0001 |
| >3 symptoms | 52 (13/25) | 0.0011 |
| >4 symptoms | 67 (10/15) | 0.0012 |
| >5 symptoms | 100 (4/4) | 0.0062 |
Note:
1Pearson Chi-Square;
2Fishers exact
*Associations with a single symptom are outlined in Table 2
Dietitian consultation and mode of consult
A full dietetic nutrition assessment occurred for 20% of patients (12/60), with 12% (7/60) requiring ongoing oral nutrition support and dietitian consult during their admission. The EMR triggered MST referrals to the dietitian and also enabled remote dietitian review via chart and telephone for 97% (36/37) of patient consults. One patient was consulted face-to-face following a lengthy admission and subsequent non-infectious (negative) COVID-19 diagnosis result.
Discussion
In contrast to the current study, many publications concerning nutrition in COVID-19 relate to those who require hospitalisation (Bedock et al, 2020; Li et al, 2020; Rouget et al, 2020; Di Filippo et al, 2021; Wierdsma et al, 2021). However, the majority of COVID-19 patients who experience mild to moderate symptoms are managed in the community (Handu et al, 2021). To the authors' knowledge, this is the first study to explore nutrition-impacting symptoms across all with a COVID-19 diagnosis.
The unique population of hospitalised, but generally ‘well’ patients, confirms that patients who experience mild to moderate COVID-19 may still experience symptoms that influence nutrition and be at risk of malnutrition or be malnourished. Therefore, nutritional care should be a fundamental part of management across the healthcare system. In this study sample of 60 patients, only three required an admission to ICU (with no mortality); however, the majority of patients on admission had symptoms that could impact their nutritional status, with two or more symptoms being significantly associated with reduced oral intake. The admission criteria resulted in a younger mean age (42.5 years) compared with other countries' studies (with a mean or median age over 59 years) (Bedock et al, 2020, Rouget et al, 2020; Di Filippo et al, 2021; Wierdsma et al, 2021). Additionally, it is acknowledged that older adults with COVID-19 experience worse outcomes and higher mortality (Barazzoni et al, 2020).
Medical nutrition therapy in this study population composed of oral nutrition support, with enteral nutrition only being necessary in the ICU setting. Again, this is reflective of this unique population, where the disease severity was not the trigger for hospital admission. Often, the ‘well’ hospitalised patients required proportionally fewer ICU admissions compared with other countries (Wierdsma et al, 2021) where higher rates of parenteral and enteral nutrition support are reported.
Early in the pandemic, symptoms that can impact nutritional intake, such as changes in taste and smell, were often reported in patients after COVID-19 infection (Carfi et al, 2020; Di Filippo et al, 2021; Doykov et al, 2021). Other symptoms, such as vomiting, nausea and cough (Carfi et al, 2020; Docherty et al, 2020), are also complaints with the potential to impact oral intake. It was found that 90% of COVID-19 patients requiring hospitalisation across several Dutch hospitals showed some form of nutritional complaint, with decreased appetite the most common at a prevalence rate of 58% (Wierdsma et al, 2021). In France, other authors identified that 46% COVID-19 patients with severe acute respitatory syndrome reported decreased oral intake, with reduced appetite (28%) being the most common complaint (Rouget et al, 2020).
The high prevalence of nutrition impact symptoms and nutrition-related complaints, combined with the systemic inflammatory response (Barazzoni et al, 2020) of this infectious respiratory disease, may contribute to reduced nutritional intake and lead to malnutrition in these patients, regardless of setting (hospitalised vs managed in the community). Malnutrition can be associated with poor health outcomes (Barazzoni et al, 2020). This study of nutrition impact symptoms confirms the value of malnutrition screening, as documented, as necessary in many COVID-19-specific global nutritional guidelines, consensus documents and recommendations for clinical practice (AuSPEN, 2020; Barazzoni et al, 2020; Cawood et al, 2020; Handu et al, 2021). It is recommended that malnutrition screening be completed for all patients with or recovering from COVID-19, regardless of whether they are managed in a hospital or community setting (Cawood et al, 2020; Handu et al, 2021), and it is emphasised that outcomes of this screening be linked to a documented nutrition management plan in both community and hospital settings (Cawood et al, 2020). Holdoway (2020) refers to practical web-based fact sheets and COVID-19-specific nutrition care pathway for professionals supporting patients at low, medium or high risk of malnutrition.
Although malnutrition risk in the sample appears relatively low (8% on admission), it is anticipated to represent the general population with the virus more closely. Studies among those admitted to hospital when unwell in other countries showed malnutrition risk prevalence, according to the Mini-Nutritional Assessment (MNA) of up to 28% (Li et al, 2020). However, the actual numbers of hospitalised individuals as a proportion of the population in these studied countries is anticipated to be far lower, due to the greater number of cases nationally (WHO, 2022a). However, if 8% of all people with COVID-19 were at malnutrition risk, the numbers within a community or a global population would be very large. Utilising the MNA, high rates of malnutrition risk in the community were evident, with 54% prevalence among COVID-19 patients discharged home to the community (Di Filippo et al, 2021). However, it should be noted that, despite the fact that the population studied by Di Filippo et al (2021) included patients managed at home in the community, most had been discharged from a medical ward (73%) and, therefore, had initially required hospitalisation.
While no malnutrition was identified during this study, 25% of patients with recorded weight loss lost >5%. According to the SGA, a weight loss of >5% in 6 months is deemed clinically significant and possibly contributes to malnutrition (Detsky et al, 1987). If SGA completion was possible, higher malnutrition prevalence may have been identified. Subsequently published literature (Vivanti, 2021) confirms the value in completing the SGA questions, even when physical assessment is not possible, as high associations with full SGA assessments (97%; 473/489 identical) were confirmed. Consequently, a valid, feasible means for remote nutrition malnutrition assessments without physical contact is possible, which is of particular importance during times such as the COVID-19 pandemic, where models of practice require adaption.
Studies of hospitalised COVID-19 patients across several countries have established high malnutrition rates (52.7% (Li et al, 2020); 42.1% (Bedock et al, 2020); 37.5% (Rouget et al, 2020); 35% (Wierdsma et al, 2021)). In contrast, a study of 213 patients discharged home and managed in the community found a malnutrition prevalence rate of 6.6% (Di Filippo et al, 2021). Interestingly, Di Filippo et al (2021) found that, independent of hospitalisation, almost 30% had weight loss of >5%, which the researchers considered significant. This is similar to this study's population, with 25% of patients recording weight loss of >5%. Furthermore, Di Filippo et al (2021) note that patients who lost weight had not yet returned to initial body weight at the follow-up visit (at 30 days), and the researchers highlight the need for nutritional evaluation and counselling or intervention—not only at diagnosis or for those hospitalised, but also for those managed at home.
Among those with higher BMIs, slowing weight loss is also important, even if weight regain would not be a goal. There is a need to be cognisant that those with high BMIs are also at risk of malnutrition. A prospective cohort study conducted in Australian and New Zealand hospitals in 2010 (Agarwal et al, 2019) revealed that one in eight hospitalised obese people were malnourished. Among obese and overweight patients (Kee et al, 2012), higher energy requirements of between 2000–3000 kcal/day (8400–12600 kJ/day) are needed simply for maintenance. Health professionals need to be alert to size bias (Nowicki et al, 2009) when providing care, including active nutritional support for overweight or obese individuals.
A COVID-19 resource that includes nutritional management in the community and a malnutrition pathway was endorsed by the British Dietetic Association, the Royal College of Nursing and the British Association of Parenteral and Enteral Nutrition (2022). Although there are no guidelines for community settings in Australasia, nutrition management guidelines exist for nutrition management in critically and acutely unwell hospitalised patients with COVID-19 in Australia and New Zealand (Australasian Society of Parenteral and Enteral Nutrition, 2020). These guidelines recommend local pathways, incorporating nutrition algorithms for implementing nutrition care in the ICU and on the medical (infectious disease or COVID-19) ward and specify the importance of malnutrition screening, weighing patients where able and optimising management of clinical symptoms. Of note is the fact that both the British and Australasian recommendations advocate that patients identified at risk of malnutrition be referred to the dietitian and a nutrition intervention protocol commenced, which includes a high-protein, high-energy diet.
Encompassing the broad population, this study identified that most people admitted with COVID-19 had symptoms influencing oral intake. Consequently, it can be recommended that all admitted COVID-19 patients be supported nutritionally and commence a high-protein, high-energy diet as standard practice. Additionally, if COVID-19 patients were managed at home, this study supports routine screening of nutrition impact symptoms and malnutrition being incorporated into standard monitoring practices, with appropriate action plans, nutrition advice and dietitian referral actioned if deemed necessary.
The difficulties with remote consultations for COVID-19 patients remain a challenge. These include remote malnutrition screening, remote malnutrition assessment without physical/body composition assessments, difficulties in weighing and weight monitoring of patients in isolation, and challenges organising appropriate dietetic care in a hospital or community setting. These limitations are common themes also generally experienced in this study population. Dietitian consultation and ongoing review was required for 12% (approximately one in eight) of patients with COVID-19. The telephone consultations and availability of electronic medical records enabled remote care provision, ensuring staff safety and minimising the use of PPE.
Limitations
Study limitations include the inability to physically assess body composition and, therefore, complete the SGA in person. Limited repeat weights and weight monitoring during admission is also a limitation. The authors were also unable to always determine if a weight recorded in the EMR was an actual, reported or estimated weight. Frequent and accurate weight monitoring is an important component of nutritional assessment, evaluation and monitoring. The degree to which the nutrition impact symptom incidence can be extrapolated to different COVID-19 strains subsequently identified since study completion is unknown.
Conclusion
Nutrition-impacting symptoms and reduced oral intake are common during the COVID-19 disease trajectory, in both mild to moderate as well as severe cases, during both the acute and recovery phases. This study established that almost all people who had tested positive to COVID-19 had at least one symptom influencing their nutritional intake. T herefore, incorporating regular monitoring of both nutrition impact symptoms that may interfere with intake, and a patient's actual nutritional intake, is of utmost importance, including among those in the community, as are improved ways of delivering nutrition support early in care.
Key points
- Of all people diagnosed with COVID 19, four out of five had at least one symptom that affected their oral intake
- Nutrition screening for all with COVID-19 is important, including for those receiving community or home care
- Incorporating simple nutrition screening tools can guide appropriate action plans, nutrition advice and dietitian involvement if deemed necessary
CPD reflective questions
- Why is nutrition important when diagnosed and recovering from COVID-19?
- How do you build nutrition screening into everyday practice?
- What resources can assist nurses to provide dietary advice during recovery?