A high proportion of pulmonology inpatients are at undernutrition risk or are undernourished despite the increasing awareness of its burden.1–4 Being male,4 older,5 side effects of treatments or drugs as well as eating and swallowing difficulties6 have already been shown to be related to undernutrition risk in hospitalized patients presenting a wide range of diagnoses. However, data regarding the factors associated with undernutrition risk on hospital admission in pulmonology inpatients is still required.

Pulmonology patients, such as chronic obstructive pulmonary disease (COPD) or lung cancer (LC) patients, frequently experienced breathing difficulties that can cause loss of appetite and decreased nutritional intake.7–9 Moreover, the state of inflammation of those patients contributes to body mass depletion,7,8 highlighting pulmonology patients as prone to be at undernutrition risk.

Undernutrition revealed to be associated with a prolonged length of hospital stay (LOS) and increased risk of morbidity and mortality in a varied sample of hospitalized patients.10 It has also been shown, in mixed inpatients samples, that undernutrition contributes to increased costs for health care6,11 and social services provision.12

Among pulmonology patients, undernutrition has been linked to poor prognosis and recurrent hospitalization.2,13–15 However, knowledge regarding the association of undernutrition with increased LOS and worse prognosis in pulmonology department inpatients was obtained from bivariable comparative analysis6,16,17 or logistic regression models,12,15 which did not take into account all the parameters that may be related to LOS and discharge destination. Some studies that included pulmonology patients reported the use of survival analysis. However, patients were all analyzed in a single group without stratified analysis for pulmonology patients10,18 or the analysis was related to a particular group of pulmonary disease patients.2,13,14 In fact, evidence regarding the nutritional factors associated with LOS and discharge destination among pulmonology department inpatients is scarce, which reinforces the need for further research.

Increasing knowledge regarding the consequences of undernutrition among pulmonology patients would be of major relevance. Predicting LOS is relevant for hospital management since it would allow hospitals to organize their resources accordingly. Moreover, it is essential for the development of a more effective health care plan.19 Shortening LOS would also reduce health care costs. Additionally, the early identification of patients’ health status and the implementation of an adequate treatment would reduce the risk of nosocomial infections and could lead to an improvement in patients’ quality of life.20

Several tools have been developed to assess undernutrition risk. Nonetheless, there is current knowledge that undernutrition screening tools differ in their performance,21 and there is no gold standard defined for identification of undernutrition risk17 and no consensus about the most appropriate tool for undernutrition screening neither for hospitalized patients nor specifically for pulmonology inpatients. Actually, a recent systematic review and meta-analysis has revealed a lack of standardized methods to assess undernutrition risk and undernutrition in pulmonology inpatients.22

The Malnutrition Universal Screening Tool (MUST) was developed by the British Association for Parenteral and Enteral Nutrition and was validated in a hospital setting.23,24

Therefore, this study aimed to quantify the association of undernutrition risk evaluated using the MUST at hospital admission with time to discharge alive.

This sample was composed of 683 patients, of which the majority were men (67.2%) with a mean (SD) age of 63.2 (16.3) years. Among the studied sample, 16.5% and 18.3% of the patients presented moderate and high undernutrition risk, respectively (34.8% of undernutrition risk).

A higher proportion of men were observed among moderate and high undernutrition risk categories (p=0.001). Across the gradient of undernutrition risk (low to high undernutrition risk), there was a decrease in the proportion of married participants and an increase of the proportion of single individuals (p=0.009). Also, an increase in the proportion of unemployed individuals was observed in the high undernutrition risk category (p<0.001) (Table 1).

There are significant differences concerning LOS and discharge destination according to undernutrition risk categories. Patients with high undernutrition risk had longer LOS (p=0.006) and among those, it was verified a higher proportion of patients who were not discharged to home (p=0.030), compared with the other groups.

In the survival analysis, we observed that patients with high undernutrition risk on admission had higher probability of remaining in the hospital at a certain point in time (p<0.001; Fig. 1a). Similarly, LC patients had lower probability of hospital discharge (p<0.001). Kaplan–Meier curves for COPD, pneumonia, bronchiectasis and other diagnoses seemed to be overlapped (Fig. 1b). Patients with high undernutrition risk had a longer LOS, with approximately 50% of patients discharged to home only after 14 days of hospitalization (p<0.001; Fig. 1a). On the other hand, patients with low and moderate undernutrition risk were discharged to home after a median of 10 and 13 days, respectively.

Figure 1.

Kaplan–Meier estimates of discharge-free survival of pulmonology inpatients according to undernutrition risk (a) and diagnosis (b). Higher values of discharge-free survival represent a lower probability of hospital discharge at a certain time point; in-hospital deaths, transfers and discharge against medical advice were censored at time of those events. Length of hospital stay was censored at 30 days.

(0.11MB).

Through Cox regression, we verified that the presence of high undernutrition risk at hospital admission was associated with a decreasing probability of discharge alive over time, regardless of sex, age, diagnosis, marital status and working status (adjusted HR=0.70; 95% CI: 0.55–0.90) (Table 2). Lung cancer patients were also shown to have a lower probability of being discharged to home.

In the present study, approximately one in three patients admitted to this pulmonology department was at undernutrition risk. Our results revealed that undernutrition risk assessed by MUST is associated with both LOS and discharge destination among pulmonology inpatients.

The prevalence of undernutrition risk found in the present analysis is similar to previous results (33.8%) of inpatients from the same department where this study has been conducted.29 However, higher frequencies have also been reported in other settings.4,30 A recent study conducted in Norway among respiratory diseases patients using the Nutritional Risk Screening—2002 (NRS-2002) showed a higher prevalence of undernutrition risk (43.5%).30 In another study carried out in a Chinese pulmonology unit, also using NRS-2002, 55.9% of patients were at undernutrition risk.4 These results reveal a high, although variable, prevalence of undernutrition risk among these patients across different settings, geographical areas and using different undernutrition screening tools. Similarly to a recent systematic review, MUST revealed to have good predictive validity for both LOS and mortality among adult inpatients.21

According to our results, the probability of discharge alive over time decreased with the presence of high undernutrition risk on admission. This is in agreement with previous studies, involving inpatients from a wide variety of medical wards31 and pulmonology inpatients.2,10,13,18 The association of undernutrition with in-hospital mortality was reported among acute care patients with respiratory diseases.10 Survival analyses, performed among tuberculosis patients2,13 using either MUST or Malnutrition Screening Tool, revealed that patients with high undernutrition risk had a lower probability of survival. This reinforces our findings, since a lower probability of being discharged to home among patients at high undernutrition risk was found.

Furthermore, through Cox analysis, it was shown that high undernutrition risk maintained its association with time to discharge, even after adjusting for sex, age, marital status, working status and diagnosis. LC patients were less likely to be discharged alive over time. These findings are similar to those previously reported in non-small cell LC outpatients, in whom the presence of undernutrition was associated with lower overall survival.14

This is a relevant study on the association of undernutrition risk with adverse clinical outcomes amongst pulmonology department inpatients. To the best of our knowledge, evidence in pulmonology inpatients is scarce and it has been mainly focused on specific patients groups and the use of different screening tools makes it difficult to make comparisons. In the majority of studies survival analysis was not carried out,6,12,15–17 and its use can be regarded as a study strength.

The concurrent validity of MUST at hospital level was previously studied.24 MUST was shown to be a reliable and useful tool for predicting adverse outcomes in tuberculosis patients13 and also in line with results from inpatients from a wide spectrum of medical wards.26,31–33

This study aimed to evaluate the association between the nutritional parameters assessed on hospital admission and time to discharge alive.

As patients with increased disease severity would certainly be those with the greatest need of hospitalization34 and as the undernutrition risk increases with disease severity,35 the inclusion of this information would be valuable for forthcoming studies, since it was unavailable for the present study. Furthermore, education level would be relevant for socio-demographic characterization36 and because this information was unavailable, the possibility of confounding cannot be discarded as it was reported that less educated patients are more likely to be undernourished.37 It is noteworthy that in some cases information on weight was reported by patients; thus, our results could be influenced by recall bias.

Additionally, selection bias could occur since patients were selected from a tertiary hospital. Nonetheless, our findings could be generalized to pulmonology inpatients with identical characteristics as those included in this study.

The major strengths of this study were the evaluation of a large and varied sample of patients from a pulmonology department and from the same geographical area, allowing the generalizability of these results for patients with similar characteristics.

The findings of this study support the need of early nutritional screening and assessment and consequently implementation of nutritional support. Further studies are still required to study specific diagnoses in order to improve nutritional status of hospitalized patients. In our study we only used data on undernutrition screening, thus it will be also relevant to use data on the diagnosis of undernutrition in future studies.

This study reveals that patients at high undernutrition risk have longer LOS and a lower probability of being discharged to home, regardless of sex, age, marital status, working status and diagnosis. Lung cancer patients had lower probability of being discharged to home, which corroborates a worse prognosis for these patients. Hospitals should follow the guidelines for nutritional screening and assessment as early nutritional interventions can improve nutritional status and therefore improve clinical outcomes.

The authors have no conflicts of interest to declare.