Association between preoperative malnutrition and hospital length of stay in adults with major lung resection: A prospective cohort study.

BACKGROUND

Postoperative infections remain a frequent complication after lung resection, and malnutrition is highly prevalent in this population. However, evidence on the impact of malnutrition on clinical outcomes following major lung surgery is limited. This study evaluated the association between preoperative malnutrition and hospital length of stay and postoperative infections in patients undergoing lung resection.

METHODS

A prospective cohort study was conducted. Nutritional status was assessed preoperatively using the seven-point Subjective Global Assessment (7p-SGA), classifying patients as malnourished or normal nutritional status. Body composition was measured by bioelectrical impedance analysis, and muscle function by handgrip strength. The primary outcome was hospital length of stay and postoperative infection the secondary outcome. Multivariable regression models were used.

RESULTS

A total of 62 patients were included; 61.2% were classified as malnourished. Despite comparable body mass index and no significant differences in fat-free mass index or handgrip strength overall, malnourished patients had a significantly longer hospital length of stay than normal nutritional status patients (median 7 vs 4.5 days, P = 0.01). After adjustment, malnutrition was independently associated with increased hospital length of stay in patients undergoing lung resection for infectious disease (β = 4.2 days). Postoperative infections occurred in 23% of patients; malnutrition was associated with a higher adjusted risk of infection (OR = 7.0, 95% CI: 1.3, 36.2, P = 0.02).

CONCLUSION

Preoperative malnutrition is highly prevalent among patients undergoing lung resection and is independently associated with prolonged hospitalization and increased postoperative infection risk. Early nutritional screening and targeted interventions may improve postoperative outcomes in this high-risk population.