Nutritional Risk, Health Outcomes, and Hospital Costs Among Chinese Immobile Older Inpatients: A National Study.

ABSTRACT

Purpose: Evidence of the impact of nutritional risk on health outcomes and hospital costs among Chinese older inpatients is limited. Relatively few studies have investigated the association between clinical and cost outcomes and nutritional risk in immobile older inpatients, particularly those with neoplasms, injury, digestive, cardiac, and respiratory conditions. Methods: This China-wide prospective observational cohort study comprised 5,386 immobile older inpatients hospitalized at 25 hospitals. All patients were screened for nutritional risk using the Nutrition Risk Screening (NRS 2002). A descriptive analysis of baseline variables was followed by multivariate analysis (Cox proportional hazards models and generalized linear model) to compare the health and economic outcomes, namely, mortality, length of hospital stay (LoS), and hospital costs associated with a positive NRS 2002 result. Results: The prevalence of a positive NRS 2002 result was 65.3% (n = 3,517). The prevalence of "at-risk" patients (NRS 2002 scores of 3+) was highest in patients with cardiac conditions (31.5%) and lowest in patients with diseases of the respiratory system (6.9%). Controlling for sex, age, education, type of insurance, smoking status, the main diagnosed disease, and Charlson comorbidity index (CCI), the multivariate analysis showed that the NRS 2002 score = 3 [hazard ratio (HR) 1.376, 95% CI 1.031-1.836] were associated with approximately a 1.5-fold higher likelihood of death. NRS 2002 scores = 4 (HR 1.982, 95% CI 1.491-2.633) and NRS scores ≥ 5 (HR 1.982, 95% CI 1.498-2.622) were associated with a 2-fold higher likelihood of death, compared with NRS 2002 scores <3. An NRS 2002 score of 3 (percentage change 16.4, 95% CI 9.6-23.6), score of 4 (32.4, 95% CI 24-41.4), and scores of ≥ 5 (36.8, 95% CI 28.3-45.8) were associated with a significantly (16.4, 32.4, and 36.8%, respectively) higher likelihood of increased LoS compared with an NRS 2002 scores <3. The NRS 2002 score = 3 group (17.8, 95% CI 8.6-27.7) was associated with a 17.8%, the NRS 2002 score = 4 group (31.1, 95% CI 19.8-43.5) a 31.1%, and the NRS 2002 score ≥ 5 group (44.3, 95% CI 32.3-57.4) a 44.3%, higher likelihood of increased hospital costs compared with a NRS 2002 scores <3 group. Specifically, the most notable mortality-specific comorbidity and LoS-specific comorbidity was injury, while the most notable cost-specific comorbidity was diseases of the digestive system. Conclusions: This study demonstrated the high burden of undernutrition at the time of hospital admission on the health and hospital cost outcomes for older immobile inpatients. These findings underscore the need for nutritional risk screening in all Chinese hospitalized patients, and improved diagnosis, treatment, and nutritional support to improve immobile patient outcomes and to reduce healthcare costs.