Hospitalization costs of injury in elderly population in China: a quantile regression analysis

Around 5.8 million people die from trauma each year, making it one of the major causes of death worldwide [1]. By 2030, there will be a rise in the number of injury-related fatalities worldwide, notably in low- and middle-income countries (LMICs) [2‐4]. In 2019, China accounts for 18% of the world's population, with 164.5 million people aged 65 and older (65 +) and 26 million people aged 80 and older (80 +), and the proportion of people aged 65 and older is expected to be as high as 30% in 2050 [5, 6]. China has an aging population, and as it does, the strain on the country's current family and public healthcare systems will only become worse. Older folks' falls are a common and raising public health issue that can result in severe injuries (such as head injuries and hip fractures) [7]. It is estimated that approximately 30% of people aged 65 and over fall each year [8, 9] and an increasing number of elderly people suffer from hip fractures [10, 11]. By 2050, the total number of hip fractures in the elderly is expected to be 1.3 million in China [12]. Between 2015 and 2019, the older population is more likely to have a traffic injury, especially if they are over 45, and there is a rise in the mortality and morbidity linked to such injuries [13]. A study has demonstrated that multimorbidity is linked to increased health care utilization and out-of-pocket expenditure among the aged population over the age of 45 in China [14]. As a result, the above-described problems, which are primarily relevant to the elderly, will saddle China with a significant medical and budgetary burden in the future.

Previous studies on the cost of injury in the elderly population have focused more on falls [15‐17], hip fractures [18‐20] and head injuries [21, 22]. Traditional regression methods such as ordinary least squares (OLS), generalized linear models (GLM) and multiple linear regression have been mainly applied in previous cost studies [16, 17, 23], which are more concerned with overall mean effects, not robust to statistical outliers and lack flexibility in analyzing health care costs. Quantile regression (QR) models, which concentrate on the conditional quartiles of the cost distribution [24, 25], may be the best option for directing targeted cost-effective policies and early interventions, and can also give analysts the flexibility to examine trauma costs predictors corresponding to quantiles of interest.

In this study, we shed light on the hospitalization costs of the elderly trauma patients in China across the spectrum of injuries. Our aim is to assess the hospitalization costs for the trauma patients, non-surgical patients and surgical patients and to identify determinants of hospitalization costs through QR. In addition, understanding the magnitude of hospitalization costs for trauma patients aged ≥ 65 years is essential for providing evidence for cost-effective interventions and reference for health insurance policies.

Trauma has steadily developed into a global health issue in the elderly population as a result of population aging, with consequences such as disability and death. Our study uses QR to clarify the diverse impacts of influencing factors on hospitalization costs and primarily describes the features of the elderly trauma group over 65 years old. According to the results of our quantile regression analysis, trauma patients have greater hospitalization costs when they have LOS, surgery, ISS (16–24), multimorbidity, thorax injury, and blood transfusion. More importantly, our results suggest that the impacts of LOS, surgery, ISS (16–24), multimorbidity, thorax injury and blood transfusion on hospitalization costs were not constant across all the quantiles. Costlier hospitalizations for non-surgical patients are related to LOS, ISS (16–24), and multimorbidity. In-hospital expenses were particularly high in patients with significant injury and multimorbidity, according to a similar research conducted in the Netherlands [23]. Actually, it is not surprising that LOS and surgery were associated with higher hospitalization costs. The length of stay (LOS) is an indicator of quality of health care [35], and a rise in LOS indicates an increase in the consumption of medical resources. The expense of hospitalization could be decreased by lowering the LOS. On November 19, 2021, the National Healthcare Security Administration released the notice of the three-year action plan for the reform of the DRGs (Diagnosis Related Groups) and DIP (Diagnosis-Intervention Packet) payment modes [36]. The country-wide rollout of DRG and DIP is intended to optimize health resources, rationally regulate hospital costs, scale back on excessive LOS, and bolster the efficiency of the health insurance fund. In China, the LOS of C-DRGs (the Chinese diagnosis-related group) patients with hip fracture have been on the decline, according to a recent study on Sanming City [37]. Future research will continue to focus on the impact of DRGs/DIP on LOS without compromising quality of care, as shorter LOS may lead to inadequate treatment, more complications, or more readmissions.

Hospitalization costs are impacted differently by various therapeutic interventions. The average cost of hospitalization was ¥20,741 ($3,072) per patient. In contrast to the average hospitalization cost of €5430 ($6,134) in a Netherlands study [23], the average hospitalization cost for non-surgical patients was ¥13,647 ($2,021). In terms of surgical patients, average hospitalization costs for surgical patients are greater than for non-surgical patients (¥44,936 ($6,655) VS ¥13,647 ($2,021)). The quantile regression results illustrate significant associations of LOS, ISS (16–24), and hip and thigh injuries with higher hospitalization costs. And these effect values differed significantly across all quantiles. The average hospitalization cost for patients who underwent surgery for hip and thigh injuries was ¥56,839 ($8,418). Palais et al. showed that the mean hospitalization costs of patients who received hip replacement was $22,076 in 2018 [38]. In our study, the average hospitalization cost for all trauma patients with hip and thigh injuries was ¥27,875 ($4,129). According to a study conducted in Ireland in 2022, the mean hospitalization cost of patients with hip fractures was approximately €11,700 [19]. The cost of treating hip injuries differs across nations due to treatment practices, health insurance systems, hospital protocols and socio-demographic factors, however it is obvious that hip fractures incur significant medical expenses. The number of elderly people undergoing complex surgeries will rise as the population ages and medical technology continues to progress, and the entire healthcare system will face challenges in caring for these patients who may have high needs and high costs [39].

On the other hand, in contrast to ordinary least square regression, our study also discovered and confirmed that quantile regression, because of its capacity to evaluate each factor's impact at any point of the hospitalization costs distribution, can provide more detailed and comprehensive information on potential correlation when exploring the factors of hospitalization costs. Other factors like usage of the ventilator, rescue, transferring to other departments, intensive care, admission to ICU, and ISS(≥ 25) were associated with increased hospitalization costs in a specific or extreme (low or high) quantile. In addition to LOS, surgery, ISS (16–24), having multimorbidity, having a thorax injury, and receiving blood transfusions, other factors also indirectly reflect the severity of the patient's disease and the patient's critical status. This suggests that when applying DRGs and DIPs, it is more worthwhile for policy makers and researchers to consider not only the disease itself and the treatment of the disease, but also factors and specific information that indirectly affect hospitalization costs, such as the severity of injury in the elderly trauma population. Therefore, in order to more effectively lower hospitalization costs, the DRG/DIP payment standard may be dynamically adjusted as needed at any moment.

At the clinical level, reference to factors affecting hospitalization costs provides a theoretical basis for rational cost control. At the medical management level, consideration of hospitalization costs and influencing factors of the elderly trauma population provides some reference for relevant medical policy proposals.

In our study, we discovered that falls were the most common cause of trauma, and the average hospitalization cost for falls was ¥19,809 ($2,934), which was less expensive than the average hospitalization cost of $29, 562 for patients with nonfatal falls in the United States [40] and the range of $5,654 to $42,840 per fall-related hospitalization in the study by Heinrich et al. [41]. Age-related brain shrinkage affects the parts of older people's brains connected to physical activity [42, 43], and age-related losses in active muscle strength and cognitive function raise the risk of falls and fractures in older people. Additionally, traffic accidents are another major external cause of trauma. The majority of the elderly population in our study was more vulnerable to traffic accidents while walking or bicycling because they may have reduced peripheral vision, reaction activity, balance, and coordination [44, 45] and are more likely to have head and thorax injuries. Exercise instruction and routine screening for pertinent risk factors in the elderly population should be enhanced from a prevention aspect, and when needed, personalized fall prevention programs should be implemented [46]. Age-related pedestrian safety awareness should be improved by public health initiatives and safety education on the use of bicycle is necessary.

Our findings indicated that the major contributor to hospitalization costs was the cost of the medicine. In recent years, our country has implemented regulations to limit the medicine costs. As a result of the measures, the percentage of medicine costs has dropped from 41.93 percent to 26.34 percent, according to the findings of our study. Therefore, it is essential to further strengthen the supervision of rational medicine and standardize medication. We found a slight increase in the supplies cost. In order to cut medical costs and financial burdens for patients with fractures or trauma in the future, China is also actively developing and implementing centralized bulk procurement of high-value medical consumables.

To our knowledge, this is the first study in China to use quantile regression to examine the impacts of trauma on hospitalization costs in elderly population. This study also had several limitations. First of all, because the study group was restricted to a single hospital, it is impossible to extrapolate the findings to other areas due to potential bias caused by disparities in injury patterns and severity. More extensive studies are needed in the future. Secondly, we examined only hospitalization costs because hospitalization costs account for a significant portion of total injury costs. However, we study patients who are hospitalized for the first time for an injury, and some patients may be transferred to other hospitals without having their medical costs fully determined. This may underestimate the true total cost of the injury. Future studies should concentrate more on post-hospital health care costs and indirect costs (e.g., transportation, lost productivity) in order to fully assess the financial burden on the elderly trauma population. Thirdly, we only studied injuries based on anatomical sites, but the study of poisoning, and certain other consequences of external causes cannot be ignored. In the future studies, we further discuss their different categories of trauma and different causes of trauma separately. Finally, future studies could further expand the sample size to explore the financial burden of hospitalization costs for the elderly trauma population under different economic conditions.

This study can effectively identify and measure the association between injuries and hospitalization costs at different quartiles in the elderly population by employing QR. Hospitalization costs in elderly patients with injuries were mainly associated with LOS, surgery, ISS (16–24), multimorbidity, thorax injury, and blood transfusions. LOS, ISS (16–24) and multimorbidity were positively associated with hospitalization costs for elderly non-surgical patients. Hospitalization costs for geriatric surgical patients were chiefly associated with LOS, ISS (16–24) and hip and thigh injuries. These findings provide support for scientific investigation of the factors influencing hospitalization costs and inform policy makers regarding the reduction of injury-related hospitalization costs in the elderly population.