Health care costs of injury in the older population: a prospective multicentre cohort study in the Netherlands

Globally, the number of older people will increase substantially in the coming decades [1]. The proportion of the population over 65 years old in the Netherlands is expected to increase from 18.8% in 2018 to 26% in 2040 [2]. Moreover, the proportion of those aged 80 years and older is expected to double (4.5% in 2018 [3] to 8.7% in 2040 [4]). At present, nearly 50% of all trauma patients in the Netherlands are 65 years or older and more than 25% are older than 80 years [5]. Additionally, an increase in injury in the older population can be expected because of the independence and active lifestyles of older people. Since people as they age become more vulnerable and are more likely to experience comorbidities, they are at greater risk of adverse health outcomes [6, 7].

Compared to the younger trauma population, older patients have longer hospital stays, have higher risk of complications, have higher health care consumption after discharge and are more likely to die due to their injury [5, 7, 8]. The high incidence of comorbidities and frailty in the older population are factors contributing to these outcomes [6, 8]. As a consequence, the need for long-term care and the associated health care costs for this population are increasing, presenting a high economic burden to individuals and society.

Several studies examined the health care costs of adults trauma patients [9‐11]. In the older trauma population, fewer costs of all cause trauma studies exist. Previous research on costs of injury in the older population has mainly focused on falls [12‐14]. Other studies have focused on costs of specific injury types like hip fractures [15‐17] and traumatic brain injury [18]. In general, studies have demonstrated that frailty [19‐21] and comorbidity [22] are associated with increased health care utilization and costs. However, studies within the trauma population are scarce. Two studies within the older trauma population found no association between frailty status and hospital costs [23, 24] The majority of costs of injury studies present only intramural health care costs, despite the fact that especially the long-term extramural health care costs can be high for the older injury population.

In this study we provided a detailed overview of non-surgical health care costs of the older trauma population in the Netherlands for the whole spectrum of injuries. We aimed to assess short-term, in-hospital and long-term, post-hospital health care costs and to identify determinants of health care consumption and costs of trauma patients aged ≥65 years.

With the ageing population, the number of older trauma patients has increased. There is limited research available on the associated health care costs of this group. In this study, we described the characteristics of Dutch trauma patients older than 65 years old and their associated health care costs up to two years after injury. We found that mean costs per patient were €12,190. Older age, higher ISS, being frail and having comorbidities were associated with higher in-hospital and post-hospital costs.

Of mean total health care costs, 80% were related to leisure-related injuries, which consisted mainly of falls. The economic burden of falls in the older population is demonstrated by several other studies [12‐14]. Falls are the most important cause of trauma in older people, especially women, and therefore a relevant topic for strategies for prevention. Apart from this substantial group of frail older people there are also older people, especially males, sustaining traffic or sport-related injuries. In the Netherlands, there is an increase in older road users. There is for example a growing use of electric bicycles by older adults [42]. This has resulted in an increase of the number of traffic related injuries among these older people [42, 43]. Additionally, older road users are at a higher risk of serious injury compared to younger road users, due to functional limitations and physical vulnerability [44]. A focus on the prevention of injury for older road users is therefore also advised.

Frailty was associated with in-hospital and post-hospital health care costs. After controlling for other factors including age, this association was weaker but still significant. Several other studies found a relationship between frailty and health care costs but most of these studies were cross-sectional and none were specifically focused on the older trauma population [19, 45]. Two studies in the older trauma population found no association between frailty status and hospital costs, contrary to our results [23, 24]. It is possible that the different outcome of these studies is related to the services included in the in-hospital costs, as in our study surgical costs were not included. Another possibility for the different outcomes are differences in the definition of a frail trauma patient used in the studies. Multiple frailty screening instruments and other frailty measurements exist, but only few are specific to trauma. Future research is needed to validate more existing instruments in the trauma population [46].

Women were more often admitted with a hip fracture than men and patients with hip fractures had higher post-hospital health care costs, as other studies have also shown [12]. The difference between non-surgical in-hospital costs of patients with and without a hip fracture was less profound. However, patients with hip fractures were more likely to have a surgical intervention. Mean health care costs were highest in ISS 9–15 group; however, the results of the analyses imply that these costs can partly be attributed to other factors, such as prolonged rehabilitation of injury and care needed due to their pre-existing comorbidities and/or frailty.

In the Netherlands, health insurance is mandatory, and both short-term and long-term medical care are covered by this insurance. Consistent with other studies, total health care costs were higher with older age and female sex [47, 48]. The increase in costs by age was mainly caused by post-hospital costs like homecare. Although differences in non-surgical in-hospital costs were small between men and women, we found that women had higher post-hospital costs. This can potentially be attributed to women having a higher life expectancy, more often being frail and with comorbidities and being more likely to sustain a hip fracture. Additional to the fact that we looked at the health care costs of the whole spectrum of injuries in the older trauma population in the Netherlands, two major strengths of the BIOS study were the high number of older patients who participated and the detailed information on both in-hospital and post-hospital costs and pre-injury characteristics. This study also had several limitations. Firstly, several assumptions had to be made in the calculations of costs. Mean total costs per case were calculated for all patients with available hospital or ICU costs. If other costs were missing, these were imputed with €0, which could have resulted in an underestimation of the true total costs. Additionally, in the interpretation of our results it is important to note that in-hospital costs did not include costs of surgical intervention. This has resulted in an underestimation of the in-hospital costs, specifically for patients with injuries like hip fractures, with a high percentage of surgical intervention. Secondly, the instructions of the iMCQ specifically stated to only report health care utilization related to the trauma. However, some of the reported post-hospital health care consumption may have been because of factors other than the injury. This may have led to an overestimation of the post-hospital costs. Thirdly, there is a possibility for non-response bias, since the non-responders were significantly older and more often female. This also suggests an underestimation of true costs, since both age and being female are associated with higher health care costs. Lastly, all variables had < 5% missing values except for GFI items, which were used to determine the level of frailty. The high mean health care costs of respondents with missing frailty data may indicate that older and more vulnerable patients were less likely to complete the GFI questionnaire.

We conclude that the economic burden of older trauma patients is substantial. These high costs are mainly caused by high post-hospital health care consumption. This study showed that both fall-related injuries and traffic-related injuries are important areas for prevention of injuries in the older population.